CDC16xxF-E
Automotive Controller
Family User Manual
CDC1605F-E
Edition March 31, 2003
6251-606-2AI
ADVANCE INFORMATION
MICRONAS
Automotive Controller
Emulator
Specification
CDC16xxF-E ADVANCE INFORMATION
2March 31, 2003; 6251-606-2AI Micronas
Contents
Page Section Title
5 1. Introduction
5 1.1. Features
9 1.2. Abbreviations
11 2. Package and Pins
11 2.1. Pin Assignment
13 2.2. Package Outline Dimensions
14 2.3. Multiple Function Pins
14 2.4. Pin Function Description
17 2.5. External Components
18 2.6. Pin Circuits
20 3. Electrical Data
20 3.1. Absolute Maximum Ratings
21 3.2. Recommended Operating Conditions
22 3.3. Characteristics
28 3.4. Recommended Crystal Characteristics
29 3.5. Flash/EMU Port Characteristics
32 4. CPU and Clock System
32 4.1. W65C816
33 4.2. Operating Modes
36 4.3. Clock System
38 4.4. EMI Reduction Module (ERM)
40 5. Memory and Boot System
40 5.1. RAM and ROM
42 5.2. Memory Banking
45 5.3. Boot System
49 6. Core Logic
49 6.1. Control Register CR
50 6.2. Reset Logic
55 6.3. Standby Registers
56 6.4. Test Registers
57 7. Multiplier
58 7.1. Functional Description
58 7.2. Registers
58 7.3. Operation of the Multiplier
59 8. Power-Saving Module (PSM)
60 8.1. Functional Description
61 8.2. Registers
66 8.3. Operation of Power-Saving Module
69 8.4. Operation of RTC Module
71 8.5. Operation of Polling Module
71 8.6. Operation of Port Wake Module
Contents, continued
Page Section Title
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 3
73 9. Memory Patch Module
73 9.1. Principle of operation
74 9.2. Registers
76 10. Interrupt Controller (IR)
76 10.1. Principle of Operation
78 10.2. Registers
80 10.3. Interrupt Assignment
83 10.4. Interrupt Timing
85 10.5. Port Interrupt Module
87 11. Ports
87 11.1. Analog Input Port 0
88 11.2. Universal Ports U1 to U7
90 11.3. Universal Port Registers
92 11.4. High Current Ports H0.0 to H3.5
93 11.5. High Current Port Registers
94 12. A/D Converter (ADC)
95 12.1. Operation
96 12.2. Registers
97 13. Timers (TIMER)
97 13.1. Timer T0
99 13.2. Timer T1 and T2
101 14. Pulse Width Modulator (PWM)
101 14.1. Principle of Operation
102 14.2. Registers
103 15. Capture Compare Module (CAPCOM)
104 15.1. Principle of Operation
106 15.2. Registers
108 16. Stepper Motor Module (SMM)
108 16.1. Functional Description
110 16.2. Registers
110 16.3. Principle of Operation
112 16.4. Rotor Zero Position Detection (RZPD)
114 17. LCD Module
114 17.1. Principle of Operation
118 17.2. Registers
118 17.3. Software Hints for Cascading LCD Modules
119 18. DMA
119 18.1. Principle of Operation
120 18.2. Registers
123 18.3. Ports
123 18.4. SW Application Hints
125 18.5. Timings
CDC16xxF-E ADVANCE INFORMATION
4March 31, 2003; 6251-606-2AI Micronas
Contents, continued
Page Section Title
129 19. Serial Synchronous Peripheral Interface (SPI)
130 19.1. Principle of Operation
131 19.2. Registers
132 19.3. Timing
133 20. Universal Asynchronous Receiver Transmitter (UART)
134 20.1. Principle of Operation
136 20.2. Timing
138 20.3. Registers
140 21. CAN Manual
141 21.1. Abbreviations
141 21.2. Functional Description
147 21.3. Application Notes
152 21.4. Bit Timing Logic
154 21.5. Bus Coupling
156 22. DIGITbus System Description
156 22.1. Bus Signal and Protocol
157 22.2. Other Features
157 22.3. Standard Functions
158 22.4. Optional Functions
159 23. DIGITbus Master Module
159 23.1. Introduction
159 23.2. Context
160 23.3. Functionality
162 23.4. Registers
165 23.5. Principle of Operation
168 23.6. Timings
169 24. Audio Module (AM)
170 24.1. Functional Description
173 24.2. Registers
175 25. Hardware Options
175 25.1. Functional Description
175 25.2. Listing of Dedicated Addresses and Corresponding Hardware Options
180 26. Register Cross Reference Table V2.1
180 26.1. CAN Registers, memory page 1C
181 26.2. I/O Register 1, memory page 1E
183 26.3. I/O Register 0, memory page 1F
186 27. Register Quick Reference
215 28. Differences
218 29. Data Sheet History
ADVANCE INFORMATION
CDC16xxF-E
Micronas
March 31, 2003; 6251-606-2AI
5
1. Introduction
Release Note: Revision bars indicate significant changes to the previous edition.
The IC is a single-chip controller for use in automotive applications. The CPU on the chip is an upgrade of the 65C02 with 16-bit internal data and 24-bit address
bus. The chip consists of timer/counters, an interrupt controller, a multichannel A/D converter, a stepper motor and LCD driver, CAN interfaces and PWM outputs.
1.1. Features
Table 1–1: CDC16xxF Family Feature List
This Docu-
ment:
Item CDC1605F-E
EMU CDC1607F-E
MCM Flash CDC1631F-E
MASK ROM CDC1605F-C
EMU CDC1607F-C
MCM Flash CDC1641F-C
Mask ROM CDC1652F-C
Mask ROM CDC1672F-C
Mask ROM
Core
CPU 16-bit 65C816, featuring software compatibility with its 8-bit NMOS and CMOS 6500-series predecessors
CPU-Active Operation
Modes FAST, SLOW and DEEP SLOW FAST and SLOW
Power Saving Modes
(CPU Inactive) WAKE and IDLE -
EMI Reduction Mode selectable in FAST mode
Oscillators 4 MHz to 12 MHz Quartz, RC 4 MHz to 12 MHz Quartz
RAM 6 KB 2 KB 6 KB 2.75 KB 4 KB 6 KB
ROM ROMless,
external pro-
gram stor-
age with up
to 16 MB,
internal 2 KB
Boot ROM
256 KB Flash,
bottom boot
configuration,
internal 2 KByte
Boot ROM
64 KB ROMless,
external pro-
gram storage
with up to
16 MB, internal
2 KB Boot ROM
256 KB Flash,
bottom boot
configuration,
internal 2 KB
Boot ROM
90 KB 128 KB 216 KB
CDC16xxF-E
ADVANCE INFORMATION
6
March 31, 2003; 6251-606-2AI
Micronas
Multiplier, 8 by 8 bit ✔-
Digital Watchdog ✔
Central Clock Divider ✔
Interrupt Controller
expanding NMI 16 inputs,16 priority levels
Port Interrupts including
Slope Selection 4 inputs
Port Wake-Up Inputs
including Slope / Level
Selection
✔-
Patch Module 10 ROM locations 5 ROM loca-
tions 10 ROM locations 5 ROM loca-
tions 6 ROM locations
Boot System allows in-system downloading of
code and data into RAM via
serial link
- allows in-system downloading of
code and data into RAM via serial
link
---
Analog
Reset/Alarm Combined Input for Regulator Input Supervision
Clock and Supply Super-
vision
✔
10-bit ADC, charge bal-
ance type 9 channels (5 channels selectable as digital input)
ADC Reference VREF Pin
Comparators P06COMP with 1/2 AVDD reference
LCD Internal processing of all analog voltages for the LCD driver
Table 1–1: CDC16xxF Family Feature List, continued
This Docu-
ment:
Item CDC1605F-E
EMU CDC1607F-E
MCM Flash CDC1631F-E
MASK ROM CDC1605F-C
EMU CDC1607F-C
MCM Flash CDC1641F-C
Mask ROM CDC1652F-C
Mask ROM CDC1672F-C
Mask ROM
ADVANCE INFORMATION
CDC16xxF-E
Micronas
March 31, 2003; 6251-606-2AI
7
Communication
DMA 1 DMA Channel for serving the
Graphics Bus interface - 1 DMA Channel for serving the
Graphics Bus interface - 1 DMA Channel for serving the
Graphics Bus interface
UART 3: UART0, UART1 and UART2 1: UART0 3: UART0, UART1 and UART2 1: UART0 3: UART0, UART1 and UART2
Synchronous Serial
Peripheral Interfaces 2: SPI0 and SPI1 1: SPI0 2: SPI0 and SPI1 1: SPI0 2: SPI0 and SPI1
Full CAN modules V2.0B 3: CAN0, CAN1 and CAN2 with
256-byte object RAM each
(LCAN000F)
1: CAN0 with
256-byte object
RAM
(LCAN000F)
3: CAN0, CAN1 and CAN2 with
256-byte object RAM each
(LCAN0009)
1: CAN0 with
256-byte object
RAM
(LCAN0009)
2: CAN0 and CAN1 with 256-byte
object RAM each (LCAN0009)
DIGITbus 1 master module - 1 master module - 1 master module
Input & Output
Universal Ports select-
able as 4:1 mux LCD
Segment/Backplane lines
or Digital I/O Ports
up to 52 I/O or 48 LCD segment lines (=192 segments),
in groups of two configurable as I/O or LCD
Universal Port Slew Rate HW preselectable
Stepper Motor Control
Modules with High-Cur-
rent Ports
5 Modules, 24 dI/dt controlled ports
8-bit PWM Modules 5 Modules: PWM0, PWM1,
PWM2, PWM3 and PWM4 3 Modules:
PWM0, PWM1,
PWM2
5 Modules: PWM0, PWM1, PWM2,
PWM3 and PWM4 2 Modules:
PWM0, PWM1 5 Modules: PWM0, PWM1, PWM2,
PWM3 and PWM4
Audio Module with auto-
decay
✔
SW select. Clock outputs 2
Polling / Flash Timer Out-
put 1 High-Current Port output operable in Power
Saving Mode -
Table 1–1: CDC16xxF Family Feature List, continued
This Docu-
ment:
Item CDC1605F-E
EMU CDC1607F-E
MCM Flash CDC1631F-E
MASK ROM CDC1605F-C
EMU CDC1607F-C
MCM Flash CDC1641F-C
Mask ROM CDC1652F-C
Mask ROM CDC1672F-C
Mask ROM
CDC16xxF-E
ADVANCE INFORMATION
8
March 31, 2003; 6251-606-2AI
Micronas
Timers & Counters
16-bit free running
counters with Capture/
Compare modules
CCC0 with 3CAPCOM
16-bit timers 1: T0
8-bit timers 2: T1 and T2
Real Time Clock, Deliver-
ing Hours, Minutes and
Seconds
✔-
Miscellaneous
Scalable layout in CAN,
RAM and ROM -✔-✔
Various randomly select-
able hardware options Most options software-program-
mable, copy from user program
storage during system start-up
Mask pro-
grammed
according to
user specifica-
tion
Most options software-programma-
ble, copy from user program stor-
age during system start-up
Core Bond-Out ✔-✔-
Supply Voltage 4.5 V to 5.5 V
Temperature Range Tcase: −40 °C to +105 °C Tamb: −40 °C to +85 °C
Package
Type Ceramic
177PGA Plastic 100QFP
0.65 mm pitch Ceramic
177PGA Plastic 100QFP
0.65 mm pitch
Bonded Pins 176 100 176 100
Table 1–1: CDC16xxF Family Feature List, continued
This Docu-
ment:
Item CDC1605F-E
EMU CDC1607F-E
MCM Flash CDC1631F-E
MASK ROM CDC1605F-C
EMU CDC1607F-C
MCM Flash CDC1641F-C
Mask ROM CDC1652F-C
Mask ROM CDC1672F-C
Mask ROM
ADVANCE INFORMATION
CDC16xxF-E
Micronas
March 31, 2003; 6251-606-2AI
9
1.2. Abbreviations
AM Audio Module
CAN Controller Area Network Module
CAPCOM Capture/Compare Module
CPU Central Processing Unit
DMA Direct Memory Access Module
ERM EMI Reduction Module
IR Interrupt Controller
LCD Liquid Crystal Display Module
P06COMP P0.6 Alarm Comparator
PINT Port Interrupt Module
PSM Power Saving Module
PWM 8-Bit Pulse Width Modulator Module
RTC Real time Clock
SM Stepper Motor Control Module
SPI Serial Synchronous Peripheral Interface
T0 16-Bit Timer 0
T1, T2 8-Bit Timers 1 and 2
UART Universal Asynchronous Receiver Transmitter
CDC16xxF-E ADVANCE INFORMATION
10 March 31, 2003; 6251-606-2AI Micronas
Fig. 1–1: Block diagram of CDC1605F-E/CDC1607F-E
65C816
CPU
ROMless
or
256kFlash *
CAN 0 *
CAN 1 *
CAN 2 *
Banking
16-Bit Timer 0
10-Bit ADC
UART 0
8-Bit PWM 2
Audio Module
Power Saving
Module
8-Bit Timer 1
Watchdog
Clock
16 Inputs
Interrupt
Controller
16-Bit
CAPCOM 0
Stepper Motor
Control
LCD Control
6k RAM *
UART 1
UART 2
UPort1UPort2UPort3UPort4UPort5UPort6UPort7
HPort0HPort1HPort2HPort3
8-Bit Timer 2
16-Bit
CAPCOM 1
16-Bit
CAPCOM 2
8-Bit PWM 4
8-Bit PWM 0
PPort0
SPI 0
8-Bit PWM 1
8-Bit PWM 3
Clock Out 0
Clock Out 1
DMA Logic
4
6
8
9
8
8
8
8
8
6
6
6
XTAL1
XTAL2
TEST
RESETQ
VSS
VDD UVDD
UVSS
VREF
AVDD
AVSS
HVDD1
HVSS1
HVDD2
HVSS2
Reset/Alarm
Test
Data-,
address- and
control bus,
Emu only
EVDD
EVSS
SPI 1
* Scalable within wide limits
Patch Module
Boot ROM
DIGITbus
ERM
RTC
RC Oscillator
Multiplier
8 by 8 bit
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 11
2. Package and Pins
2.1. Pin Assignment
Fig. 2–1: Pin Map of CPGA177 Package
92 88 85 83 79 75 71 70 66 64 60 55 51 49 48
93 91 87 84 80 76 74 69 65 61 59 54 50 47 44
95 94 90 86 81 78 73 68 63 58 56 53 46 43 41
99 98 97 89 82 77 72 67 62 57 52 45 42 40 39
104 103 100 96 38 37 36 35
108 105 102 101 33 34 32 31
110 109 107 106 28 29 30 27
114 113 112 111 23 24 25 26
115 118 117 116 18 19 21 22
119 120 122 121 13 14 17 20
123 124 125 126 812 15 16
127 128 130 133 140 145 150 155 160 165 170 1 9 10 11
129 131 134 141 144 146 151 156 161 166 169 174 267
132 135 138 142 147 149 153 157 162 164 168 172 175 3 5
136 137 139 143 148 152 154 158 159 163 167 171 173 176 4
151413121110987654321
A
B
C
D
E
F
G
H
J
K
L
M
N
P
Q
177
151413121110987654321
A
B
C
D
E
F
G
H
J
K
L
M
N
P
Q
Pin 1
A 1
Top View
Top View
CDC16xxF-E ADVANCE INFORMATION
12 March 31, 2003; 6251-606-2AI Micronas
Fig. 2–2: Pin Assignment for CPGA177 Package
Pin Functions Co-
ord.
Pin
No.
Bus
Mode
LCD
Mode
Port
Special Out
Port
Special In
Basic
Function
EADB21 M8 155
EADB20 N8 156
NC P8 157
NC Q8 158
EPH2 Q7 159
EBE M7 160
EVPA N7 161
EVDA P7 162
ESTOPCLK Q6 163
UVSS P6 164
UVDD M6 165
ADB7 SEG3.7 T2-OUT U3.7 N6 166
ADB6 SEG3.6 CC1-OUT U3.6 Q5 167
ADB5 SEG3.5 SPI1-CLK-OUT SPI1-CLK-IN U3.5 P5 168
ADB4 SEG3.4 T0-OUT WP0 U3.4 N5 169
ADB3 SEG3.3 CC2-OUT U3.3 M5 170
ADB2 SEG3.2 DIGIT-OUT DIGIT-IN U3.2 Q4 171
ADB1 SEG3.1 CO1 SPI1-D-IN U3.1 P4 172
ADB0 SEG3.0 SPI1-D-OUT U3.0 Q3 173
NC N4 174
NC P3 175
NC Q2 176
Extra pin for avoiding wrong chip insertion. Connect to System Ground. E5 177
NC M4 1
NC N3 2
NC P2 3
NC Q1 4
SEG6.7 CAN0-TX MULTI_TEST_IN U6.7 P1 5
SEG6.6 PINT1-OUT CAN0-RX/WP1 U6.6 N2 6
SEG6.5 T1-OUT SPI0-D-IN U6.5 N1 7
Connect to System Ground L4 8
SEG6.4 SPI0-D-OUT U6.4 M3 9
TEST M2 10
RESETQ M1 11
XTAL2 L3 12
XTAL1 K4 13
VSS K3 14
VDD L2 15
SEG6.3 SPI0-CLK-OUT SPI0-CLK-IN U6.3 L1 16
SEG6.2 T1-OUT PINT2-IN/WP5 U6.2 K2 17
SEG6.1 LCD-CLK-OUT PINT1-IN/WP4 U6.1 J4 18
SEG6.0 LCD-SYNC-OUT PINT0-IN/WP3 U6.0 J3 19
WEQ SEG1.7 CAN1-TX U1.7 K1 20
CEQ SEG1.6 CAN1-RX/WP2 U1.6 J2 21
ITSTOUT SEG1.5 LCD-CLK-OUT U1.5 J1 22
RWQ SEG1.4 LCD-SYNC-OUT U1.4 H4 23
PH2 BP3 U1.3 H3 24
OEQ BP2 U1.2 H2 25
BE BP1 U1.1 H1 26
RDY BP0 ITSTOUT U1.0 G1 27
EADB17 G4 28
EADB0 G3 29
EDB7 G2 30
EDB6 F1 31
EDB5 F2 32
EDB4 F4 33
EDB3 F3 34
EDB2 E1 35
EDB1 E2 36
EDB0 E3 37
Connect to System Ground E4 38
EVSS1 D1 39
NC D2 40
NC C1 41
NC D3 42
NC C2 43
NC B1 44
EVDD1 D4 45
EOEQ C3 46
ECEQ B2 47
EADB1 A1 48
STOPCLK SMB1+ H1.5 A2 49
VPQ SMB1- H1.4 B3 50
VPA SMB2+ H1.3 A3 51
Connect to System Ground D5 52
VDA SMB2- SMB-COMP H1.2 C4 53
DB7 SME1+/PWM2 H1.1 B4 54
DB6 SME1-/PWM0 H1.0 A4 55
HVDD1 C5 56
HVSS1 D6 57
DB5 SME2+ H0.5 C6 58
DB4 SME2- SME-COMP H0.4 B5 59
DB3 SMA1+ H0.3 A5 60
DB2 SMA1- H0.2 B6 61
DB1 SMA2+ H0.1 D7 62
DB0 SMA2- SMA-COMP H0.0 C7 63
SMD1+ H3.5 A6 64
SMD1- H3.4 B7 65
SMD2+ H3.3 A7 66
Pin
No.
Co-
ord.
Pin Functions
Basic
Function
Port
Special In
Port
Special Out
LCD
Mode
Bus
Mode
154 Q9 EADB22
153 P9 EADB23
152 Q10 U7.0 SEG7.0
151 N9 U7.1 SEG7.1
150 M9 U7.2 GRDQ SEG7.2
149 P10 U7.3 GWRQ SEG7.3
148 Q11 U4.0 CAN2-RX/WP7 SEG4.0 ADB8
147 P11 U4.1 CAN2-TX SEG4.1 ADB9
146 N10 U4.2 UART2-RX SEG4.2 ADB10
145 M10 U4.3 UART2-TX SEG4.3 ADB11
144 N11 U4.4 UART0-RX/WP8 SEG4.4 ADB12
143 Q12 U4.5 UART0-TX SEG4.5 ADB13
142 P12 U4.6 CC2-IN CC1-OUT SEG4.6 ADB14
141 N12 U4.7 CC1-IN SEG4.7 ADB15
140 M11 Connect to System Ground
139 Q13 U5.0 CC0-IN CO1 SEG5.0
138 P13 U5.1 INT-TEST-IN CC0-OUT SEG5.1
137 Q14 U5.2 LCD-CLK-IN AM-PWM SEG5.2
136 Q15 U5.3 LCD-SYNC-IN AM-OUT SEG5.3
135 P14 U5.4 IRQ UART1-TX SEG5.4
134 N13 U5.5 ABORTQ CO0 SEG5.5
133 M12 NC
132 P15 NC
131 N14 NC
130 M13 NC
129 N15 NC
128 M14 U5.6 PINT3/WP6 PWM2 SEG5.6
127 M15 U5.7 PINT3/UART1-RX PINT0-OUT SEG5.7
126 L12 U2.0/GD0 SEG2.0 ADB16
125 L13 U2.1/GD1 SEG2.1 ADB17
124 L14 U2.2/GD2 SEG2.2 ADB18
123 L15 U2.3/GD3 SEG2.3 ADB19
122 K13 U2.4/GD4 SEG2.4 ADB20
121 K12 U2.5/GD5 SEG2.5 ADB21
120 K14 U2.6/GD6 SEG2.6 ADB22
119 K15 U2.7/GD7 SEG2.7 ADB23
118 J14 AVSS
117 J13 AVDD
116 J12 VREF
115 J15 NC
114 H15 P0.1 P0.1 digital input
113 H14 P0.2 P0.2 digital input
112 H13 P0.3 P0.3 digital input
111 H12 P0.4 P0.4 digital input
110 G15 P0.5 P0.5 digital input
109 G14 P0.6 P0.6 Comparator input
108 F15 P0.7
107 G13 P0.8
106 G12 P0.9
105 F14 EADB16
104 E15 EADB15
103 E14 EADB14
102 F13 EADB13
101 F12 EADB12
100 E13 EADB11
99 D15 EADB10
98 D14 EADB9
97 D13 EWEQ
96 E12 Connect to System Ground
95 C15 EADB18
94 C14 EVSS2
93 B15 NC
92 A15 NC
91 B14 NC
90 C13 NC
89 D12 NC
88 A14 NC
87 B13 NC
86 C12 EVDD2
85 A13 EADB19
84 B12 EADB8
83 A12 EADB7
82 D11 EADB6
81 C11 EADB5
80 B11 EADB4
79 A11 EADB3
78 C10 EADB2
77 D10 H2.0 SMC-COMP SMC2-
76 B10 H2.1 SMC2+
75 A10 H2.2 SMC1-
74 B9 H2.3 SMC1+
73 C9 H2.4 WP9 PWM0
72 D9 H2.5/POL PWM4
71 A9 HVSS2
70 A8 HVDD2
69 B8 H3.0 PWM1
68 C8 H3.1 PWM3
67 D8 H3.2 SMD-COMP SMD2-
1
44
132
88
155
66
154
67
177
87
133
45
NC = not connected, leave
vacant
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 13
2.2. Package Outline Dimensions
Fig. 2–3: CPGA177 Ceramic Pin Grid Array 177-Pin (Weight approx. 14g)
PIN No. 1
INDEX
2.540 0.15±
ABCDEFG JKLMNPQ
15
14
13
12
11
10
9
7
6
5
4
3
2
1
28 18 838
52
62
72
82
106 116 12696
170
160
150
140
2.540 x 14 = 35.56 0.2±
2.540 0.15±
2.540 x 14 = 35.56 0.2±
0.4 0.1±
2.1 0.20±
0.8
1. 0.1±
0.46 0.05±
0.2
0.2
4.80 0.20±
10.1±
19.8
21 0.2±
38 0.4±
7.6
11.7
D0022-B/1
MIS-INSERT
COVERING PIN
H
8
1
CDC16xxF-E ADVANCE INFORMATION
14 March 31, 2003; 6251-606-2AI Micronas
2.3. Multiple Function Pins
2.3.1. U-Ports
Apart from their basic function (digital I/O), Universal Ports
(prefix “U”) have overlaid alternative functions (see Fig. 2–2
on page 12).
How to enable Basic Function, Special In and Special Out
mode is explained in the functional description of the U-
Ports. How to enable LCD mode is explained in the func-
tional descriptions of LCD module and U-Ports.
Bus Mode is used for testing purposes only. It is controlled by
the Control Register (CR) setting. Refer to section “Control
Word” for more information.
2.3.2. H-Ports
Apart from their basic function (digital I/O), High Current
Ports (prefix “H”) have overlaid alternative functions (see Fig.
2–2 on page 12).
How to enable Basic Function, Special In and Special Out
mode is explained in the functional description of the H-
Ports.
The Bus Mode is used for testing purposes only. It is con-
trolled by the Control Register (CR) setting. Refer to section
“Control Word” for more information.
2.3.3. Emulator Bus
In the CPGA177 package, the Emulator Bus (prefix “E”)
serves as connection to an external emulation hardware.
In the PQFP100 MCM Package it is internally connected to
the Flash program storage and is not available to the out-
side.
Its function is controlled by register CR. Refer to section
“Control Word” for more information.
2.4. Pin Function Description
ABORTQ
The Abort input serves as the CPU’s ABORTQ input pro-
vided that flag EXTIR in register SR2 is set. Active LOW.
ADB0 to ADB23
These 24 lines form the address bus for memory and I/O
exchange. The function is controlled by register CR.
AM-OUT
This is the output signal of the AM.
AM-PWM
This is the output signal of the 8-bit PWM of the AM. It is
intended for testing only.
AVDD
This is the positive power supply for ADC, P06COMP and
ERM. AVDD should be kept at VDD ±0.5 V.
AVSS
This is the negative reference for the ADC and the negative
power supply for ADC, P06COMP and ERM. Connect to sys-
tem ground.
BE
The Bus Enable output signal shows the state of the internal
CPU.
1: Internal CPU has bus access.
0: Internal CPU has no bus access.
BP0 to BP3
These pin functions serve as Backplane drivers for a 4:1
multiplexed LCD.
CAN0-RX, CAN1-RX, CAN2-RX
These signals provide the input lines for the CAN0, CAN1
and CAN2 modules.
CAN0-TX, CAN1-TX, CAN2-TX
These signals provide the output lines for the CAN0, CAN1
and CAN2 modules.
CC0-IN, CC1-IN, CC2-IN
These signals are the capture inputs of the CAPCOM0,
CAPCOM1 and CAPCOM2 modules.
CC0-OUT, CC1-OUT, CC2-OUT
These signals are the compare outputs of the CAPCOM0,
CAPCOM1 and CAPCOM2 modules.
CEQ
This output signal connects to external memory’s CEQ pin
and reduces its power consumption when CPU operates in
SLOW mode. Active LOW.
CO0, CO1
The signals Clock Out 0 and 1 provide frequency outputs.
Both are connected to internal prescaler and multiplexer.
They can be hard-wired by HW Option. Refer to section
“Hardware Options” for setting the CO0 bytes 00FFABh,
00FFB2h, 00FFB3h, 00FFB6h and the CO1 byte 00FFACh.
For testing purposes it is possible to drive clocks and other
signals of internal peripheral modules out of CO0 and CO1.
Selection is done via register TST2.
DB0 to DB7
The eight bidirectional Data Bus lines provide the 8-bit data
bus for use during data exchanges between the micropro-
cessor and external memory or peripherals. The function is
controlled by register CR.
DIGIT-IN
This is the receive input line of the DIGITbus module.
DIGIT-OUT
This is the transmit output line of the DIGITbus module.
EADB0 to EADB23
These 24 lines form the address bus for external memory
access on the Emulator Bus. The function is controlled by
register CR.
EBE
The Emulator Bus Enable output signal shows the state of
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 15
the internal CPU.
1: Internal CPU has bus access.
0: Internal CPU has bus no access.
ECEQ
Emulator Bus Chip Enable output signal connects to external
memory’s CEQ pin and reduces its power consumption
when CPU operates in SLOW mode. Active LOW.
EDB0 to EDB7
These eight bidirectional Emulator Data Bus lines provide
the 8-bit data bus for use during data exchanges between
the microprocessor and external memory or peripherals.
EOEQ
The Emulator Bus Output Enable signal connects to the
OEQ pin of external memory for read access. Active LOW.
EPH2
This output is the system clock of the Emulator Bus. It pro-
vides the timing for external memory access.
ESTOPCLK
If the Emulator Stop Clock input signal is HIGH, all the
peripheral modules are halted (fOSC is stopped), whereas the
clock PH2 and the CPU remain active. Thus it is possible to
read the registers and memory for debugging purposes.
For normal operation connect ESTOPCLK to System Ground
or leave it floating (internal pull-down).
EVDA
The Emulator Bus Valid Data Address output signal shows
the state of the internal CPU. It must be considered together
with the signal EVPA (Emulator Bus Valid Program Address).
Each signal indicates a valid memory address when high.
EVDD1, EVDD2
These two lines form the positive power supply of the Emula-
tor Bus drivers.
EVPA
Refer to EVDA.
EVSS1, EVSS2
These two lines form the negative supply of the Emulator
Bus drivers. Connect to system ground.
EWEQ
The output signal Emulator Bus Write Enable connects to the
external memory’s WEQ pin and activates it for write access.
Active LOW.
GD0 to GD7
These eight bidirectional Graphics IC Data lines provide an
8-bit DMA controlled data link to an external IC.
GRDQ
This Graphics IC Read line provides the control signal for
read accesses via the GD7 to GD0 bus. Active LOW.
GWRQ
This Graphics IC Write line provides the control signal for
write accesses via the GD7 to GD0 bus. Active LOW.
H0.0 to H3.5
The High Current Ports are intended for use as digital I/O
which can drive higher currents than the Universal Ports.
Each of the high current ports H0 to H3 is six bits wide.
HVDD1, HVDD2
The pins HVDD1 and HVDD2 are the positive power supply
of the high current ports H0.0 to H3.5. HVDD1 feeds ports
H0 and H1. HVDD2 feeds ports H2 and H3. HVDD1 and
HVDD2 should be kept at VDD ±0.5 V. Be careful to design
the PCB traces for carrying the considerable operating cur-
rent on these pins.
HVSS1, HVSS2
The pins HVSS1 and HVSS2 are the negative power supply
for the high-current ports H0.0 to H3.5. HVSS1 feeds ports
H0 and H1. HVSS2 feeds ports H2 and H3. Both have to be
hard-wired to system ground. Be careful to layout sufficient
PCB traces for carrying the considerable operating current
on these pins.
INT-TEST-IN
Test input signal for Interrupt Controller.
IRQ
IRQ serves as the CPU’s IRQ input, provided that flag EXTIR
in register SR2 is set. Active LOW.
LCD-CLK-IN
The Clock input of the LCD module receives the clock of an
optional external LCD master driver which is used to extend
the LCD driver capability. This input is active if the internal
LCD module is configured as slave and the external LCD
driver operates as master.
LCD-CLK-OUT
The Clock output of the LCD module provides a clock signal
to optional external LCD slave drivers if the internal LCD
module is configured as master and the other LCD drivers
are slaves.
LCD-SYNC-IN
The Synchronization input of the LCD module receives the
sync signal from an optional external LCD master driver. This
input is active if the internal LCD module is configured as
slave and the external LCD driver serves as master.
LCD-SYNC-OUT
The Synchronization output of the LCD module provides a
sync signal to optional external LCD slave drivers if the inter-
nal LCD module is configured as master and the other LCD
drivers are slaves.
MULTI-TEST-IN
This is a test input line. It is intended for factory test only. The
application should not use this signal.
OEQ
The Output Enable output signal connects to the OEQ pin of
external memory for read access. Active LOW.
PH2
The System Clock output signal provides timing for external
Table 2–1: Valid Memory Address
EVDA,
VDA EVPA,
VPA State
0 0 Internal Operation. Address and data
bus available. Address bus may be
invalid.
0 1 Valid program address. May be used
for program cache control.
1 0 Valid data address. May be used for
data cache control.
1 1 Opcode fetch. May be used for pro-
gram cache control and single step
control.
CDC16xxF-E ADVANCE INFORMATION
16 March 31, 2003; 6251-606-2AI Micronas
read or write operations. Addresses are valid (after the
Address Setup Time) following the negative transition of
PH2. The function is controlled by register CR.
PINT0-IN, PINT1-IN, PINT2-IN
The Port Interrupt 0, 1 and 2 inputs serve as inputs to the
PINT.
PINT3
The Port Interrupt 3 input serves as input to the PINT. HW
option FFC2h has to be set to determine whether U5.6 or
U5.7 is the source of PINT3-IN.
PINT0-OUT, PINT1-OUT
The Port Interrupt 0 and 1 outputs carry the output signals of
the PINT.
POL
Output of the Polling Module.
PWM0 to PWM4
These are the outputs of the PWM. Some of these PWM sig-
nals are directed to two pins.
P0.1 to P0.9
These 9 analog ports are the multiplexed input channels of
the ADC. Analog port P0.6 is additionally input to the
P06COMP.
The analog ports P0.1 to P0.5 can also be used as five digi-
tal input lines. The digital function of the analog ports should
be disabled (P0DIN in register SR1 set to 0) when operating
these ports as analog inputs to avoid leakage currents.
RDY
This output signal shows the state of the internal CPU.
1: CPU active
0: CPU halted by IR or DMA
RESETQ
This bidirectional signal is used to initialize all modules and
start program execution.
Two comparators distinguish three input levels:
– A low level resets all internal modules.
– A medium level activates all internal modules and starts
program execution. An alarm signal is generated which
can be directed to the interrupt controller.
– A high level keeps all internal modules active and cancels
the alarm signal.
The RESETQ input signal must be held low for at least two
clock cycles after VDD reaches operating voltage.
Internal reset sources output their reset request on the
RESETQ pin via an internal open drain pull-down transistor.
Thus RESETQ can be wire-ored with external reset sources.
The internally limited pull-down current allows direct connec-
tion to large capacitors. The connection of such a capacitor
(e.g. 10 nF) is recommended to reduce the capacitive influ-
ence of the neighboring XTAL2 pin.
RESETQ must be pulled up by an external pull-up resistor
(e.g. 10 kΩ).
RWQ
This input/output signal controls data exchange in coopera-
tion with DB and ADB. It can be driven from the external
CPU if the internal CPU is switched off.
1: CPU reads data.
0: CPU writes data.
SEG1.4 to SEG7.3
These pin functions serve as Segment drivers for a 4:1 multi-
plexed LCD.
SMA to SME
These lines are intended for driving stepper motors. They
are the outputs of the SM. Two of these lines together with
an external coil form an H-bridge. Thus each of the signals
SMA to SME can drive a two-phase bipolar stepper motor.
SMA-COMP to SME-COMP
These lines are comparator inputs that connect to one line
each of the SMA to SME lines. They serve to distinguish
rotation from stand-still during zero detection in each stepper
motor.
SPI0-CLK-IN, SPI1-CLK-IN
The Serial Synchronous Peripheral Interface Clock input
receives the bit clock from an external master, to shift data in
or out of SPI0 resp. SPI1 in slave mode. This means that the
external master controls the bit stream.
SPI0-CLK-OUT, SPI1-CLK-OUT
The Serial Synchronous Peripheral Interface Clock output
supplies the bit clock of SPI0 resp. SPI1 to an external slave,
to shift data in or out of SPI0 resp. SPI1 in master mode.
This means that the internal SPI controls the bit stream.
SPI0-D-IN, SPI1-D-IN
These are the data input lines of the SPI0 and SPI1 mod-
ules.
SPI0-D-OUT, SPI1-D-OUT
These are the data output lines of the SPI0 and SPI1 mod-
ules.
STOPCLK
If the input signal Stop Clock is HIGH, all the peripheral mod-
ules are halted (fOSC is stopped). But the clock PH2 and the
CPU remain active. Thus it is possible to read the registers
and memory for debugging purposes. TEST must be held
high to enable STOPCLK.
TEST
The main function of the Test pin is to define the source for
the Control Word fetch during reset. If TEST is held low dur-
ing reset, the Control Word is fetched via the Emulator Bus
(from internal Flash program storage in MCM package). If
TEST is held high during reset, the Control Word is fetched
via the Test Bus.
For normal operation with internal code connect TEST to
System Ground or leave it floating (internal pull-down).
TEST must be held high in active mode to enable STOPCLK.
T0-OUT
The Timer 0 output is connected to the zero output of T0 by a
divide by 2 scaler. The scaler generates a 50% pulse duty
factor.
T1-OUT, T2-OUT
These signals are connected to the zero outputs of T1 and
T2. T1-OUT is directed to several pins.
UART0-RX, UART1-RX, UART2-RX
These are the Receive input lines of UART0, UART1 and
UART2. Polarity of the signals is settable by HW options
FFB4h, respectively FFB5h.
UART0-TX, UART1-TX, UART2-TX
These are the data output lines of UART0, UART1 and
UART2. Polarity of signals is settable by HW options FFB4h,
respectively FFB5h.
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 17
UVDD
The pin UVDD is the positive supply voltage for the U-Port
output stages (see Fig. 2–4 for external connection).
UVSS
The pin UVSS is the negative power supply for the U-Port
output stages. It has to be connected to system ground (see
Fig. 2–4).
U1.0 to U7.3
Universal ports are intended for use as digital I/O or as LCD
driver outputs. The ports U1 to U6 consist of eight bit lines
each. Port U7 is 4 bits wide.
VDA
The signal Valid Data Address must be considered together
with the signal VPA (Valid Program Address). They show the
state of the internal CPU. If the internal CPU is switched off,
both signals may be driven from the outside. Together they
indicate a valid memory address when HIGH (see Table 2–1
on page 15).
VDD
The pin VDD is the positive supply voltage for the internal
digital modules (see Fig. 2–4 for external connection).
VPA
Refer to VDA.
VPQ
The Vector Pull output indicates that the CPU addresses a
vector location during an interrupt sequence. VPQ is low dur-
ing the last two interrupt sequence cycles, during which time
the CPU reads the interrupt vector.
VREF
This pin is the positive reference input for the ADC. The volt-
age at this pin must be set to a level between 2.56 Volts and
AVDD.
VSS
The pin VSS is the negative supply terminal of the internal
digital modules (see Fig. 2–4 for external connection).
WEQ
The Write Enable output signals a write access to external
memory. Active LOW.
WP0 to WP9
The Wake Port inputs are inputs to the Port Wake Module
inside the Power Saving Module. They serve as Wake Ports
during power saving modes and as port interrupt inputs dur-
ing CPU-active modes.
XTAL1
This is the quartz oscillator or clock input pin (see Fig. 2–4
for external connection).
XTAL2
This is the quartz oscillator output pin for two pin oscillator
circuits (see Fig. 2–4 for external connection).
2.5. External Components
Fig. 2–4: Recommended external supply and quartz connection for low electromagnetic interference (EMI)
To provide effective decoupling and to improve EMC behav-
ior, the small decoupling capacitors must be located as close
to the supply pins as possible. The self-inductance of these
capacitors and the parasitic inductance and capacitance of
the interconnecting traces determine the self-resonant fre-
quency of the decoupling network. A frequency too low will
C
18 p
18 p
UVDD
VDD
VSS
UVSS
XTAL1
XTAL2
IC
System
Ground
L
47 n
Resetq
+5 V
+5 V
4.7 k
AVSS
VREF
AVDD
HVSS 0 to 1
HVDD 0 to 1
EVSS 0 to 1
EVDD 0 to 1
RESETQ
C
C
2 * C
10 n
2 * C
+5 V
+5 V
+5 V Analog
System
Ground
System
Ground
Analog
Ground
C = 100 n to 150 n
C = 100 n to 150 n
CDC16xxF-E ADVANCE INFORMATION
18 March 31, 2003; 6251-606-2AI Micronas
reduce decoupling effectiveness, increase RF emissions and
may affect device operation adversely.
XTAL1 and XTAL2 quartz connections are especially sensi-
tive to capacitive coupling from other PC board signals. It is
strongly recommended to place quartz and oscillation capac-
itors as close to the pins as possible, and to shield the
XTAL1 and XTAL2 traces from other signals by embedding
them in a VSS trace.
The RESETQ pin adjacent to XTAL2 should be supplied with
a 47nF capacitor, to prevent fast RESETQ transients from
being coupled into XTAL2, to prevent XTAL2 from coupling
into RESETQ, and to guarantee a time constant of ≥200 µs
sufficient for proper Wake Reset functionality.
2.6. Pin Circuits
Fig. 2–5: Input Pins
Fig. 2–6: Input Pins with Pull-Down
Fig. 2–7: Push Pull I/O Pins
Fig. 2–8: Open Drain I/O
Fig. 2–9: Push Pull Output Pins
GNDOUT
VSUPOUT
VSUPIN
GNDIN
Input
Logic
Input
Logic
GNDOUT
VSUPOUT
VSUPIN
GNDIN
weak
GNDOUT
VSUPOUT
VSUPIN
GNDIN
Input
Logic
Input
Logic
GNDOUT
VSUPOUT
VSUPIN
GNDIN
GNDOUT
VSUPOUT
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 19
Table 2–2: I/O Supply Catalog
Pin Names Figure VSUPOUT GNDOUT VSUPIN GNDIN
XTAL1, XTAL2 2–5 UVDD UVSS VDD VSS
ESTOPCLK 2–6
TEST
U-Ports 2–7
H-Ports HVDD HVSS
P-Ports AVDD UVSS
EDB0 to EDB7 EVDD EVSS
RESETQ 2–8 UVDD UVSS
EADB0 to EADB19, ECEQ, EOEQ, EWEQ 2–9 EVDD EVSS
EADB20 to EADB23, EPH2, EVDA, EVPA, EBE UVDD UVSS
CDC16xxF-E ADVANCE INFORMATION
20 March 31, 2003; 6251-606-2AI Micronas
3. Electrical Data
3.1. Absolute Maximum Ratings
1) This condition represents the worst case load with regard to the intended application
Stresses beyond those listed in the “Absolute Maximum Ratings” may cause permanent damage to the device. This is a stress
rating only. Functional operation of the device at these or any other conditions beyond those indicated in the “Recommended
Operating Conditions/Characteristics” of this specification is not implied. Exposure to absolute maximum ratings conditions for
extended periods may affect device reliability.
Table 3–1: UVSS=HVSS1=HVSS2=EVSS1=EVSS2=AVSS=0V
Symbol Parameter Pin Name Min. Max. Unit
VSUP Core Supply Voltage
Port Supply Voltage
Analog Supply Voltage
SM Supply Voltage 1
SM Supply Voltage 2
Flash Port Supply Voltage 1
Flash Port Supply Voltage 2
VDD
UVDD
AVDD
HVDD1
HVDD2
EVDD1
EVDD2
−0.3 6.0 V
∆VDD Voltage Difference between VDD and
AVDD resp. UVDD
VDD, AVDD
UVDD −0.5 0.5 V
ISUP Core Supply Current
Port Supply Current
Flash Port Supply Current
VDD, VSS
UVDD, UVSS
EVDD1, EVSS1
EVDD2, EVSS2
−40 40 mA
IASUP Analog Supply Current AVDD, AVSS −20 20 mA
IHSUP SM Supply Current
@TCASE=105C, Duty Factor=0.71 1)HVDD1, HVSS1
HVDD2, HVSS2 −380 380 mA
Vin Input Voltage U-Ports,
XTAL,RESETQ,
TEST
UVSS−0.5 UVDD+0.7 V
P0-Ports
VREF UVSS−0.5 AVDD+0.7 V
H-Ports HVSS−0.5 HVDD+0.7 V
E-Ports EVSS−0.5 EVDD+0.7 V
Iin Input Current all Inputs 0 2 mA
IoOutput Current U-Ports
E-Ports −55 mA
H-Ports −60 60 mA
toshsl Duration of Short Circuit in Port SLOW
Mode to UVSS or UVDD U-Ports except
U3.2 in DP Mode indefinite s
TjJunction Temperature under Bias −45 115 °C
TsStorage Temperature −45 125 °C
Pmax Maximum Power Dissipation 0.8 W
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 21
3.2. Recommended Operating Conditions
Table 3–2: UVSS = HVSS1 = HVSS2 = EVSS1 = EVSS2 = AVSS = 0V
Symbol Parameter Pin Name Min. Typ 1) Max. Unit
VDD Supply Voltage
Port Supply Voltage
Analog Supply Voltage
Flash Port Supply Voltage 1
Flash Port Supply Voltage 2
VDD
UVDD
AVDD
EVDD1
EVDD2
4.5 5 5.5 V
HVDD SM Supply Voltage 1
SM Supply Voltage 2 HVDD1
HVDD2 4.75 5 5.25 V
∆VDD Voltage Difference between VDD
and AVDD resp. UVDD VDD, AVDD
UVDD -0.2 0.2 V
dAVDD AVDD Ripple, Peak-to-Peak AVDD 200 mV
fXTAL XTAL Clock Frequency XTAL1 4 12 MHz
XTAL Clock Frequency
using ERM XTAL1 4 10 MHz
TjJunction Temperature -40 110 C
Vil Low Input Voltage U-Ports
H-Ports
P0-Ports
TEST
0.51*VDD V
E-Ports 0.8 V
Vih High Input Voltage U-Ports
H-Ports
P0-Ports
TEST
0.86*VDD V
E-Ports 2.2 V
RVil Reset Active Input Voltage RESETQ 0.9 V
WRVil Reset Active Input Voltage during
Power Saving Modes and Wake
Reset
RESETQ 0.6 V
RVim Reset Inactive and Alarm Active
Input Voltage RESETQ 1.6 2.1 V
RVih Reset Inactive and Alarm Inactive
Input Voltage RESETQ 2.9 V
WRVih Reset Inactive during Power Sav-
ing Modes RESETQ UVDD -
0.4V V
VREFi ADC Reference Input Voltage VREF 2.56 AVDD V
P0ViP0 ADC Input Port Input Voltage P0-Ports 0 VREFi V
Clock Input from External Generator
XVil Clock Input Low Voltage XTAL1 0.2*VDD V
XVih Clock Input High Voltage XTAL1 0.8*VDD V
DXTAL Clock Input High-to-Low Ratio XTAL1 0.45 0.55
CDC16xxF-E ADVANCE INFORMATION
22 March 31, 2003; 6251-606-2AI Micronas
3.3. Characteristics
Table 3–3: UVSS = HVSS1= HVSS2 = EVSS1 = EVSS2 = AVSS = 0 V, 4.5 V< VDD = AVDD = UVDD = EVDD1= EVDD2 < 5.5 V,
4.75 V < HVDD1 = HVDD2 < 5.25 V, TCASE = −40 °C to +105 °C, fXTAL= 10 MHz
Symbol Parameter Pin Na. Min. Typ 1) Max. Unit Test Conditions
Package
Rthjc Thermal Resistance from
Junction to Case 15 C/W
Supply Currents CMOS levels on all
Inputs,
no Loads on Outputs
difference between any
two VDDs within ±0.2 V
IDDF VDD FAST Mode Supply
Current VDD 15 25 mA
IDDS VDD SLOW Mode Supply
Current VDD 1.7 mA all Modules OFF 2), 6)
IDDD VDD DEEP SLOW Mode
Supply Current VDD 1.4 all Modules OFF 2), 6)
IDDI VDD IDLE Mode Supply
Current VDD 50 75 µAf
xtal = 4 MHz 6)
60 90 µAf
xtal = 10 MHz 6)
70 100 µA internal RC oscill.
IDDW VDD WAKE Mode Supply
Current VDD 30 50 µA
UIDDa UVDD Active Supply
Current UVDD 0.3 mA no Output Activity,
LCD Module ON
AIDDa AVDD Active Supply
Current AVDD 0.2 0.4 mA ADC ON, ERM OFF
12mAERM ON,
fXTAL= 8.4 MHz
AIDDq Quiescent Supply Current AVDD 1 10 µA ADC and ERM OFF
UIDDq UVDD 1 10 µA no Output Activity,
LCD Module OFF
EIDDq EVDD1
EVDD2 110µA no Output Activity, LCD
Module OFF
HIDDq Sum of
all
HVDD1
HVDD2
120µA no Output Activity,
SM Module OFF
Inputs
1) Typical values describe typical behavior at room temperature (25 °C, unless otherwise noted), with typical
Recommended Operating Conditions applied (derived from device characterization, not 100% tested).
2) Value may be exceeded with unusual Hardware Option setting
3) Design value only, the actually observable hysteresis may be lower due to system activity and related supply noise
4) When the ERM is active, this time value is increased by 0.121/fXTAL, e.g. 15.125 ns @8 MHz.
5) When the ERM is active, this time value is decreased by 0.121/fXTAL, e.g. 15.125 ns @8 MHz.
6) Measured with external clock. Add 170 µA @ 4 MHz, 200 µA @ 10 MHz for operation on typical quartz with
SR3.XTAL = 0 (Oscillator RUN mode).
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 23
Vilh Low to High Input
Threshold Voltage all Inp.
except
EPorts,
XTAL
0.68*
VDD
0.76*
VDD
0.84*
VDD
V
Vihl High to Low Input
Threshold Voltage all Inp.
except
EPorts,
XTAL
0.53*
VDD
0.61*
VDD
0.69*
VDD
V
Vilh-Vihl Input Hysteresis all Inp.
except
EPorts,
XTAL
0.1*
VDD
0.15*
VDD
0.2*
VDD
V3)
IiInput Leakage Current U-Ports
H-Ports
P0-Port
P06
VREF
E-Ports
−1
−10
−1
−0.2
−1
−1
1
10
1
0.2
1
1
µA 0 < Vi < UVDD
0 < Vi < HVDD
0 < Vi < AVDD
0 < Vi < AVDD
0 < Vi < AVDD
0 < Vi < EVDD
Ipd Input Pull-Down Current TEST
E-Ports 25
25 80
80 170
170 µAVi = UV
DD
Vi = EVDD, when unused
Ipu Input Pull-Up Current E-DB −170 −80 −25 µA Vi = 0, when unused
Outputs
Vol Port Low Output Voltage U-Ports 0.4 V Io = 2 mA
E-Ports 0.4 V Io = 0.5 mA
H-Ports 0.125 0.5 V Io = 27 mA
Io = 40 mA@TCASE =
-40 °C
Io = 30mA@TCASE =
25 °C
∆Vol Spread of Vol Values within
one SM Driver Module H-Ports −50 50 mV
Voh Port High Output Voltage U-Ports UVDD−
0.4 VIo= −2mA
E-Ports EVDD−
0.4 VIo= −0.5 mA
H-Ports HVDD−
0.5 HVDD−
0.125 VIo = −27 mA
Io = −40 mA@TCASE =
−40 °C
Io = −30 mA
@TCASE = 25 °C
∆Voh Spread of Voh Values within
one SM Driver Module H-Ports −50 50 mV
Table 3–3: UVSS = HVSS1= HVSS2 = EVSS1 = EVSS2 = AVSS = 0 V, 4.5 V< VDD = AVDD = UVDD = EVDD1= EVDD2 < 5.5 V,
4.75 V < HVDD1 = HVDD2 < 5.25 V, TCASE = −40 °C to +105 °C, fXTAL= 10 MHz
Symbol Parameter Pin Na. Min. Typ 1) Max. Unit Test Conditions
1) Typical values describe typical behavior at room temperature (25 °C, unless otherwise noted), with typical
Recommended Operating Conditions applied (derived from device characterization, not 100% tested).
2) Value may be exceeded with unusual Hardware Option setting
3) Design value only, the actually observable hysteresis may be lower due to system activity and related supply noise
4) When the ERM is active, this time value is increased by 0.121/fXTAL, e.g. 15.125 ns @8 MHz.
5) When the ERM is active, this time value is decreased by 0.121/fXTAL, e.g. 15.125 ns @8 MHz.
6) Measured with external clock. Add 170 µA @ 4 MHz, 200 µA @ 10 MHz for operation on typical quartz with
SR3.XTAL = 0 (Oscillator RUN mode).
CDC16xxF-E ADVANCE INFORMATION
24 March 31, 2003; 6251-606-2AI Micronas
LVol LCD Port Zero Output Volt-
age U-Ports −0.05 0.05 V no load
LVo1 LCD Port Low Output Volt-
age U-Ports 1/3
*UVDD
−0.05
1/3
*UVDD
+0.05
Vno load
LVo2 LCD Port High Output Volt-
age U-Ports 2/3
*UVDD
-0.05
2/3
*UVDD
+0.05
Vno load
LVoh LCD Port Full Output Volt-
age U-Ports UVDD
−0.05 UVDD
+0.05 Vno load
ΣLIo1 Internal LCD-Low Supply
Short Circuit Current U-Ports 0.3
−0.3
mA Pin Short to 2/3*UVDD
Pin Short to UVSS
ΣLIo2 Internal LCD-High Supply
Short Circuit Current U-Ports 0.3
−0.3
mA Pin short to UVDD
Pin short to 1/3*UVDD
Ishf Port Fast Short Circuit Cur-
rent U-Ports TBD 9 TBD mA Pin Short to UVDD or
UVSS, Port FAST Mode
Ishs Port Slow Short Circuit Cur-
rent U-Ports TBD 2.5 TBD mA Pin Short to UVDD or
UVSS, Port SLOW
Mode
Ishsd Port Slow Short Circuit Cur-
rent, DP Mode U3.2 TBD 5 TBD mA Pin Short to UVDD, Port
SLOW and Double Pull-
Down Modes
Comparators
VBG Internal Reference Voltage 1.125 1.25 1.375 V
tBG Internal Voltage Reference
Setup Time after Power-Up 10 µs
VREFR RESET Comparator Refer-
ence Voltage RESETQ 1*VBG 1*VBG V
RVlh−
RVhl
RESET Comparator Hyster-
esis, symmetrical to VREFR RESETQ 0.25 0.375 V 3)
WRVihl Reset Active high to low
Voltage during Power Sav-
ing Modes and Wake Reset
RESETQ 0.6 1.1 V
VREFA ALARM Comparator Refer-
ence Voltage RESETQ 2*VBG 2*VBG V
AVlh−
AVhl
ALARM Comparator Hyster-
esis, symmetrical to VREFA RESETQ 0.1 0.15 V 3)
Table 3–3: UVSS = HVSS1= HVSS2 = EVSS1 = EVSS2 = AVSS = 0 V, 4.5 V< VDD = AVDD = UVDD = EVDD1= EVDD2 < 5.5 V,
4.75 V < HVDD1 = HVDD2 < 5.25 V, TCASE = −40 °C to +105 °C, fXTAL= 10 MHz
Symbol Parameter Pin Na. Min. Typ 1) Max. Unit Test Conditions
1) Typical values describe typical behavior at room temperature (25 °C, unless otherwise noted), with typical
Recommended Operating Conditions applied (derived from device characterization, not 100% tested).
2) Value may be exceeded with unusual Hardware Option setting
3) Design value only, the actually observable hysteresis may be lower due to system activity and related supply noise
4) When the ERM is active, this time value is increased by 0.121/fXTAL, e.g. 15.125 ns @8 MHz.
5) When the ERM is active, this time value is decreased by 0.121/fXTAL, e.g. 15.125 ns @8 MHz.
6) Measured with external clock. Add 170 µA @ 4 MHz, 200 µA @ 10 MHz for operation on typical quartz with
SR3.XTAL = 0 (Oscillator RUN mode).
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 25
VREFPOR VDD Power On Reset
Threshold VDD 2.88*
VBG 2.88*
VBG V
VREFP06 P06 Comparator Reference
Voltage P06 0.49*
AVDD
0.51*
AVDD
V
P06Vlh−
P06Vhl
P06 Comparator Hystere-
sis, symmetrical to
VREFP06
P06 0.1 0.24 V 3)
VREFSM SM Comparator Reference
Voltage H00,
H04,
H12,
H20,
H32
1/9*
HVDD
−0.07
1/9*
HVDD
+0.07
V
tCDEL RESET, ALARM, P06, SM
Comparator Delay Time RESETQ
P06
H00
H04
H12
H20
H32
100 ns Overdrive = 50 mV
ADC
LSB LSB Value VREF
/1024 V
INL Integral Non-Linearity:
difference between the out-
put of an actual ADC and
the line best fitting the out-
put function (best-fit line)
−3
−2.5
3
2.5
LSB VREF = 2.56 V
VREF = 5.12 V
ZE Zero Error:
difference between the out-
put of an ideal and an actual
ADC for zero input voltage
−11LSBV
REF = 2.56 V
and 5.12 V
FSE Full-Scale Error:
difference between the out-
put of an ideal and an actual
ADC for full-scale input volt-
age
−11LSBV
REF = 2.56 V
and 5.12 V
TUE Total Unadjusted Error:
maximum sum of integral
non-linearity, zero error and
full-scale error
−4
−3
4
3
LSB VREF = 2.56 V
VREF = 5.12 V
QE Quantization Error:
uncertainty because of ADC
resolution
−0.5 0.5 LSB VREF = 2.56 V
and 5.12 V
Table 3–3: UVSS = HVSS1= HVSS2 = EVSS1 = EVSS2 = AVSS = 0 V, 4.5 V< VDD = AVDD = UVDD = EVDD1= EVDD2 < 5.5 V,
4.75 V < HVDD1 = HVDD2 < 5.25 V, TCASE = −40 °C to +105 °C, fXTAL= 10 MHz
Symbol Parameter Pin Na. Min. Typ 1) Max. Unit Test Conditions
1) Typical values describe typical behavior at room temperature (25 °C, unless otherwise noted), with typical
Recommended Operating Conditions applied (derived from device characterization, not 100% tested).
2) Value may be exceeded with unusual Hardware Option setting
3) Design value only, the actually observable hysteresis may be lower due to system activity and related supply noise
4) When the ERM is active, this time value is increased by 0.121/fXTAL, e.g. 15.125 ns @8 MHz.
5) When the ERM is active, this time value is decreased by 0.121/fXTAL, e.g. 15.125 ns @8 MHz.
6) Measured with external clock. Add 170 µA @ 4 MHz, 200 µA @ 10 MHz for operation on typical quartz with
SR3.XTAL = 0 (Oscillator RUN mode).
CDC16xxF-E ADVANCE INFORMATION
26 March 31, 2003; 6251-606-2AI Micronas
AE Absolute Error:
difference between the
actual input voltage and the
full-scale weighted equiva-
lent of the binary output
code, all error sources
included
−4.5
−3.5
4.5
3.5
LSB VREF = 2.56 V
VREF = 5.12 V
R Conversion Range P0-Ports AVSS VREF V 2.56 V < VREF < AVDD
A Conversion Result INT
(Vin/
LSB)
hex AVSS < Vin < VREF
000 hex Vin < = AVSS
3FF hex Vin > = VREF
tcConversion Time 4 µs
tsSample Time 2 µs
Ci Input Capacitance during
Sampling Period 15 pF
Ri Serial Input Resistance
during Sampling Period 5kOhm
SPI (Fig. 3–1, Fig. 3–2)
tsoci Data out Setup Time with
internal clock U3.0,
U6.4 60 4) ns @Cl = 30 pF, port fast
mode
thoci Data out Hold Time with
internal clock U3.0,
U6.4 −60 60 4) ns @Cl = 30 pF, port fast
mode
tsoce Data out Setup Time with
external clock U3.0,
U6.4 3/ fXTAL
+60 4) ns @Cl = 30 pF, port fast
mode
thoce Data out Hold Time with
external clock U3.0,
U6.4 2/ fXTAL
−60 ns @Cl = 30 pF, port fast
mode
tsi Data in Setup Time with
external clock U3.1,
U6.5 1/
fXTAL+
60 4)
ns
thi Data in Hold Time with
external clock U3.1,
U6.5 1/
fXTAL+
60 4)
ns
Table 3–3: UVSS = HVSS1= HVSS2 = EVSS1 = EVSS2 = AVSS = 0 V, 4.5 V< VDD = AVDD = UVDD = EVDD1= EVDD2 < 5.5 V,
4.75 V < HVDD1 = HVDD2 < 5.25 V, TCASE = −40 °C to +105 °C, fXTAL= 10 MHz
Symbol Parameter Pin Na. Min. Typ 1) Max. Unit Test Conditions
1) Typical values describe typical behavior at room temperature (25 °C, unless otherwise noted), with typical
Recommended Operating Conditions applied (derived from device characterization, not 100% tested).
2) Value may be exceeded with unusual Hardware Option setting
3) Design value only, the actually observable hysteresis may be lower due to system activity and related supply noise
4) When the ERM is active, this time value is increased by 0.121/fXTAL, e.g. 15.125 ns @8 MHz.
5) When the ERM is active, this time value is decreased by 0.121/fXTAL, e.g. 15.125 ns @8 MHz.
6) Measured with external clock. Add 170 µA @ 4 MHz, 200 µA @ 10 MHz for operation on typical quartz with
SR3.XTAL = 0 (Oscillator RUN mode).
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 27
Fig. 3–1: SPI: Send and Receive Data with Internal Clock. Timing is valid for inverted clock too (Data valid at positive edge).
Fig. 3–2: SPI: Send and Receive Data with External Clock. Timing is valid for inverted clock too (Data valid at positive edge).
CAN (Fig. 3–3)
tsrx rx-strobe Time CAN rx 0 10 4) ns reference is XTAL1 ris-
ing edge
tdtx tx-drive Time CAN tx 15 60 4) ns reference is XTAL1 fall-
ing edge
@Cl = 30 pF, port fast
mode
DIGITbus (Fig. 3–4)
tbtj Bit Time jitter DIGIT-
OUT ±10 4) ns rising edges, internal
clock master
tfed Falling edge delay DIGIT-
OUT 15 5) tBIT/64
+60 4) ns reference is nominal fall-
ing edge
Table 3–3: UVSS = HVSS1= HVSS2 = EVSS1 = EVSS2 = AVSS = 0 V, 4.5 V< VDD = AVDD = UVDD = EVDD1= EVDD2 < 5.5 V,
4.75 V < HVDD1 = HVDD2 < 5.25 V, TCASE = −40 °C to +105 °C, fXTAL= 10 MHz
Symbol Parameter Pin Na. Min. Typ 1) Max. Unit Test Conditions
1) Typical values describe typical behavior at room temperature (25 °C, unless otherwise noted), with typical
Recommended Operating Conditions applied (derived from device characterization, not 100% tested).
2) Value may be exceeded with unusual Hardware Option setting
3) Design value only, the actually observable hysteresis may be lower due to system activity and related supply noise
4) When the ERM is active, this time value is increased by 0.121/fXTAL, e.g. 15.125 ns @8 MHz.
5) When the ERM is active, this time value is decreased by 0.121/fXTAL, e.g. 15.125 ns @8 MHz.
6) Measured with external clock. Add 170 µA @ 4 MHz, 200 µA @ 10 MHz for operation on typical quartz with
SR3.XTAL = 0 (Oscillator RUN mode).
tsoci
thoci
SPI-CLK-OUT
SPI-D-OUT
SPI-D-IN
tsi thi
SPI-CLK-IN
SPI-D-IN
tsoce
thoce
SPI-CLK-IN
SPI-D-OUT
CDC16xxF-E ADVANCE INFORMATION
28 March 31, 2003; 6251-606-2AI Micronas
Fig. 3–3: CAN I/O Timing.
Fig. 3–4: DIGIT bus I/O Timing
3.4. Recommended Crystal Characteristics
Table 3–4: UVSS = HVSS1 = HVSS2 = EVSS1 = EVSS2 = AVSS = 0 V, 4.5 V < VDD = AVDD = UVDD = EVDD1 = EVDD2 < 5.5 V,
4.75 V< HVDD1 = HVDD2 < 5.25 V, TCASE = −40 °C to +105 °C
Symbol Parameter Min. Typ. Max. Unit Test Conditions
fPParallel Resonance Frequency
@ CL = 12 pF 412MHz
R1Series Resonance Res. for
50 ms Oscillation Start-Up time
@CL = 12 pF
@ fP = 4 MHz 380
320 Ohm START-UP
RUN
@ fP = 6 MHz 230
160 Ohm START-UP
RUN
@ fP = 8 MHz 150
95 Ohm START-UP
RUN
@ fP = 10 MHz 100
60 Ohm START-UP
RUN
CEXT External Oscillation Capaci-
tances for CL = 12 pF,
connected to VSS
18 pF
XTAL1
TX
RX
tsrx
tcyc
tdtx
nominal pulse
DIGIT-IN
DIGIT-OUT
tBIT
tNPL
tfed
tNPL: Nominal programmed Pulse Length. Depends on programmed phase, Baudrate and transmitted sign (0, 1, T).
Should be 1/4 for sign 0, 1/2 for sign 1 and 3/4 for sign T of tBIT.
tbtj tbtj
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 29
3.5. Flash/EMU Port Characteristics
Table 3–5: EVDD = VDD = 4.5 V to 5.5 V, TCASE = −40 °C to 85 °C
Symbol Parameter Min. Typ. Max. Unit Test Conditions
tPHD internal PH2 delay 6 8 ns
tADS Address Setup Time 15
+ 0.5
20
30
19
+ 0.7
26
40
ns
ns/pf
ns
ns
CEADB = 0 pF
CEADB = 10 pF
CEADB = 30 pF
tADH Address Hold Time 8 10 ns CEADB = 10 pF
tDSR Data Setup Read Time 6 12 ns
tDSW Data Setup Write Time 9
+0.5
14
24
14
+ 0.7
21
35
ns
ns/pf
ns
ns
CEDB = 0 pF
CEDB = 10 pF
CEDB = 30 pF
tDHR Data Hold Time Read 6 8 ns CEDB = 0 pF
tDHW Data Hold Time Write 8 10 ns CEDB = 0 pF
tWOS Write/Output Enable Setup 6 10 ns CEOQ,EWQ = 0 pF
tWOH Write/Output Enable Hold 6 9 ns CEOQ,EWQ = 0 pF
tBES Bus Enable Setup 14 24 ns CEBE = 0 pF
tBEH Bus Enable Hold 7 10 ns CEBE = 0 pF
tVS EVDA, EVPA Setup Time 21 30 ns CEVDA,EVPA = 0 pF
CDC16xxF-E ADVANCE INFORMATION
30 March 31, 2003; 6251-606-2AI Micronas
Fig. 3–5: Emu Bus Timing
Fig. 3–6: Memory Access Signals Emulator ports: CE (=ECEQ) may be used for low power
mode. Input data are latched with the rising edge of CE.
Test ports: Be careful with CE working in low power mode.
XTAL1
ESTOPCLK
Ph2
EPh2
READ DATA
WRITE DATA WRITE DATA
READ DATA
EWE, EOE
EADB,
EVDA, EVPA
tPHD
tcyc
tADH
tADS, tVS
tACC
tDSR
tDHR
tDHW tDSW
tSCLS tSCLH
tWOH
tWOS
tEPHD
EBE, ECE
tBEH, tBES
fOSC
Ph2
RW
CE
WE
OE
FAST mode SLOW mode
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 31
There are no input data latches at the test ports. Always pull
CE down.
Fig. 3–7: Stop Clock (only if TEST Pin = 1)
STOPCLK is used for emulation and test mode. It is active
with TEST input pulled to VDD only. With the stopped fOSC all
peripheral modules, such as timers and UARTs, are frozen.
But with the runni ng Ph2 clock the CPU is able to read and
write the internal registers and memory. The pin ESTOPCLK
has an internal pulldown. The pin STOPCLK (H1.5, Test
mode) has no internal pulldown, so in test mode an external
pulldown is needed.
Fig. 3–8: Internal/External Memory Access
XTAL1
STOPCLK
fOSC
Ph2
XTAL1
EPh2
RW
EADB
EDB
EOE
EWE
EDB
EOE
EWE
external external internal internal
internal signal
Control word
Bit 5 = 1
(ext. mem. mode)
Control word
Bit 5 = 0
(emu mode)
CDC16xxF-E ADVANCE INFORMATION
32 March 31, 2003; 6251-606-2AI Micronas
4. CPU and Clock System
The core basically consists of the CPU, RAM and ROM.
A Memory Banking module is included to allow access to
more than 64 k memory with an address bus limited to
16 bits.
ROM is subdivided in Boot ROM and Flash EEPROM.
In normal operation, after RESET, the CPU starts executing
boot loader SW code from the Boot ROM and then jumps
into the Flash EEPROM. Please find detailed information in
section Boot System.
In mask-ROM-derivative ICs, the code execution starts in
factory-defined mask ROM.
4.1. W65C816
The CPU is fully compatible to WDC’s W65C816 micropro-
cessor. This is a processor with 16-bit registers/accumulator,
an 8-bit data bus and a 24-bit address bus. A software switch
determines whether the processor is in the 8-bit emulation
mode, supporting SW compatibility to its 8-bit predecessors,
or in the 16-bit native mode. For further information on the
CPU core, please refer to the WDC W65C816 Data Sheet.
Table 4–1: Major Differences between Processors
and Modes
Item 65C816
Emulation 65C816
Native 65C02
ABS, X ASL,
LSR, ROL,
ROR with no
page cross-
ing
7 cycles 7 cycles 6 cycles
jump indi-
rect,
operand =
XXFF
5 cycles 5 cycles 6 cycles
branch
across Page 4 cycles 3 cycles 4 cycles
decimal
mode no addi-
tional cycle no addi-
tional cycle add 1 cycle
RWQ signal
during read-
modify-write
instructions
RWQ=0
during mod-
ify and write
cycles
RWQ=0
only during
write cycle
RWQ=0
only during
write cycle
abort signal yes yes no
accumulator 8/8 bits 16 or
8/8 bits 8 bits
addressing
modes 25 25 16
address
space 16 M 16 M 64 K
bank
registers yes yes none
block moves of little use yes none
break flag yes no yes
break vector FFFE.FFFF FFE6.FFE7 FFFE.FFFF
direct page
indexed wraps crosses
page wraps
flags after
interrupt D=0 D=0 D=0
flags after
reset D=x D=0 D=0
index
registers 8 bits 8 or 16 bits 8 bits
instructions 256 256 178
interrupts FFF4.FFFF FFE4.FFEF FFFA.FFFF
mnemonics 92 92 64
special page direct page direct page zero page
stack page 1 bank 0 page 1
unused
opcodes none none NOP
Table 4–1: Major Differences between Processors
and Modes
Item 65C816
Emulation 65C816
Native 65C02
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 33
4.1.1. Processor Modes
The 65C816 CPU allows operation in two modes:
– Native mode: 16-bit mode
– Emulation mode: 8-bit mode for emulation of 65C02 prop-
erties
Table 4–1 gives some details on the differences between
modes. Refer to the WDC W65C816 Data Sheet for more
information and on how to switch between modes.
After reset, Emulation mode is active.
However, Native mode is recommended for normal use. To
make maximum usage of the CPU’s 16-bit properties, switch
to Native mode after reset.
4.1.2. Emulation of 65C02
When using the Emulation mode to design software for a
derivative containing only the 65C02 8-bit CPU, care has to
be taken to use only the features available with that CPU.
Table 4–1 gives a comparing overview. Refer to the WDC
W65C816 Data Sheet for more information.
4.2. Operating Modes
To adapt to the large variety of CPU speed and current con-
sumption requirements, the device offers a number of oper-
ating modes:
– CPU-active modes, where the CPU is clocked at select-
able speeds.
– Power-saving modes, where the CPU is kept reset and
only certain portions of the circuit are powered.
Fig. 4–1 shows how the various modes are accessed in an
operating modes state diagram.
Fig. 4–1: Operating Modes State Diagram
4.2.1. CPU-Active Modes
The CPU can be operated in three different CPU-active
modes (Table 4–2). Core modules that are also affected by
CPU-active modes are:
1. Interrupt Controller with all internal and external interrupts
2. RAM, ROM/Flash and DMA
3. Watchdog
Table 4–2 shows the operability of the peripheral modules in
the various CPU-active modes.
WAKE/IDLE
DEEP SLOW
FAST
All reset sources
WAKE/IDLE RTC, Wake Ports
< 50 ms (if 4 ... 12MHz XTAL was off)
Global
Reset
RESETQ
< 50 ms (if 4 ... 12MHz XTAL was off)
~ 0,5 ms @ 8 MHz XTAL (if XTAL stays on)
w/o Ports
Global
Reset
with Ports
Power up
SLOW All reset sources
SLOW
FAST
~ 0,5 ms @ 8 MHz XTAL (if XTAL stays on)
SLOW DEEP SLOW
CDC16xxF-E ADVANCE INFORMATION
34 March 31, 2003; 6251-606-2AI Micronas
4.2.1.1. FAST Mode
After reset the CPU is in FAST mode. The CPU clock and
the I/O clock both equal the oscillator frequency fXTAL.
4.2.1.2. SLOW Mode
To considerably reduce power consumption, the user can
reduce the internal CPU clock frequency to 1/256 of the nor-
mal fXTAL value. In this CPU SLOW mode, program execu-
tion is reduced to 1/256 of the normal speed, but clocking of
most other modules remains unaffected. The modules that
are affected by CPU SLOW mode are:
1. CPU and Interrupt Controller with all internal and external
interrupts
2. RAM, ROM and DMA
3. Watchdog
Some modules must not be operated during CPU SLOW
mode (e.g. CAN). Refer to module sections for details (see
Table 4–2 on page 35).
CPU SLOW mode is enabled by clearing flag CPUFST in
standby register SR1. The CPU clock frequency reduction to
fXTAL/256 will take effect after a maximum delay of 256 fXTAL
periods.
Returning CPU to FAST mode is done by setting flag
CPUFST to HIGH. The CPU clock frequency will immedi-
ately change to its normal fXTAL value.
4.2.1.3. DEEP SLOW Mode
To further reduce power consumption beyond SLOW mode,
DEEP SLOW mode also disables most of the internal periph-
eral clocking system. Table 4–2 shows which peripheral
modules can be operated in DEEP SLOW mode.
Only peripheral module clocks f5 and slower are available
from the divider chain. T0 can be operated only with this limi-
tation.
To prevent undefined behavior don’t switch between DEEP
SLOW and FAST mode directly but select SLOW mode in
between.
Changing to or from DEEP SLOW mode is done by writing
SR3.FCLO (see Section 6.3. "Standby Registers" on
page 55 for details).
4.2.2. Power-Saving Modes
Power-saving modes are activated by the CPU. The com-
plete core logic will immediately terminate operation and
power will be reduced. The result is a device current con-
sumption that is greatly reduced, to the amount of leakage
currents. The internal SRAM and CAN-RAM keep their pro-
grammed data and all U, P, and H-Port registers keep their
programmed state.
However, a means to leave these modes has to be provided.
As the CPU is no longer active, either an external or internal
wake signal has to be generated. The external wake neces-
sitates no device current, but to generate an internal wake
requires an internal oscillator and a Real Time Clock (RTC)
to run, which will cost a small amount of supply current.
Please note that inadvertently entering a power-saving
mode, e.g. by an external electrical overstress (EOS) condi-
tion, when no wake source has been configured previously
as recovery path from this state, renders the device locked in
this power saving mode. Only a RESETQ pin reset or a com-
plete power removal and reapplication recovers the device
from this state. Sufficient external shielding measures must
avoid this hazard.
4.2.2.1. WAKE Mode
The WAKE mode is the most current-saving operation mode.
All device circuits are stopped or powered down except the
Port Wake Module (Table 4–3).
The Port Wake Module allows the CPU to configure up to ten
fixed device ports (see the device pinout for details) as Wake
Ports (WP).
To prepare for WAKE mode, the CPU has to switch off the
RTC and to configure the desired Wake Port(s) (see chapter
“Power Saving Module”, sections “Port Wake Module” and
“RTC Module”).
To enter WAKE mode, the CPU sets SR3.WAID.
The device will immediately enter WAKE mode by resetting
all circuitry, stopping all clocks, and powering down all ana-
log circuitry. As long as all Port inputs - except analog inputs
via P-ports P0.1 to P0.9 - are kept at CMOS input levels
(Vil = xVSS ± 0.3 V and Vih = xVDD ± 0.3 V), the supply cur-
rents will be minimal. The device may be kept in this state
indefinitely.
To exit WAKE mode, the previously configured Wake Port
has to switch. Immediately a Wake Reset sequence will be
started internally, which pulls the RESETQ pin low and
releases it as soon as all internal reset sources have become
inactive. See chapter “Core Logic” for details on internal
reset sources. After reset, the CPU starts in FAST mode as
usual.
4.2.2.2. IDLE Mode
IDLE mode allows to configure an internal wake source that
wakes after a preselected period. As clock sources, either a
current-saving, but imprecise internal RC oscillator, or the
precise, but more current-consuming XTAL oscillator, are
selectable. These circuits and the RTC are kept alive (Table
4–3) as well as the Port Wake Module.
The RTC allows the CPU to select from one-second to one-
day clocks as wake signal (see Section 8.4. "Operation of
RTC Module" on page 69 for details). Apart from serving as
wake source, the CPU may use the RTC as a real-time clock
that is not halted by resets.
A Polling Module, driven by a selectable RTC clock, may be
configured to generate a polling pulse on H2.5 and sample
the Wake Ports immediately after. Thus a periodical polling
of Wake Ports may be achieved, with no continuous power
consumption in external circuitry.
To prepare for IDLE mode, the CPU has to configure the
desired RTC wake clock (see Section 8.4. "Operation of RTC
Module" on page 69 for details), beside the desired Wake
Port(s) (see Section 8.6. "Operation of Port Wake Module"
on page 71 for details).
To enter IDLE mode, the CPU sets SR3.WAID.
The device will immediately enter IDLE mode by resetting all
circuitry, stopping the unused clocks, and powering down all
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 35
analog circuitry. As long as all Port inputs - analog inputs via
P-ports P0.1 to P0.9 excepted - are kept at CMOS input lev-
els (Vil = xVSS ±0.3 V and Vih = xVDD ±0.3 V), the supply cur-
rents will only amount to leakage and the requirement of the
enabled oscillator(s) and the slow-clocked RTC and Polling
Modules.
To exit IDLE mode, the previously configured Wake source
has to switch. Immediately a Wake Reset sequence will be
started internally, that pulls the RESETQ pin low and
releases it as soon as all internal reset sources have become
inactive. See chapter “Core Logic” for details on internal
reset sources. After reset, the CPU starts in FAST mode, as
usual.
Table 4–2: CPU-Active Modes and their effect on peripheral modules
Module FAST SLOW DEEP SLOW
Core
Digital Watchdog ✔✔✔
IR Interrupt Controller Unit ✔✔✔
Port Interrupts ✔✔✔
Port Wake Module ✔✔✔
Memory Patch Module ✔✔✔
Analog
A/D Converter ✔✔
ALARM, P06 and Comparators ✔✔✔ 3)
LCD Module ✔✔✔ 2)
Communication
DMA ✔✔✔
UART ✔✔
SPI ✔✔
CAN ✔✔ 3)
DIGITbus ✔✔ 3)✔ 2) 3)
Input & Output
Ports ✔✔✔
Stepper Motor Module ✔
PWM ✔✔
Audio Module ✔
Clock Outputs ✔✔✔ 2)
Timers & Counters
Capture Compare Module ✔✔ 1)✔ 2) 3)
Timers ✔✔✔ 2)
RTC/Polling Module ✔✔✔
1) Avoid write access to CCxI
2) Only clocks f5 and slower are available from Clock Divider
3) Don’t access registers or CAN RAM
CDC16xxF-E ADVANCE INFORMATION
36 March 31, 2003; 6251-606-2AI Micronas
4.3. Clock System
This IC contains a quartz oscillator circuit that only requires
external connection of a quartz and of 2 oscillation capaci-
tors.
The XTAL Oscillator generates a 4 to 12 MHz reference sig-
nal from an external quartz resonator, cf. section “Electrical
characteristics” for quartz data.
A reset sets the module to START-UP mode, where, at the
expense of a higher current consumption, marginal quartzes
receive more drive to ease start-up of oscillation.
After start-up of the CPU program, register SR3.XTAL may
be cleared by SW to set the XTAL Oscillator to RUN mode,
where current consumption is at its standard level.
Switching between START-UP and RUN modes must not be
done @ FQUARZ >= 10 MHz or with the ERM active, as this
might lead to unpredictable behavior of the clock system.
This module is permanently active except during power sav-
ing mode, where continued operation may be selected in
register OSC.XM.
Fig. 4–2: Clock System
The oscillator clock drives a system clock divider that sup-
plies the various modules with its specific clock. Module
clock selection is software-defined in some cases, hardware
or HW option-defined in other cases. The module descrip-
tions give details.
Table 4–3: Power Saving Modes and related functionality vs. CPU-Active Modes
Operating Mode Modules which can be activated: SRAM,
CAN-RAM Port Regis-
ters Available
Wake
Sources
Power
Saving
Modes
WAKE - Port Wake Module data
retained state
retained Wake Ports
IDLE all WAKE Mode modules plus:
- 4 ... 12 MHz XTAL and/or 20..50 k RC Oscillator
- Real-Time Clock
- Polling Module
Wake Ports
and RTC
CPU-Active Modes in principle all, for limitations see Table 4–2 active active Wake
sources
usable as
interrupt
Table 4–4: Operating Mode Selection and Effect on Clocks
SR1.CPUFST
SR3.FCLO
SR3.WAID
Operat-
ing Mode PH2 f0
000SLOW f
XTAL/256 fXTAL
100FAST f
XTAL fXTAL
010DEEP
SLOW fXTAL/256 f0 to
f4 = VDD
101WAKE/
IDLE VDD VDD
fXTAL = 4 ... 12MHz
1/256
’1’
XTAL Oscillator
Peripheral
Modules
’0’
Divider Chain
f0
f1
f2
f3
f4
fn
SR3.FCLO = ’1’
VDD
SR1.CPUFST
.
.
.
PH2
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 37
Section “HW Options” gives details about HW option con-
trolled clocks, their selection and their activation.
Note that specifying 1/1.5 and 1/2.5 prescaled clocks results
in clock signals with 33%, respectively 20% duty factor.
Two Clock Output signals CO0 and CO1 provide external
visibility of internal clocks.
Fig. 4–3: Clock Outputs Diagram
Signal CO0 is the output of a prescaler and a 4 to 1 multi-
plexer. Prescaler and input for the multiplexer are selectable
by HW options (see Table 4–5). The output selection of the
multiplexer is done by register CO0SEL, bits CO01 and
CO00. The outputs of the prescalers are fed not only to the
ports, but may also serve as interrupt source. The U-Ports
assigned to function as clock outputs (see Table 4–5) have to
be configured Special Out.
The interrupt source output of this module is routed to the
Interrupt Controller logic. But this does not necessarily select
it as input to the Interrupt Controller. Check section “Interrupt
Controller” for the actually selectable sources and how to
select them.
CO0 and CO1 are not affected by CPU SLOW mode.
CO00, CO01 Clock Out Bit 0 and 1
w: Clock selection
CO10 WAKE/IDLE
w 0 : Clock Out
w 1: SMX Out
2
CO0
CO0SEL.CO00,
1/1
1/1.5
1/2.5
4:1
Mux
CO0SEL.CO01
CO1
CO0
Interrupt
Source
CO1
Interrupt
Source
F1Mux0
F1Mux1
F1Mux2
SMXout
HW Options
HW Option
HW Option
Clock Out1 1/1
1/1.5
1/2.5
F1Mux3
2:1
Mux
CO1SEL.CO10
FFAB [6:5]
HW Option
FFAC [6:5]
2
2
HW Option
FFAC [4:0]
Clock Out1
Table 4–5: HW Options and Ports
Signal HW Options Initialization
Item Address Item Setting
CO0 CO0
Prescaler FFABh CO0
output U5.5
special out
FMux0 FFABh
FMux1 FFB2h
FMux2 FFB3h
FMux3 FFB6h
CO1 CO1
Prescaler FFACh CO1
output U3.1 or
U5.0
special out
Clock Out FFACh
SMX Out -
CO0SEL Clock Out 0 Selection
76543210
Table 4–6: Clock Out 0 Selection
CO01 CO00
00F1
Mux0
01F1
Mux1
10F1
Mux2
11F1
Mux3
CO1SEL Clock Out 1 Selection
76543210
wxxxxxxCO01CO00
xxxxxx00Res
wxxxxxxxCO10
xxxxxxx0Res
CDC16xxF-E ADVANCE INFORMATION
38 March 31, 2003; 6251-606-2AI Micronas
4.4. EMI Reduction Module (ERM)
The EMI Reduction Module distributes the electromagnetic
energy radiated from VLSI embedded controllers among
many spectral lines in the frequency spectrum.
The module performs modulation of the quartz clock phase
by selecting delay taps in a delay-locked loop in a pseudo
random manner, to arrive at the system clock. The individual
system clock edges thus appear delayed by pseudo ran-
domly selected time values ranging from 0% to 10% of the
quartz signal cycle length.
Differening from PLL-based EMI reduction devices, the max-
imum dislocation of a system clock edge from its quartz ref-
erence is only +10% of the quartz signal cycle length.
Due to its inherent frequency stability the module is fully
applicable to ICs containing clocks, timers and synchronous
communication links to other systems, e. g., CAN interfaces.
Features
– Reduction of Clock Related Electromagnetic Interference
– DLL-Based Modulation of Clock Phase
– Full Quartz Frequency Stability
– Maximum Clock Edge Dislocation +10% of Cycle
Fig. 4–4: Block Diagram
4.4.1. Principle of Operation
4.4.2. General Remarks
The edges of the input clock are discretely delayed by 1/71
to 15/71 (21.1%) of the low time of a clock cycle (see Fig. 4–
5). The selection of the clock edge is done by means of a
pseudo-random sequence.
The integral parts of the ERM are the Delay Locked Loop
(DLL) and the Random Number Generator (RNG).
The DLL consists of a chain of 71 controllable delay ele-
ments. Within a locked loop, the total delay time of the chain
is controlled to equal the low time of the input clock. To
reduce the influence of supply noise the DLL is operated on
AVDD.
The RNG generates a sequence of pseudo-random values.
Three bits serve to select one output from the first eight
delay elements. The selected output represents the modu-
lated clock. The selection of the three bits achieves a “high
randomness” even for short time intervals. In the worst case
the same output is selected three times in succession.
4.4.3. Initialization
After Reset the EMI Reduction Module is in standby mode
(inactive). In standby mode, all internal registers are reset.
Setting the standby bit SR1.ERM activates the ERM.
ttt tttt t
123 13 14 15 16 71
Control
Random
8:1 MUX
0
1
3
XTAL1
delay
12345678
Delay Locked Loop
SR.ERM enable
modulated
system
ERMC.CLKSEL
Number
Generator
clock
clock
clk
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 39
Before entering operation, a setup time of at least 100 µs has
to elapse.
4.4.4. Operation
Setting register ERMC to 01h immediately selects the phase
modulated clock as system clock. Flag ERMC.CLKSEL is
readable and indicates the busy state.
4.4.5. Inactivation
To deactivate the ERM first deselect the modulated clock by
resetting register ERMC to 00h. Then SR1.ERM may be
reset to return the whole module to standby mode.
4.4.6. Precautions
For all modules using the modulated system clock or derived
clocks, the following facts have to be kept in mind:
– The maximum operating frequency is reduced.
– The sampling time point of clocked inputs is modulated.
For this reason, e.g., a CAN or UART module may appear
less failure-tolerant.
– The output timepoint of clocked output signals is modu-
lated as well.
– The sampling time of an ADC is modulated slightly.
4.4.7. Results of Application
According to product spectrum measurements of supply cur-
rent and electromagnetic radiation, the EMI Reduction Mod-
ule is capable of reducing the energy in spectral lines of the
frequency spectrum which occur on multiples of the oscillator
frequency.
The ERM distributes the spectral energy contained in oscilla-
tor harmonics over the spectrum between them. Thus the
noise level is increased by a slight 1 to 2 dB.
The higher the share of system events that are in synchro-
nism with the oscillator (current peaks due to clocking in digi-
tal circuits), the more distinct the reduction in the spectrum.
The reduction takes effect from the 6th oscillator harmonic
up, and reaches values from 6 to12 dB over a wide range of
frequencies.
Fig. 4–5: Timing of modulated clock
4.4.8. Registers
CLKSELClock Select
r/w0:System clock is unmodulated
r/w1:System clock is modulated
Other bits in this register are used for test purposes. They
must all be written to zero for proper operation.
OSCCLK
OSCCLKM
1/71 15/71
(21,1 %)
100 %
1/71 15/71
(21,1 %)
100 %
ERMC EMI Reduction Module Control Regis-
ter
76543210
rxxxxxxxCLKSEL
wxx00000CLKSEL
xx001000Res
CDC16xxF-E ADVANCE INFORMATION
40 March 31, 2003; 6251-606-2AI Micronas
5. Memory and Boot System
5.1. RAM and ROM
On-chip RAM is composed of static RAM cells. It is protected
against disturbances during reset as long as the specified
operating voltages are available.
The boot ROM contains up to 2 KB of firmware boot code.
The functionality is described in section Boot System.
The 100PQFP Multi Chip Module also contains a 256 KB
Flash EEPROM of the AMD Am29F200AB type (bottom boot
configuration) or Am29F200AT type (top boot configuration).
These devices exhibit electrical byte program and sector
erase functions. Erase sectors are of various sizes (Fig. 5–
1). Refer to the AMD data sheet for details.
Future mask ROM derivatives will contain no Boot ROM and
may be specified to contain less internal RAM and ROM than
this IC. RAM will always grow upwards from physical
address 000000h. ROM will grow down from 00FFFFh to
002000h and then upwards from 010000h.
5.1.1. Address Map
Emulation mode: E=1
Native mode: E=0
Table 5–1: Reserved (physical) Addresses
Addresses Mode Usage
00FFA0 - C3 Emu &
Native HW Options
00FFC4 - E3 Emu &
Native Interrupt Controller Vectors
00FFE4 - E5 Native
only (reserved) COP
00FFE6 - E7 Native
only BRK
00FFE8 - E9 Native
only ABORT
00FFEA - EB Native
only NMI (expanded by Interrupt
Controller)
00FFEC - ED Emu &
Native reserved
00FFEE - EF Native
only IRQ
00FFF0 - F1 Emu &
Native Manufacturer ROM Identifica-
tion
00FFF2 Emu &
Native reserved
00FFF3 Emu &
Native Control word
00FFF4 - F5 Emu
only (reserved) COP
00FFF6 - F7 Emu &
Native reserved
00FFF8 - F9 Emu
only ABORT
00FFFA - FB Emu
only NMI (expanded by Interrupt
Controller)
00FFFC - FD Emu &
Native RESET
00FFFE - FF Emu
only IRQ/BRK
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 41
Fig. 5–1: Address Map
6KB RAM
Reserved
256KB
Flash
EEPROM
mirrored
Flash
EEPROM
000000
001800
002000
042000
FFFFFF
EMU MCM
008000
010000
018000
028000
030000
038000
040000
phys.addr.
020000
Bank 5
Bank 8
Bank 0
Bank 1
Bank 2
Bank 3
Bank 4
log.addr.
Alternative Native
log.addr.
000000
010000
00FFFF
020000
01FFFF
030000
02FFFF
040000
03FFFF
041FFF
7FFF
0000
FFFF
8000
8000
9FFF
FFA0
Reserved Addr.
CPGA177 PQFP100
Bank 0
004000
006000
Sector 0,
Sector 1,
Sector 2,
Sector 3,
Sector 4, 64k
Sector 5,
Sector 6,
Sector 0,
001900 CAN2-RAM
CAN1-RAM
CAN0-RAM
CAN-Regs
Ext. I/O
I/O-Reg1
I/O-Reg0
001A00
001B00
001C00
001D00
001E00
001F00
Bank 6
FFFF
8000
Bank 7
FFFF
8000
Bank 1
FFFF
8000
Bank 2
FFFF
8000
Bank 3
FFFF
8000
Bank 4
FFFF
8000
reserved
Reserved
CAN2-RAM
CAN1-RAM
CAN0-RAM
CAN-Regs
Ext. I/O
I/O-Reg1
I/O-Reg0
mirrored
Flash
EEPROM
MCM
PQFP100
6KB RAM
Reserved
CAN2-RAM
CAN1-RAM
CAN0-RAM
CAN-Regs
Ext. I/O
I/O-Reg1
I/O-Reg0
Top Boot Config.
6KB RAM
Bottom Boot Config.
256KB
Flash
EEPROM
Sector 0,
Sector 1, 64k
Sector 2,
Sector 3,
Sector 4, 8 KB
Sector 5, 8 KB
Sector 6, 16 KB
Sector 0,
Flash Boot Loader
F800
Boot ROM Boot ROM Boot ROM
64 KB
64 KB
32 KB
64 KB
lower 8 KBlower 8 KB
32 KB
8 KB
8 KB
upper 8 KB upper 56 KB
CDC16xxF-E ADVANCE INFORMATION
42 March 31, 2003; 6251-606-2AI Micronas
5.2. Memory Banking
The 24-bit address bus of the 65C816 CPU allows access to
16 MB of memory space. The upper 8 bits of the addresses
are delivered as bank address, which is multiplexed with the
data value on the internal data/address lines of the CPU.
This kind of native banking is supported by the 65C816 inter-
nal bank registers (PBR, DBR) by using the processor’s
native 16-bit instruction set.
The CPU may be used to emulate the behavior of the 8-bit
processor W65C02. This CPU only allows access to 64 KB
of memory space. To allow access to the expanded memory
range above 64 KB, an alternative banking logic is imple-
mented.
Fig. 5–2: Block diagram Memory Banking
The banking mode is toggled with flag ABM of the
SR2 register:
ABM = ’0’: Native Banking mode
(default after RESET),
ABM = ’1’: Alternative Banking mode.
The alternative bank no. is programmed in the Alternative
Banking Register = ABR.
5.2.1. Native Banking Mode
The bank address is present on the 65C816 data/address
lines during the first half of each processor bus cycle. To get
the whole address bus available during the second half of
the bus cycles, the bank address is demultiplexed with a
transparent latch, triggered with the processor clock,
PHI2(in). This mode is supported by the native instruction set
since it supports the CPU’s internal bank registers PBR and
DBR, which values are multiplexed on the data/address
lines, no matter if the processor is running in native or emula-
tion mode. If the internal bank registers are not supplied,
e.g., if only the 8-bit instruction set is applied, the address
range is limited to 64 KB, the size of one Native Bank,
because the banking registers keep their initial 0-values
which they get during RESET.
After RESET the native banking mode is active and the
whole 65C816 address range of 24 bits is available.
By setting ABM to “1” the native banking mode is switched
off, and the alternative banking mode is activated instead.
5.2.2. Alternative Banking Mode
To use the CPU with the 8-bit instruction set of the 65(C)02
and reaching more than the addressable 64 KB, a specific
banking hardware is implemented: The physical address
range above 32 KB is separated into several banks of which
only one at a time is enabled and selected by the 8-bit ABR
(Alternative Banking Register), which is programmable like
any other standard 8-bit peripheral register by writing the
desired value into its specific address. The contents of the
ABR are also readable, so the software may check the cur-
rent bank at any time. The applied software is responsible for
programming the ABR with the correct bank number at the
right time. Since the upper 32 KB range is switched immedi-
ately after programming the ABR, correct function is not
guaranteed if it´s changed by a program sequence running in
a switched bank. ABR settings need to be done in the lower
32 KB, which is the non-switchable master bank, resp., alter-
native bank number 0.
With ABR = 0, the lower 32 KB appear as bank 0 in the
upper address range, so bank 0 is identical to the non-swit-
chable master bank. Be careful when operating bank 0 in the
upper 32 KB area. RAM, I/O pages and reserved addresses
may be manipulated unintentionally.
RESET initializes ABR = 1, so as to have control byte and
reset vector in the same physical addresses as with active
native banking mode. Also Interrupt vectors have to reside in
A0 ... A14
Alternative
Banking
Mode =
off
on
Banking
Register
A15 ... A23
65C816
AD0 ... AD14
*AD15
*D0 ... D7 MSB = ’0’
ABA0 ... ABA7
NBA0 ... NBA7 Interrupt
Controller,
A0 ... A14
A15 ... A23
Address
Decoder,
Memory,
I/O
*DBA0 ... DBA7
*Processor internal Bus
modified
LSB = *AD15
ABM
DMA
Logic
Alternative
Native
Bank
Latch
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 43
the alternative bank number 1, because the interrupt control-
ler generates the appropriate address of bank 1, but does
not change the contents of the ABR. Interrupt functions have
to reside in the non-switchable master bank (alternative bank
0). Otherwise they need to be in each used bank, as after
getting the vector, the unchanged contents of the ABR deter-
mine the current bank which is valid if A15 is “1”.
After RESET the alternative banking mode is inactive. By
setting ABM to “1”, the alternative banking mode is switched
on.
5.2.3. Memory Banking Mode Selection
In the Native Banking Mode the logical address range is
identical to the physical one. In the Alternative Banking Mode
A15 is lost, because a non-switched address range is neces-
sary. This cuts the addressable space in half. To retain the
physical address range without holes as is in Native Banking
Mode, the addresses are multiplexed.
Fig. 5–3: Physical Address change depending on the
Banking Mode
ABR Alternative Banking Register
76543210
r/w Alternative Bank Address
00000001Res
NBA7
’0’ A23
ABM
NBA6
ABA7
NBA4
NBA5
ABA5
ABA6
A22
A21
A20
A15
A16
A17
A18
A19
AD15
NBA0
NBA1
NBA2
NBA3
ABA0
ABA1
ABA2
ABA3
ABA4
CDC16xxF-E ADVANCE INFORMATION
44 March 31, 2003; 6251-606-2AI Micronas
Fig. 5–4: Banking Map shown with the max. size of addressable memory (different from the implemented amount)
Alternative Bank 255
.......
Alternative Bank 3
Alternative Bank 2
Native Bank 255
( not reachable as Alternative Bank )
......
( not reachable as Alternative Bank )
Native Bank 127
Alternative Bank 254 *Alternative Bank 255
......
................
Native Bank 2
Alternative Bank 4 *Alternative Bank 5
Native Bank 1
Alternative Bank 2 *Alternative Bank 3
Native Bank 0
Alternative Bank 0 *Alternative Bank 1
( also Alternative Bank 0 * )Alternative Bank 1
Native Bank 255
.......
Native Bank 2
Native Bank 1
Native Bank 0
Alternative BankingNative Banking
FFFFH
8000H
0H
10000H
20000H
.......
7F0000H
.......
FF0000H
8000H
FFFFH
0H
* The logical address range is 8000H ... FFFFH
Native Banks: logical addresses = physical addresses
Alternative Banks: logical address range = 8000H ... FFFFH
Banking Cross Reference
FFFFH
0H
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 45
5.3. Boot System
The Boot System offers a very flexible method for the con-
troller to receive and store data and programs, no matter if
the Flash memory contains a working or faulty application, or
no application at all. With the Boot System it is possible to fill
the RAM with data and functions and to start execution from
any address. Tasks like flash programming and diagnostic
routines may be downloaded and started, either in lab, fac-
tory or in system.
After RESET, the CPU executes code from the Boot ROM.
Its firmware, the Boot Loader, checks whether the application
software in the flash memory must be started or if data must
be downloaded via one of 6 receivers: 3 UARTs and 3 CANs.
5.3.1. Principle of operation
After RESET the Boot ROM is active, if the Test pin is left
vacant or connected to ground. The Boot Loader is started
and checks whether there is a wake-up from Power Saving
Mode (see Table 6–7 on page 54), i. e., if CSW2.WKID is
set. If so, it starts the application software in Flash Memory
(or ROM). Otherwise, it tries to detect a download condition.
If no appropriate data can be received via either of its 6 inter-
face receivers within a predefined time, the Boot Loader ter-
minates itself by copying a program sequence into the RAM
and executing it. This software part switches the Boot ROM
off and starts the application software assumed within the
Flash Memory. The detour via RAM is necessary because
the Boot ROM hides the start-up address area of the flash
memory that contains the processor Control Word and
RESET Vector of the application.
When detecting a download condition, the time-out condition
to terminate the boot sequence is extended with the detec-
tion of every new appropriate data until the download is com-
pleted. The Boot Loader jumps to a start address, which is
assumed part of the code-header downloaded before.
Neither the Boot ROM is switched off, nor the Control Word
or RESET Vector of the flash memory are treated automati-
cally if a download occurs. This leaves a maximum of flexibil-
ity during and after the boot procedure.
5.3.2. The Boot ROM
The 2 kB Boot ROM covers the address range from 0F800H
to 0FFFFH. With Test pin left vacant or pulled to low level at
RESET, the Control Word is read out of the Boot ROM and
copied into the Control Register. With the flag FLASH of the
Control Register set to ’0’ the Boot ROM stays switched on
and its control program, the Boot Loader, is started.
In a standard application (no download request) the control
program in the flash memory is started after the Boot Loader
has set the flag FLASH in the Control Register to ’1’ to switch
the Boot ROM off at the appropriate time, which enables the
Flash ROM in the same address range as the Boot ROM
before.
Fig. 5–5: Principle of the boot system
Fig. 5–6: Boot ROM Selection
RESET
Boot Loader
start at downloaded
Download
data to
receive ?
no
yes
Boot ROM
RAM
Termination
Start
address
Flash ROM
Application
Start
affected by SW
affected by HW
CR.FLASH
Boot ROM
Control Word
RESET Vector
Interrupt Vectors
Flash ROM
Control Word
RESET Vector
Interrupt Vectors
RAM
or downloaded
Start application
code
enableenable
Application
CR.IROM &
&
CDC16xxF-E ADVANCE INFORMATION
46 March 31, 2003; 6251-606-2AI Micronas
5.3.3. The Boot Loader
Fig. 5–7: Boot Loader flow-chart
During the first 480000 clock cycles (= 6 ms at 8 MHz pro-
cessor clock) after reset, the Boot Loader tries to detect a
start of the download condition, the so-called Boot Identifier,
at one of the receivers of the 3 UARTs and 3 CANs.
If the complete Boot Identifier has been received within the
appropriate time, the watchdog is initialized (with the value 6
for effective 6.144 ms at fXTAL = 8 MHz) to supervise further
time-out conditions. This is not only to increase security, but
also because the Boot Loader could have been started by a
reset other than power-on, from a running application in
which the watchdog was programmed before. Therefore the
Boot Loader has to program it, too, otherwise it generates a
reset after the max. WD time-off, which is default after reset.
Within further time-out conditions additional Boot Identifiers
may occur, followed by the Download Header and the Down-
load Data including separate Check Sums. At the end of the
procedure the processor executes a jump to the start
address which is part of the Download Header.
If a time-out condition is met, an unexpected byte has been
received or if a Check Sum error occurs, a reset is forced by
the watchdog to start the Boot Loader again. This subse-
quently will start the application because of absence of the
Boot Identifier.
If the Boot Loader does not detect the Boot Identifier within
the time limit, the application software assumed in the Flash
ROM is started. For thís purpose a software part is copied
into RAM and started. It toggles the flag FLASH of the Pro-
cessor Control Register to switch the BOOT ROM off and
reads the Control Word out of the Flash ROM, which is
accessible now, and programs the processor’s Control Reg-
ister with that value. Last, it starts the application by a jump
to the address defined in the RESET Vector of the Flash
ROM.
The application can program its own watchdog value, as it
was not treated by the Boot Loader before.
5.3.4. Used RAM
The Boot Loader uses and thus destroys the contents of
addresses F5H to FFH in zero page, as well as 100H to
10AH and 1E8H to 1FFH in (stack) page 1, and 200H to
20DH in page 2.
RESET
yes
no
init. UART and CAN receivers
Application
no
yes
no
no
yes
yes
init. Watchdog
no
yes
no
yes
Start downloaded code
Boot
Identifier
time-out
Download
Header
time-out
Data
Length
time-out
reset UART and CAN interfaces
init time-out
refresh Watchdog
no
yes
Boot
Identifier
force RESET
refresh Watchdog
download data
= 6ms @ fXTAL = 8 MHz
yes
no
CSW2.WKID = ’1’
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 47
5.3.5. Interfacing to a Download Source
Either one of the UARTs or CANs can be used to activate a
download. The Boot Loader assumes a single-bus line as
connection to the host. So either wired-or Rx/Tx interfaces or
separate driver circuits may be used.
To allow download to several targets connected to the same
bus, the Boot Loader does not generate a receipt. Since the
UART and CAN busses work asynchronously, target
response would result in bus collisions.
5.3.6. Interface Protocol
The interface protocol falls into three sections:
1. Boot Identifier. It is continuously transmitted by the host.
After reset, the Boot Loader waits for the Boot Identifier.
2. Download Header. It contains the start address, target
address and data length.
3. Download Data. After the Download Header, the
announced number of data bytes are transmitted by the host.
The first byte is stored in the address defined in target
address, the next one in target address + 1, and so on. After
reception of the last byte the processor immediately exe-
cutes a jump to the address in start address. Because the
watchdog is treated during download, the downloaded pro-
gram must also refresh it with the same value (= 6 for effec-
tive 6.144 ms at fXTAL = 8 MHz), starting with 06H, followed
by its complement 0F9H, 06H, 0F9H, ... and so on.
No more than 6 ms (at fXTAL = 8 MHz) delay is allowed as
delay between the single telegrams during boot identification
and download, otherwise the Boot Loader stops reception
and starts the application in Flash.
Time-out value (6 ms) and baud rate of UART and CAN
relate to the processor clock (fXTAL).
As start of a download, the host has to keep generating Boot
Identifiers while the targets get a reset. After a minimum of
time, when the target running is stable, the host sends the
Download Header, followed by the Download Data.
5.3.6.1. UART Protocol and Timing
Fig. 5–8: UART Timing
The expected telegram format is: 1 Start bit, 8 Data bits, 1
even parity bit and 1 Stop bit at a rate of 9600Bd at 8 MHz
processor clock. To use standard K-bus drivers the UART
transmitter inverters are switched on.
Boot Identifier
After reset the Boot Loader tries to detect the Boot ID which
consists out of 3 bytes with the values 26H, D6H and A3H
respectively, of which each is expected within 6 ms (at fXTAL
= 8 MHz) after the preceding one. The telegram may start
with any value out of this queue, but the bytes have to
appear in the right order. Any other value or the event of a
bigger delay being detected, stops the Boot ID detection and
starts the application.
Download Header
The Download Header has to follow the last byte of the Boot
ID, the A3H, and starts with an ASCII SOH = 01H. It is fol-
lowed by the start and target addresses of the download pro-
gram and its length as 16-bit values of which the low bytes
are transmitted first. The Download Header terminates with a
16-bit Check Sum, low byte transmitted first, which is the 2’s
complement of the sum, calculated from SOH up to and
including Data Length. The sum of all bytes in the Download
Header (including the Check Sum) should be zero.
Boot Identifier
A3H 26H D6H A3H SOH
Download Header, Data
RESET to <= 6ms @ fXTAL = 8 MHz
to
Rx
n bytes of Boot Identifier
26H D6H A3H
D6H A3H 26H
A3H 26H D6H
or
or 26H
D6H
toto
in correct order
D6H A3H
A3H 26H
totototototo
Table 5–2: Download Header
Field Size Meaning
SOH 8 bit start of Download Header
Start Address 16 bit Boot Loader jumps to this
address after download.
Target
Address 16 bit Boot Loader writes data to this
address.
Data Length 16 bit Number of data bytes trans-
mitted after Download Header
(Check Sum excluded).
Check Sum 16 bit 16-bit Check Sum from SOH
to data length
CDC16xxF-E ADVANCE INFORMATION
48 March 31, 2003; 6251-606-2AI Micronas
Download Data
As many data bytes as defined in Data Length are expected,
plus a 16-bit Check Sum, low byte transmitted first, which is
the 2’s complement of the sum, calculated over all data. The
sum of all bytes (including the Check Sum) should be zero.
5.3.6.2. CAN Protocol and Timing
Fig. 5–9: CAN Timing
The Boot Loader expects CAN telegrams in standard format
(11 bit identifier) at a bit rate of 125kBd at 8 MHz processor
clock with control parameters set as follows: BPR = 7,
TSEG1 = 3, TSEG2 = 2, SJW = 3, rx not inverted.
Boot Identifier (ID = 7FFH, DLC = 0)
After reset the Boot Loader expects the Boot Identifier within
6ms (@ fXTAL = 8 MHz), what consists out of a telegram
with ID = 7FFH and DLC = 0. As the transmitter is not
active, at least one additional working CAN node has to
work correctly on the bus, to generate the CAN bus
acknowledge bits! If the Boot ID can not be detected in
time, the Boot Loader starts the application.
Download Header (ID = 627H, DLC = 6)
The Download Header telegram has to follow within 6 ms (@
fXTAL = 8 MHz) after the Boot ID. It contains the start and
target addresses of the download program and its length as
16-bit values with low bytes first.
Download Data (ID = 626H, DLC = 1 ... 8)
As many data bytes as defined in Data Length are expected,
plus a 16-bit Check Sum, low byte transmitted first, which is
the 2’s complement of the sum, calculated over all data. The
sum of all bytes (including the Check Sum) should be zero.
RESET
Rx 7FF 627 6267FF 626
Download
tototo
Boot Identifier
to
to <= 6ms @ fXTAL = 8 MHz
DLC = 0 DLC = 6
7FF
DLC = 0 DLC = 1 ... 8 DLC = 1 ... 8
Header
Download
Data
to
n
Boot
Identifiers
Download
Data
Table 5–3: Download Header
Field Size Meaning
Start Address 16 bit Boot Loader jumps to this
address after download.
Tar get
Address 16 bit Boot Loader writes data to this
address.
Data Length 16 bit Number of data bytes trans-
mitted after Download Header
(Check Sum excluded).
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 49
6. Core Logic
6.1. Control Register CR
The Control Register CR serves to configure the ways, by
which certain system resources are accessed during opera-
tion. The main purpose is to obtain a variable system config-
uration during IC test.
Upon each HIGH transition on the RESETQ pin internal
hardware reads data from address location 00FFF3h and
stores it to the CR. The state of the TEST and ESTOPCLK
pins at this point in time, specifies which program storage
source is accessed for this read.
The system will thus start up according to the configuration
defined in address location 00FFF3h, automatically copied to
register CR.
RESLNG Reset Pulse Length
r/w1: Pulse length is 4095/FXTAL
r/w0: Pulse length is 16/FXTAL
This bit specifies the length of the reset pulse which is output
at pin RESETQ following an internal reset. If pin TEST is 1
the first reset after power on is short. The following resets
are as programmed by RESLNG. If pin TEST is 0 all resets
are long.
TSTTOG TEST Pin Toggle (Tables 6–2 and 6–3)
This bit is used for test purposes only. If TSTTOG is true in
IC active mode, pin TEST can toggle the multifunction pins
between Bus mode and normal mode.
EBTRI Emulator Data Bus Tristate (Table 6–3)
MFM Multifunction pin Mode
(Tables 6–2 and 6–3)
TSTROM TestROM (Table 6–4)
FLASH FLASH EEPROM (Table 6–5)
IROM Internal ROM (Tables 6–4 and 6–5)
Table 6–1: Control byte source
TEST Control byte source
0 or NC internal BOOT ROM
(standard for stand-alone operation)
1 external via Multifunction pins in Bus
mode (for test purposes only)
CR Control Register
76543210
r/w RESLNG TSTTOG x MFM TSTROM IROM IRAM ICPU ROM
r/w RESLNG TSTTOG EBTRI MFM FLASH IROM IRAM ICPU Emu
Value of 00FFF3h Res
Table 6–2: TSTTOG and MFM usage in mask ROM parts
TSTTOG MFM TEST pin Multifunction
Pins
0 0 x Bus mode
100Bus mode
1 normal mode
x 1 x normal mode
Table 6–3: TSTTOG, EBTRI and MFM usage in Flash and
EMU parts
TST-
TOG EBT
RI MFM TEST
pin Multi-
function
Pins
Emula-
tor Bus
Pins
0x0xBus
mode Flash
mode
1x00Bus
mode Flash
mode
1 normal
mode
x 0 1 x normal
mode Eulator
mode
1Flash
mode
Table 6–4: TSTROM and IROM usage in mask ROM parts
TSTROM IROM selected program storage
1 1 internal ROM
0 internal TestROM
x 0 external via Multifunction pins
in Bus mode
CDC16xxF-E ADVANCE INFORMATION
50 March 31, 2003; 6251-606-2AI Micronas
IRAM Internal RAM
r/w1: Enable internal RAM.
r/w0: Disable internal RAM.
ICPU Internal CPU
r/w1: Enable internal CPU.
r/w0: Disable internal CPU.
6.2. Reset Logic
6.2.1. Alarm Function
An alarm comparator on the pin RESETQ allows the detec-
tion of a threshold higher than the reset threshold. An alarm
interrupt can be triggered with the output of this comparator.
The interrupt source output of this module is routed to the
Interrupt Controller logic. But this does not necessarily select
it as input to the Interrupt Controller. Check section “Interrupt
Controller” for the actually selectable sources and how to
select them.
The intended use of this function is made, when a system
uses a 5 V regulator with an unregulated input. In this case,
the unregulated input, scaled down by a resistive divider, is
fed to the RESETQ pin. With falling regulator input voltage
this alarm interrupt is triggered first. Then the reset threshold
is reached and the IC is reset before the regulator drops out.
The time interval between the occurrence of the alarm inter-
rupt and the reset may be used to save process data to non-
volatile memory. In addition, power saving steps like turning
off stepper motor drivers may be taken to increase the time
interval until reset. The alarm interrupt is a level triggered
interrupt. The interrupt is active as long as the voltage on pin
RESETQ remains between the two thresholds of alarm and
reset (see Fig. 6–1 on page 51).
6.2.2. Internal Reset Sources
This IC contains three internal circuits that are able to gener-
ate a system reset: watchdog, supply supervision and clock
supervision.
All three internal resets are directed to the open drain output
of pin RESETQ. Thus a “wired or” combination with external
reset sources is possible. The RESETQ pin is current limited
and therefore large external capacitances may be con-
nected.
All internal reset sources initially set a reset request flag.
This flag activates the pull-down transistor on the RESETQ
pin. An internal reset prolongation counter starts, as soon as
no internal reset source is active any more. It counts 4096
FXTAL periods (for alternative settings refer to register CR)
and then resets the reset request flag, thus releasing the
RESETQ pin.
As long as the reset input comparator on the pin RESETQ
detects the low level, all IC registers are held in reset state.
6.2.2.1. Supply Supervision
An internal bandgap reference voltage is compared to VDD.
A VDD level below the Supply Supervision threshold VREF-
POR will permanently pull the pin RESETQ low and thus
hold the IC in reset (see Fig. 6–2 on page 52). With HW
Option FFB8h, bit6 = 0, this reset source can be enabled/dis-
abled by flag CMA in register CSW0 (see Section 6.2.2.2.
on page 50).
6.2.2.2. Clock Supervision
The Clock Supervision monitors the frequency at the oscilla-
tor input XTAL1. A frequency level below the clock supervi-
sion threshold of approx. 200 kHz will permanently pull the
pin RESETQ low and thus hold the IC in reset (see Fig. 6–2
on page 52). With HW Option FFB8h, bit6 = 0, this reset
source can be enabled/disabled by flag CMA in register
CSW0.
A frequency exceeding the specified IC frequency is not
detected.
There are two general operation options which can be
selected in the HW Options field (address FFB8h):
1. Bit6 = 1:
Clock and Supply Supervision are permanently active. They
can not be deactivated. The Watchdog must be serviced by
SW. This mode is recommended for all stand-alone applica-
tions requiring high operational reliability.
2. Bit6 = 0:
Clock and Supply Supervision are active after reset, but can
be enabled/disabled by the clock monitor active flag CMA of
register CSW0. The Watchdog must be serviced only after a
first write access to register CSW1. This mode is recom-
mended for test and evaluation purposes only.
Table 6–5: FLASH and IROM usage in FLASH and EMU
parts
FLASH IROM selected program storage
1 1 internal FLASH EEPROM
resp. Emulator Bus
0 internal BOOT ROM
x 0 external via Multifunction pins
in Bus mode
Table 6–6: Some commonly used settings for address
location 00FFF3h. A copy is automatically transferred to the
CR during IC start-up.
Code TEST
Pin Operation Mode
FFh 0 Stand-alone with internal ROM or Flash
ABh 1 External program storage connected to
Multifunction pins in Bus mode
DFh 0 Emulator mode (CPGA177 package)
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 51
Fig. 6–1: UVDD Section
6.2.2.3. RESET Comparator
During power saving mode, the comparator function is not
available and is bypassed by a simple CMOS Schmitt input.
Full CMOS levels are thus required on this input in this
mode.
VBG Generator
10% UVDD
UVSS
RESET/
ALARM
Interrupt
RESETQ
en
+
-
en ALARM Comp.
-
+
en
RESET Comp.
RES
Supply Supervision
en
TST.MIX
VBG
UVDD
Logic
-
+
Source
SR3.WAID
-
+
-
+
LCD Supply
en
1/3UVDD
2/3UVDD
SR0.LCD
XTAL Oscillator
XTAL1
XTAL2
VSS
TEST
POR
&
≥1
en
≥1
OSC.XM
UVDD
X
RESBP
≥1
MOSPOR
SR3.XTAL run
≥1
DBG.DSC
’1’
’0’
RTCCLK
OSCCLK
ON
VDD
2*VBG
2.88*VBG
VBG
CDC16xxF-E ADVANCE INFORMATION
52 March 31, 2003; 6251-606-2AI Micronas
Fig. 6–2: Reset Logic Block Diagram
6.2.2.4. Watchdog
The Watchdog module serves to monitor undisturbed pro-
gram execution. A failure of the program to retrigger the
Watchdog within a preselectable time will pull the pin
RESETQ low and thus reset the IC (Fig. 6–2 and 6–3). With
HW Option FFB8h, bit6 = 0, this reset source is only enabled
after a write access to register CSW1 (see Section 6.2.2.2.
on page 50).
>1
SQ
R
clock
supervision
VREFR port reset
RESETQ
reset in WDRES
SQ CSW2.CLM
CSW2.POR
CSW2.PIN
R
SQ
R
SQ
R
power on
&
≥
1
CM.WCM
HW option
CSW0.CMA
VDD 1
0
POR
CLS
fXTAL
fSUP
Reset extension
16 or 4096
fXTAL clocks
&
wr CSW2
≥
1
res CSW2
SQ
R
CSW2.FHR
DB4
core reset
RES
≥
1
-
+
SR3.WAID
RESINT
WAKE_RES
Watchdog
SQ
R
CSW2.WKID
≥
1
&
DBG.DCS
COMPRES RQ
G
D
MOSPOR
&
WKX
≥
1
&
MOSPOR
SR3.WAID
&
SQ
R
CSW1.WDRES
≥
1
&
wr CSW1
&
RESINTOUT
RESIR
MOSPOR
RESCORE
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 53
Fig. 6–3: Watchdog Block Diagram
The Watchdog contains a down-counter that generates a
reset when it wraps from zero to FFh. It is reloaded with the
content of the watchdog timer register, when, on a write
access to register CSW1, watchdog trigger registers 1 and 2
contain bit complemented values. An IC reset resets the
watchdog timer register to FFh, thus forcing the Watchdog to
create a maximum reset interval.
The Watchdog is controlled by register CSW1. The first write
access to it loads the timer register value setting the Watch-
dog’s unretriggered reset interval. The desired interval can
be programmed by setting the CSW1 value to:
The resolution of the Watchdog is 8192/fCPU.
The second and all following even-numbered write accesses
load watchdog trigger register 1, the third and all following
odd numbered write accesses load watchdog trigger
register 2.
In all future, the SW has to write alternatingly to register
CSW1 value and bit complement value, thus retriggering the
up-counter. Failure to retrigger will result in an overflow of the
up-counter generating a Watchdog reset.
It is not allowed to change a chosen value. Writing a wrong
value to CSW1 immediately sets the flag WDRES in register
CSW1 and prohibits further retriggering of the watchdog
counter.
WDRES is true after a Watchdog reset. Only a Supply
Supervision reset or a write access to register CSW1 clears
it.
6.2.3. External Reset Sources
As long as the reset input comparator on the pin RESETQ
detects the low level, the overall IC is reset. On this pin,
external reset sources may be wire-ored with the IC internal
reset sources, leading to a system-wide reset signal combin-
ing all system reset sources.
Trigger Reg1 Timer RegisterTrigger Reg2
8-Bit-Counter
2.write 3.write
1. write
8
≥1
8 8
zero
reset out
CPUFST=1: fosc * 2-13
clk
load
reset in
wr CSW1
power on SQ
R
CSW1.WDRES
CSW1
& even & odd
CSW1CSW1
&
SQ
R
&
=
1. write
VDD
HW Option
≥1
CPUFST=0: fosc * 2-13 * 2-8
1. write
C
Q
S
≥1
&
D
2.write & even
3.write & odd
Value Interval fCPU
×
8192
------------------------------------1–=
CDC16xxF-E ADVANCE INFORMATION
54 March 31, 2003; 6251-606-2AI Micronas
6.2.4. Summary of Module Reset States
After reset the IC modules are set to the reset state
(Fig. 6–7)
6.2.5. Reset Registers
This register controls the Supply and Clock Supervision
modules.
CMA Clock and Supply Monitor Active
w1: Both Enabled.
w0: Both Disabled.
This register controls the Watchdog module. Only values
between 1 and 255 are allowed.
WDRES Watchdog Reset Source
r1: Watchdog was reset source.
Any write access to CSW1 resets this flag.
First write the desired watchdog time value to this register.
On further writes, to retrigger the Watchdog, alternatingly
write a value (not necessarily the former time value) and its
bit complemented value. Using other values or changing the
order of both values will cause the watchdog to generate a
RESET.
The RTC can not be reset.
Table 6–7: Status after Reset
Module Status
CPU CPU FAST mode (fOSC).
Interrupt
Controller Interrupts are disabled. Priority regis-
ters, request flip flops and stack are
cleared.
U-Ports Normal mode. Output is tristate.
High current
ports Normal mode. Output is low.
LCD module Registers are reset. No display.
Watchdog Depends on mask option.
EMU option: Switched off. SW activa-
tion is possible.
Stand-alone option: Permanently
active.
Clock monitor Depends on mask option.
EMU option: Active. SW may toggle.
Stand-alone option: Permanently
active.
CSW0 Clock, Supply & Watchdog Register 0
76543210
CSW1 Clock, Supply & Watchdog Register 1
76543210
wxxxxxxxCMA
xxxxxxx1Res
rxxxxxxxWDRES
wWatchdog Time and Trigger Value
11111111Res
Table 6–8: Source of last Hardware Reset
*)
CSW2.WKID
CSW2.FHR
*)
CSW2.CLM
*)
CSW2.PIN
*)
CSW2.POR
CSW1.WDRES
Reset Source
000100external from RESETQ pin
000101internal Watchdog
001100internal Clock Supervision
001110internal Supply Supervision
010100internal Forced Hardware
1 x x 1 x x wake-up from WAKE/IDLE
The registers sum up the source of all HW resets that
occurred since the last write to register CSW2.
*)Any write access to CSW2 resets these flags to 0.
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 55
TST TEST Pin State
r1: TEST is 1.
r0: TEST is 0.
WKID Wake Reset from WAKE/IDLE
Mode (Table 6–8)
FHR Forced Hardware Reset
r0: (Table 6–8)
r1: (Table 6–8)
w1: force Reset
w0: no action
CLM Clock Supervision Reset (Table 6–8)
PIN RESETQ Pin Reset (Table 6–8)
POR Supply Supervision Reset (Table 6–8)
6.3. Standby Registers
The Standby Registers SR0, SR1, SR2 and SR3 allow the
user to switch on/off power or clock supply of single mod-
ules. With these flags it is possible to greatly influence power
consumption and its related electromagnetic interference.
For details about enabling and disabling procedures and the
standby state refer to the specific module descriptions.
The minimum IC current consumption is obtained with SR0,
SR1, SR2 and SR3 set to 00h.
SM Stepper Motor
r/w1: Module active.
r/w0: Module off.
PWM1 Pulse Width Modulator 1
r/w1: Module active.
r/w0: Module off.
PWM0 Pulse Width Modulator 0
r/w1: Module active.
r/w0: Module off.
UART2 UART 2
r/w1: Module active.
r/w0: Module off.
SPI1 SPI 1
r/w1: Module active.
r/w0: Module off.
CAN0 CAN Module 0
r/w1: Module active.
r/w0: Module off.
CCC Capture Compare Counter
r/w1: Module active.
r/w0: Module off.
SPI0 SPI 0
r/w1: Module active.
r/w0: Module off.
LCD LCD Module
r/w1: Module active.
r/w0: Module off.
CPUFST CPU FAST Mode
r/w1: FAST mode: FCPU = FXTAL
r/w0: SLOW mode: FCPU = FXTAL / 256
PSLW Port Slow Mode
r/w1: Slow mode.
r/w0: Fast mode.
UART0 UART 0
r/w1: Module active.
r/w0: Module off.
ADC ADC Module
r/w1: Module active.
r/w0: Module off.
P0DIN Port 0 Digital Input
r/w1: Enable digital inputs on analog ports P0.1 to
P0.5
r/w0: Disable digital inputs on analog ports P0.1 to
P0.5.
TIM1 Timer 1
r/w1: Module active.
r/w0: Module off.
ERM EMI Reduction Module
r/w1: Module active.
r/w0: Module off.
CSW2 Clock, Supply & Watchdog Register 2
76543210
rTST xWKID FHR CLM PIN POR x0
wx x 0 FHR 000x0
-x------Res
SR0 Standby Register 0
76543210
r/w SM PWM1 PWM0 UART2 SPI1 CAN0 CCC SPI0
00000000Res
SR1 Standby Register 1
76543210
SR2 Standby Register 2
76543210
r/w LCD CPUFST PSLW UART0 ADC P0DIN TIM1 ERM
01000000Res
r/w TIM2 PWM3 PWM2 UART1 PWM4 DGB EXTIR ABM
00000000Res
CDC16xxF-E ADVANCE INFORMATION
56 March 31, 2003; 6251-606-2AI Micronas
TIM2 Timer 2
r/w1: Module active.
r/w0: Module off.
PWM3 Pulse Width Modulator 3
r/w1: Module active.
r/w0: Module off.
PWM2 Pulse Width Modulator 2
r/w1: Module active.
r/w0: Module off.
UART1 UART 1
r/w1: Module active.
r/w0: Module off.
PWM4 Pulse Width Modulator 4
r/w1: Module active.
r/w0: Module off.
DGB DIGITbus Master
r/w1: Module active.
r/w0: Module off.
EXTIR External Interrupt
r/w1: Enable functions IRQ on pin U5.4 and
ABORTQ on pin U5.5
r/w0: Disable functions IRQ on pin U5.4 and
ABORTQ on pin U5.5
ABM Alternative Banking Mode
r/w1: Alternative Banking mode enabled
r/w0: Native Banking mode enabled.
XTAL Quartz Oscillator Mode
r/w1: Start-Up Mode active (default after Reset).
r/w0: Run Mode active.
WAID WAKE/IDLE
(Reset with COMPRES || MOSPOR)
r/w 0 : WAKE/IDLE mode off
r/w 1: WAKE/IDLE mode on
FCLO Fast Clock off
r/w 0 : Fast clock on
r/w 1: Fast clock off
CAN1 CAN Module 1
r/w1: Module active.
r/w0: Module off.
CAN2 CAN Module 2
r/w1: Module active.
r/w0: Module off.
6.4. Test Registers
Test registers are for factory test only. They must not be writ-
ten by the user with values other than their reset values
(00h). They are valid independent of the TEST input state.
In all applications where a hardware reset may not occur
over long times, it is good practice to force a software reset
on these registers within appropriate intervals.
SR3 Standby Register 3
76543210
r/w x x x XTAL WAID FCLO CAN2 CAN1
xxx10
*) 000Res
TST1 Test Register 1
76543210
TST2 Test Register 2
76543210
TST3 Test Register 3
76543210
wFor testing purposes only
00000000Res
wFor testing purposes only
00000000Res
wFor testing purposes only
00000000Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 57
7. Multiplier
The Multiplier provides the following function:
- Calculation of 8 bits x 8 bits = 16 bits
Features
– needs no wait states
– no status signals necessary to be checked separately by
the application program
– for the application program the result is immediately avail-
able after writing into the multiplier register
– the registers for multiplicand and multiplier are not only
writable but also readable
Fig. 7–1: Block diagram of the multiplier
+
Data Bus
8
+
multiplicand
multiplier
8
double shift right
load
sh2
8
&
8
8
1 0
&
8
bit
COA, SUMA[7:1]
COB, SUMB[7:1]
latch
clck
8
LSByte
Result
product
16
16
MSByte
Result
2-bit counter
16
8
SUMA[0]
SUMB[0]
8
8
CDC16xxF-E ADVANCE INFORMATION
58 March 31, 2003; 6251-606-2AI Micronas
7.1. Functional Description
To obtain a 16-bit product, the application program first has
to write the first factor into the 8-bit multiplicand register and
then the second factor into the multiplier register. Writing the
multiplier starts the multiplication.
The execution of the multiplication takes 4 CPU clock (PH2)
cycles. CPU instructions with absolute addressing modes
take at least 4 cycles to access the product register. Accord-
ing to this, immediately after writing into the multiplier regis-
ter, the result is available for the application program.
7.2. Registers
Two 8-bit registers serve as input and a 16-bit register deliv-
ers the result: Writing into the Multiplier register starts the execution of the
multiplication.
The Product register MULPROD holds the result of the multi-
plication after 4 CPU clock (PH2) cycles have elapsed.
7.3. Operation of the Multiplier
Since the execution of a multiplication takes less CPU cycles
than an application program needs to access the result reg-
ister, multiplications are done by simply writing the multipli-
cand followed by the multiplier and reading the product.
Precautions have to be taken in application programs using
the multiplier, and in application programs which may be
interrupted by a service routine also using the multiplier. If
the procedure of writing the multiplicand and multiplier until
reading the result register is interrupted, and the interrupt
service routine overwrites the multiplier input registers, the
result read by the background program is wrong.
To solve this problem, the multiplier input registers have to
be saved before and restored after a multiplication takes
place in the interrupt service routine.
MULCAND Multiplicand
76543210
MULPLIER Multiplier
76543210
r/w multiplicand
xxxxxxxxRes
r/w multiplier
xxxxxxxxRes
MULPROD Multiplication Product
76543210
Offs
rProduct high byte 1
rProduct low byte 0
xRes
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 59
8. Power-Saving Module (PSM)
To reduce the power consumption one of two Power-Saving
Modes (WAKE and IDLE) can be selected. Non-power-sav-
ing modes are DEEP SLOW, SLOW and FAST, which differ
from Power-Saving Modes by having the CPU running
instead of a stopped CPU clock during WAKE/IDLE active.
Most of the core logic is switched off in a Power-Saving
Mode. Only hardware necessary for WAKE/IDLE is supplied.
Features
– Power reduction down to leakage current of wake mode
possible
– Real-Time Clock (RTC) Module
– Clock source: Built-in RC oscillator or XTAL
– Up to 10 edge and level triggered Wake Ports
– Wake sources: RTC Module and/or Wake Ports
– Polling Module for cyclic scan of the Wake Ports
– Interrupt outputs of RTC Module and Port Wake Module
Fig. 8–1: Power-Saving Module
The major task of the Power-Saving Module is to supply a
wake-up signal (WAKE_RES) for the main system or to gen-
erate an interrupt, as shown in figure 8–1. WAKE_RES is
necessary to get the IC out of a Power-Saving Mode. Apart
from WAKE_RES, a Power-Saving Mode can only be left by
activating RESETQ at pin or by a power on reset.
WAKE_RES is generated by a Port Wake Module for an
event-driven wake-up, combined with an RTC Module for a
cyclic wake-up.
The Power-Saving Module is active all the time, during
power-saving mode as well as during non-power-saving
mode (CPU active modes). The WAKE_RES output signal
can be generated during power-saving mode only.
The RTC Module has to provide the time of the day accu-
rately down to a second, and generate an output signal,
which can be used to trigger an interrupt or a wake-up signal.
The RTC Module can be clocked by the on-chip quartz oscil-
lator or by the Power-Saving Module built-in RC oscillator.
The Polling Module cyclically outputs a high pulse of pro-
grammable duration at port H2.5. Some of the RTC Module
taps are connected to the Polling Module for deriving the
pulse period and duration.
The Port Wake Module merges several Wake Ports and out-
puts a signal that can be used to trigger an interrupt or a
wake-up signal.
H2.5
4 ... 12MHz
Port
Wake
RTC
WAKE_RES
Up to
ten
Wake
Ports
≥1&
enable Core
Logic
UVDD
Analog
Sections
RC
Oscillator
20..50kHz
Reset
Logic
IR
RTC
IR
WAPI
Module
Polling
Module
Module
AVDD,
RTC wake-up
Port wake-up
CDC16xxF-E ADVANCE INFORMATION
60 March 31, 2003; 6251-606-2AI Micronas
Fig. 8–2: Power-Saving Module Block Diagram
8.1. Functional Description
The power-saving logic combines all wake-up sources. It
contains an RC oscillator, a 20-bit sub second counter, an
RTC, multiplexers to select different taps of the sub second
counter or of the RTC, a Polling Module and the logic for up
to ten Wake Ports.
The RTC Module output can generate an interrupt or a wake-
up signal. All Wake Ports can generate a collective interrupt
or a wake-up reset. The Polling Module drives an H-Port and
generates a strobe signal to enable a Wake Port to trigger on
a dedicated input level. Several internal clock signals can be
output via CO1 (SMX_out).
8.1.1. RTC Module
To work as real-time counter, a sub-second counter (SSC)
with its reload register (SSR) and a seconds, minutes and
hours counter are part of the Power-Saving Module. As input
4...12MHz XTAL
Oscillator
RC
Oscillator20..50kHz
H2.5
Sub Sec Cnt SSC
20-Bit Rel Reg SSR
1Hz SEC MIN HR
RTCC.SEL Mux
POL.PER Mux
wake-up
Delay
Counter
poll period
Poll Clk
10..19 1248163212481632124816243 7 10 14
CLK Mux
load
Edge/Level
Trigger Mode
& Enable
wake in
UPort
PPort
CMOS Threshold
CANx-RX 10x Port Wake Module
≥1
≥1
S
R
Q
wake out
IR_RTC
IR_WAPI
trigger
fSS
fPC fPP
1
Mux
S
R
Q
fS
fmfhfd
en
en
OSC.XM
OSC.RC
&
&
OSC.SRC
POL.OE
POL.ENA
&
&
WSC.RTC
WUS.RTC
WUS.WPx
POL.DEL
SMX.MUX
SMX.BYP
wake-up
fPP
Polling Module
WAID WAKE_RES
&
WSC.P
poll out
RTC out
SMX_out
WPMx.MOD
1/8 1/2
OSC.PRE
Poll Clk
RTC Module
WSC.AST
RTCCLK
VDD
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 61
for the sub second counter (fSS), either a clock of the quartz
oscillator divided by 8 or 16 or that of the module built-in RC
oscillator can be selected. The SSR has to be programmed
with a value that yields a one Hertz beat fS as output signal of
the SSC to clock the seconds. fS is the reload signal for SSC
as well as the input clock to the seconds counter (RTC.SEC).
An underrun of the down-counting SSC triggers the seconds
counter to count up. 60 seconds trigger a minutes up-count
(RTC.MIN), followed by the hours (RTC.HR).
All stages of the three up-counters can be selected to gener-
ate an interrupt or a wake-up signal.
As opposed to other internal clocks, the RTC can not be
stopped (during emulation) by ESTOPCLK.
8.1.2. Polling Module
The polling logic periodically activates the output signal
Wake Out which can be enabled by SW to drive port H2.5
(Poll Out). The rising edge of the Polling Period input (fPP)
defines the start and the Polling Clock (fPC). Together with
the delay counter, it defines the duration of the high time.
This can be used to cyclically flash a LED or drive external
circuitry.
The falling edge of the Wake Out signal is used to scan the
input levels of that Wake Ports (WPx) which are configured
for high or low level trigger mode. Those configured ports will
set the corresponding WPx flag in register WUS with the fall-
ing edge of the strobe signal.
Please refer to figure 8–6 for timing details about the Wake
Out and the strobe signals.
Due to the adjustment mechanism by the 20-bit reload regis-
ter, the polling period is not always constant. Depending on
the reload value, the polling period may vary between 0.5
and 1.5 nominal polling periods at the point of reloading.
The control unit is designed for a polling period to be set
equal to or greater than four times the polling delay.
8.1.3. Port Wake Module
There is a trigger mode logic (level or edge sensitive) and a
wake source flag for each Wake Port. The Wake Out input is
a signal from the Polling Module. The falling edge generates
a strobe pulse which is used to sample the level of the Wake
In input and sets the corresponding wake source flag if nec-
essary. Instead of the strobe signal, WSC.AST may be used,
e.g., if no RTC subsystem (with Polling Logic) is imple-
mented. The corresponding WPx flag in register WUS will be
forced to high as long as the programmed condition (high or
low level) is met at the Wake Port. Please see figure 8–2 for
details. The selected strobe signal source is valid for all
Wake Ports. Mixing of the strobe signal sources (polling and
alternative) is not possible.
The wake flags of all Wake Ports are located in the wake-up
source register WUS. The trigger events which can set the
wake flags can be configured in the wake-up pin mode regis-
ters WPM0 to 8 either in field MOD0 or MOD1. Please refer
to Table 8–9 for details about allocation of mode registers
and Wake Ports.
The output of each Wake Port is connected to an or gate,
whose output can generate a Wake Port interrupt as well as
a wake-up signal.
8.2. Registers
RC RC oscillator
r/w1: enable
r/w0: disable
XK External 32kHz XTAL (not available)
r/w1: enable
r/w0: disable
Write to zero for future compatibility.
XM 4 ... 12MHz XTAL
r/w1: always enabled
r/w0: disabled during power-saving mode
LD Load SRC and SSC
r: Always read as zero
w1: Immediately selects the oscillator source
according to SRC and loads SSC with SSR.
w0: No action
PRE 4 ... 12MHz XTAL / 8 or / 16
r/w1: 4 ... 12MHz XTAL / 16
r/w0: 4 ... 12MHz XTAL / 8
SRC Oscillator Source Select
r/w0: 4 ... 12MHZ XTAL (divided by 8 or 16)
r/w1: Don’t use, factory test only
r/w2: RC oscillator
r/w3: Ground
With OSC.LD set, writing to SRC selects a new oscillator
source immediately. When OSC.LD is not set, a new oscilla-
tor source does not get valid before the next reload of the
SSC. Consider that a read access returns the current source
select, not a possibly requested one by a write with OSC.LD
not set!
For typical settings, please refer to tables 8–1 to 8–3. The
values 0 and 1 are not allowed. To avoid programming val-
ues not expected, never write single bytes of the SSR on
their own, but always all 3 bytes without an interrupt by SSC
read. This is necessary, as for reading SSC the register
OSC Oscillator Source Register
76543210
Offs
r/w RC XK XM x LD PRE SRC 0
1 1 1 No HW reset Res
SSR Sub Second Reload Register
76543210
Offs
r/w xxxxxxxx3
r/w x x x x Bit 19 to 16 2
r/w Bit 15 to 8 1
r/w Bit 7 to 0 0
No HW reset Res
CDC16xxF-E ADVANCE INFORMATION
62 March 31, 2003; 6251-606-2AI Micronas
hardware (master/slave) uses the same buffer registers as
for writing SSR. Single bytes could still be filled with interme-
diate SSC values of a previous read. Writing SSR does not
load the SSC immediately. This will be done automatically
together with the next reload of the SSC. It can be forced
immediately by setting OSC.LD. A new value does not get
valid before the bus cycle has been written into its register.
Therefore please wait for one fIO cycle between write and
read access, e. g., to verify a value just programmed.
A read access to byte 0 of the SSC latches the bytes 1 and
2. This mechanism grants consistent read access to the
SSC.
Reading SEC latches MIN and HR. Writing HR simulta-
neously saves MIN and SEC in the corresponding counters.
This mechanism allows consistent read and write access.
Since read and write use the same latches, don’t mix these
access types. A new value gets valid not before the following
bus cycle has been written into its register. Therefore, please
wait for one fIO cycle between write and read access, e. g., to
verify a value just programmed.
HR Hours Counter
r/w0 to 23:
MIN Minutes Counter
r/w0 to 59:
SEC Seconds Counter
r/w0 to 59:
SEL Select RTC Output (Table 8–4)
RTC Real Time Clock
r1: RTC was trigger source
r0: No trigger
w1: Clear
w0: No modification
WPx Wake Port x
r1: Wake Port was trigger source
r0: No trigger
w1: Clear
w0: No modification
For proper interrupt generation some peculiarities in operat-
ing this register have to be considered. All set bits must be
cleared by writing back the whole pattern that was read
before. Always read and clear (write back) the whole regis-
ter, byte 0 first, which will become valid when writing byte 1,
even if only flags in byte 0 are in use. Every write access to
byte 1 will produce an interrupt, as long as WUS contains a
one.
MODy Trigger Mode (Table 8–5)
Trigger mode for Wake Port WPx+y. For assignment of Wake
Port and mode field please refer to table 8–9.
CLK Select Polling Clock (Table 8–7)
PER Select Polling Period (Table 8–6)
ENA Enable Polling Module
r/w1: enable
r/w0: disable
OE Enable Polling Output
r/w1: enable
r/w0: disable
DEL Select Polling Delay Time
r/w1 to 31: Delay time = DEL/fPC
r/w0: Delay time = 32/fPC
A write access to DEL immediately loads the 5-bit down
counter. The delay time defines the duration of the Wake Out
signal (Fig. 8–6).
SSC Sub Second Counter
76543210
Offs
RTC Real Time Counter
76543210
Offs
RTCC RTC Control Register
76543210
Offs
rxxxxxxxx3
rxxxx Bit 19 to 16 2
rBit 15 to 8 1
rBit 7 to 0 0
No HW reset Res
r/w xxxxxxxx3
r/w xxx HR 2
r/w xx MIN 1
r/w x x SEC 0
No HW reset Res
r/w x x x SEL 0
No HW reset Res
WUS Wake-Up Source Register
76543210
Offs
WPMx Wake Port x Mode Register
76543210
Offs
POL Polling Register
76543210
Offs
r/w RTCxxxxxWP9WP81
r/w WP7 WP6 WP5 WP4 WP3 WP2 WP1 WP0 0
No HW reset Res
r/w xMOD1xMOD00
No HW reset Res
r/w xCLKx PER 1
r/w ENA OE x DEL 0
0x00 Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 63
AST Alternative to Strobe
r/w1: Alternative input
r/w0: Wake out signal from Polling Logic
RTC RTC Wake-up Enable
r/w1: enable
r/w0: disable
Neither WUS.RTC nor the RTC interrupt source output are
affected.
P Port Wake-up Enable
r/w1: enable
r/w0: disable
Neither WUS.WPx nor the WAPI interrupt source output are
affected.
BYP Bypass SSC
r/w1: RTC is clocked by fSS (test)
r/w0: RTC is clocked by fS (1Hz)
MUX Signal Multiplexer (Table 8–8)
Defines a signal which is output as SMX_out.
WSC Wake Source Control
76543210
Offs
r/w xxxxxASTRTCP0
0x00 after UVDD power on Res
SMX Signal Multiplexer Register
76543210
Offs
r/w BYP x x x x MUX 0
0x00 Res
Table 8–1: SSR values for fs = 1Hz with XTAL and XTAL/8 (Maximum adjusted resolution)
XTAL [MHz] XTAL/8 [KHz ] Period [us] Reload Value Resolution [±ppm] Per Day [±ms]
4 500 2.00 500000 1.00 86.40
5 652 1.60 625000 0.80 69.12
6 750 1.33 750000 0.67 57.60
8 1000 1.00 1000000 0.50 43.20
Table 8–2: SSR values for fs = 1Hz with XTAL and XTAL/16 (Maximum adjusted resolution)
XTAL [MHz] XTAL/16 [KHz ] Period [us] Reload Value Resolution [±ppm] Per Day [±ms]
4 250 4.00 250000 2.00 172.80
5 313 3.20 312500 1.60 138.24
6 375 2.67 375000 1.33 115.20
8 500 2.00 500000 1.00 86.40
10 625 1.60 625000 0.80 69.12
12 750 1.33 750000 0.67 57.60
Table 8–3: SSR values for fs = 1Hz with RC (Maximum adjusted resolution)
RC [KHz] Period [us] Reload Value Resolution [±ppm] Per Day [±ms]
20 50.00 20000 25.00 2160
32,768 30.52 32768 15.26 1318
50 20.00 50000 10.00 864
CDC16xxF-E ADVANCE INFORMATION
64 March 31, 2003; 6251-606-2AI Micronas
Table 8–4: SEL Usage
RTCC.
SEL Tap# Activation
0GndNever
1
second counter taps
1 Every second
2 2 Every 2 seconds
3 4 Every 4 seconds
4 8 at second 8, 16, 24, 32, 40, 48, 56
5 16 at second 0, 16, 32, 48
6 32 at second 0, 32
7
minute counter taps
1 Every minute
8 2 Every 2 minutes
9 4 Every 4 minutes
10 8 at minute 8, 16, 24, 32, 40, 48, 56
11 16 at minute 0, 16, 32, 48
12 32 at minute 0, 32
13
hour counter taps
1 Every hour
14 2 Every 2 hours
15 4 Every 4 hours
16 8 Every 8 hours
17 16 at hour 0, 16
18 24 Every day
19..31 Don’t use
SEL = 4, 5, 6, 10, 11, 12 and 17 do not produce isoch-
ronous intervals.
Table 8–5: MOD Usage
WPMx.
MODy Trigger Modes
2 1 0
x 0 0 Disabled
001Rising edge
0 1 0 Falling edge
0 1 1 Rising and falling edge
1 0 1 High level 1)
110Low level 1)
1 1 1 Both levels (every strobe signal) 1)
1) not in WAKE mode
Table 8–6: PER Usage
POL.
PER TAP# fPP
0
Sub Second Counter
10 fSS/2TAP#
111
212
::
919
10
Second Counter
11Hz
11 2 0.5Hz
12 4 0.25Hz
13 8 at second 8, 16, 24, 32, 40,
48, 56
14 16 at second 0, 16, 32, 48
15 32 at second 0, 32
Isochronous intervals can only be achieved by PER =
10, 11 or 12.
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 65
Table 8–7: CLK Usage
POL.
CLK Sub Sec Tap# fPC
03 f
SS/2TAP#
17
210
314
Table 8–8: MUX Usage
SMX.
MUX Name
0V
DD
1f
SS SSC input (calibration)
2f
S (1Hz) SSC output (adjustment)
3 (wake-up) Test (Factory use only)
4(wake-up)
5 wake-up
6f
PP
7Poll Clk
Table 8–9: Wake Ports
Name Basic
Funct.
WPMx
WPMx.MODy
Special Functions
WP0 U3.4 0 0 T0-OUT / SEG3.4
WP1 U6.6 1 CAN0-RX / PINT1-
OUT / SEG6.6
WP2 U1.6 2 0 CAN1-RX / SEG1.6
WP3 U6.0 1 PINT0-IN/LCD-
SYNC_OUT/SEG6.0
WP4 U6.1 4 0 PINT1-IN/LCD-
CLK_OUT/SEG6.1
WP5 U6.2 1 PINT2-IN/T1-OUT/
SEG6.2
WP6 U5.6 6 0 PINT3-IN/PWM2/
SEG5.6
WP7 U4.0 1 CAN2-RX/SEG4.0
WP8 U4.4 8 0 UART0-RX/SEG4.4
WP9 H2.4 1 PWM0
CDC16xxF-E ADVANCE INFORMATION
66 March 31, 2003; 6251-606-2AI Micronas
8.3. Operation of Power-Saving Module
Before entering a Power-Saving Mode, the necessary wake-
up sources have to be configured carefully. The reset/wake-
up reason in register CSW2 and the wake-up source register
WUS have to be cleared. Please see section “CPU and
Clock System” for information on entering a power-saving
mode.
8.3.1. Configuration of Wake Sources
8.3.1.1. Port Wake Module
If an external event-driven wake-up is necessary, the Port
Wake Module has to be configured according to section 8.6.
The register WUS has to be cleared. Flag WSC.P has to be
set, enabling the Port Wake Module output signal to gener-
ate a wake-up signal by setting signal WAKE_RES.
If a Wake Port is to be operated in polling mode (level trig-
gered), configuration of RTC Module and Polling Module is
necessary as described in sections 8.4. and 8.5.
8.3.1.2. RTC Module
If a cyclic wake-up is necessary, the RTC Module has to be
configured according to section 8.4. Flag WUS.RTC has to
be cleared. Flag WSC.RTC has to be set, enabling the
WUS.RTC output signal to generate a wake-up signal by set-
ting signal WAKE_RES.
8.3.2. Configuration of Interrupts
During CPU active modes, the RTC Module and the Port
Wake Module can be operated as interrupt sources. The
interrupts have to be configured according to section “Inter-
rupt Controller (IR)”.
8.3.3. WAKE/IDLE
With setting SR3.WAID (mode = WAKE/IDLE) a core reset
signal is generated immediately.
A wake-up signal sets WAKE_RES to one, immediately pull-
ing pin RESETQ to low. This sets SR1.CPUFST and dis-
ables the WAKE_RES output of the Power-Saving Module.
When VDD and PH2 are detected as stable, the signal CLS is
cleared and the reset extension is started. After the reset
extension has finished, the pin RESETQ is released. The
Core reset signal gets inactive and CSW2.WKID is set.
The CPU starts execution at the reset vector address and
can read the reset/wake-up reason in registers CSW1 and
CSW2. The wake-up source can be read in register WUS
and should be cleared thereafter, otherwise Wake Port inter-
rupts are not possible.
8.3.4. Precautions
The SW has to guarantee the ability of wake-up. In the best
case, a stable initialization and configuration of the wake
logic is executed immediately before a power-saving mode is
activated.
The necessary Wake Ports have to be configured as inputs.
Enabling wake-up by pin only is dangerous. Inadvertently or
accidentally entering a power-saving mode with no wake
source enabled, e. g. by a software bug, or if a flag is modi-
fied by electrical overstress (EOS) in a power-saving mode,
a status may be reached which can only be terminated by a
manually generated reset with low level at pin RESETQ or
with a power on reset. Neither the watchdog nor the clock
supervision or any other internal reset source terminates a
power-saving mode.
Because neither the VBG generator nor the RESET compar-
ator are enabled during power-saving mode, proper CMOS
input levels (Vil=UVSS±0.3 V and Vih=UVDD±0.3 V) are
required on pin RESETQ during a wake-up reset. The exter-
nal circuitry must allow the device to establish WRVil on that
pin.
The specified power-saving mode current consumption val-
ues are only obtainable with CMOS input levels
(Vil=xVSS±0.3 V and Vih=xVDD±0.3 V) applied to all ports -
analog inputs via P-ports P0.1 to P0.9 excepted - not only
the Wake ports.
If an RTC-/Polling module is not used, it is advisable to
switch it off to reduce current consumption. Its outputs
should be disabled (see Section 8.4.3. on page 70).
8.3.5. Debug Register
To allow debugging during an active power-saving mode,
e.g., read or change the contents of the RAM, or read or
change the contents of I/O or processor registers, the sys-
tem clock must not be switched off as done normally when a
Power-Saving Mode is switched on. By setting DBG.DCS the
CPU keeps on running in a power-saving mode.
DCS DISABLE CPU STOP
r/w1: Disabled CPU stop (clock off) during
power-saving mode (debugging)
r/w0: Stop CPU during power-saving modes
(standard, no debugging)
DBG Debug Register
76543210
Offs
r/w xxxxxxxDCS0
xxxxxxx0POR
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 67
8.3.6. Timing
Fig. 8–3: Power-on Reset
RESETQ
MOSPOR
UVDD
WAKE_RES
RESCORE
RESPORT
COMPRES
SR3.WAID
POR
CLS
RESINTOUT
RESINT
RESIR
WKX
16 or 4096 XTAL cycles
1.5 ... 3.2V = Vbaspor
Bandgap and Reset Comparator active
2 ... 10us = tbaspor
PH2
ON
CDC16xxF-E ADVANCE INFORMATION
68 March 31, 2003; 6251-606-2AI Micronas
Fig. 8–4: Switching to WAKEIDLE with DBG.DCS inactive and Wake-up
RESETQ
MOSPOR
UVDD
WAKE_RES
RESCORE
RESPORT
COMPRES
SR3.WAID
POR
CLS
RESINTOUT
16 or 4096 XTAL cycles
RESIR
RESINT
CPU
wake-up logic
WKX
ON
DBG.DCS
PH2
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 69
Fig. 8–5: Switching to WAKEIDLE with DBG.DCS active and Wake-up
8.4. Operation of RTC Module
8.4.1. Reset
With the exception of the oscillators, which are enabled by
every reset (OSC.RC = OSC.XM = 1), most parts of the logic
will never be reset. Therefore, the oscillator not required has
to be switched off after every reset. Furthermore, the whole
logic has to be initialized after a power-on reset.
8.4.2. Oscillator Source and Sub Second Counter
The Sub Second Counter SSC generates a 1 Hz output sig-
nal at underflow (0x000000 to 0xFFFFFF). This signal loads
the SSC with the content of the SSR register and switches
the oscillator source select multiplexer to the desired oscilla-
tor according to the SRC field in the OSC register. This load
signal can be forced by writing a one to flag LD in register
OSC.
RESETQ
MOSPOR
UVDD
WAKE_RES
RESCORE
RESPORT
COMPRES
SR3.WAID
POR
CLS
RESINTOUT
16 or 4096 XTAL cycles
RESIR
RESINT
CPU
wake-up logic
WKX
ON
DBG.DCS
PH2
CDC16xxF-E ADVANCE INFORMATION
70 March 31, 2003; 6251-606-2AI Micronas
On three occasions it is necessary to change SSR and
OSC.SRC:
– Starting the SSC for the first time (after power-on).
As OSC.SRC is not reset by HW, an oscillator source
must be selected and enabled. The SSR has to be loaded
with the reload value necessary for a 1 Hz output fre-
quency. Writing the desired oscillator source in field SRC,
enabling it if necessary and setting flag LD in register
OSC immediately selects the new oscillator source and
loads SSR to SSC.
– Changing the SSR
Due to temperature or other dependencies of the oscilla-
tor it may be necessary to adjust the reload value in the
SSR register from time to time. This can be done within
the RTC interrupt service routine (RTC-ISR). Write the
new reload value to the SSR register. Make sure that this
will be completed before the next underflow of the SSC
which happens to each second and simultaneously loads
SSC with the new SSR value and switches the oscillator
source.
– Changing the oscillator source
Due to switching to a Power-Saving Mode it may be nec-
essary to change the oscillator source. This can be done
within the RTC-ISR. Select the desired oscillator source in
OSC.SRC and write the corresponding reload value to the
SSR register. Make sure that this will be completed before
the next underflow of the SSC which happens to each
second and simultaneously loads SSC with the new SSR
value and switches the oscillator source.
Precaution: Changing the oscillator source may cause a
fragmentary clock pulse. This may result in wrong SSC and/
or RTC values and unwanted interrupt or wake pulses.
Changing the oscillator source makes it necessary to:
1. Disable all output signals of the Power-Saving Module
(POL.OE = WSC.RTC = WSC.P = 0) and disable RTC and
WAPI interrupts.
2. Switch to the new oscillator source.
3. Initialize SSC, RTC and POL.
4. Clear WUS (WUS = 0xFFFF).
5. Enable the output signals again.
8.4.3. Disabling the RTC Module
This has to be done by selecting the RC oscillator
(OSC.SRC=2) and disabling this source (OSC.RC=0).
Selecting ground (OSC.SRC=3) disables the 32 kHz sub-
system too, but this should be avoided to be compatible with
future extensions.
8.4.4. Access to SSC and RTC
SSC and RTC are periodically altered by clock pulses. Even
if the 32 kHz subsystem is clocked by the 4 ... 12 MHz oscil-
lator, a CPU access to SSC and RTC can be corrupted by a
clock pulse. Because this situation can’t be avoided, the
SSC or the RTC register have to be read twice. If there is a
difference between the two accesses the read has to be
repeated. After a write to register RTC it has to be read and
compared to the desired value. If there is a difference, write,
read and compare have to be repeated. Since the RTC is
clocked not faster than in second distance, a read or write
access to register RTC can be done in the RTC-ISR. Such
an access is safe and guarantees a correct result as long as
the RTC-ISR is finished before the next clock alters the RTC.
8.4.5. RTC Output Multiplexer
All the taps of the second, minute and hour counters are con-
nected to a multiplexer (Table 8–4) and can be selected as
output by register RTCC.SEL. The output of this multiplexer
can generate an RTC interrupt as well as a wake-up signal.
8.4.6. RTC Interrupt
The IR has to be initialized as described above.
The RTC has its own interrupt vector, thus further investiga-
tion of the interrupt source is not necessary. After an inter-
rupt, the flag WUS.RTC is set. The register WUS.RTC is not
necessary for the RTC ISR, and does not have to be handled
by the RTC ISR.
Precaution: Please be aware that modifying RTC or RTCC
may result in addtional negative edges on RTC Out. If no
measures are taken, these edges will generate unwanted
interrupts.
A solution that would not affect the intended interrupts is
reading SSC and modifying RTC or RTCC only if sufficient
time is available to intercept the unwanted interrupt before
the next 1 Hz clock pulse occurs.
In this situation, the RTC interrupt may safely be temporarily
disabled.
8.4.7. Signal Multiplexer
Various internal signals can be switched to SMX_out. The
internal signal can be selected via register SMX.MUX. The
possible signals are shown in table 8–8. Only the signals fSS
and fS are of general interest. The remaining signals may be
used, but are intended for testing purposes.
The signal fSS is useful to measure the quartz frequency and
calculate the corresponding reload value for the sub-second
counter with external equipment.
The signal fS is useful for re-adjustment of the sub second
counter, with the help of, e.g., an XTAL-driven CAPCOM
counter, when driven by the internal RC oscillator.
The bypass switch (SMX.BYP) allows to bypass the SSC
and directly feed fSS into the second counter. This feature is
intended for testing purposes only. Applications normally
keep SMX.BYP cleared.
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 71
8.5. Operation of Polling Module
8.5.1. Reset
The whole logic is cleared by every kind of reset, even wake-
up from power-saving mode resets the Polling Module. This
means that the logic has to be initialized after every reset.
8.5.2. Initialization and Start
The Polling Module needs the RTC Module running,
because the Polling Clock fPC is derived from sub-second
counter taps, and the Polling Period fPP is derived from sub-
second counter taps or second counter taps. See section
8.4. for RTC Module initialization.
The enable input (POL.ENA) and the output (POL.OE) has
to be disabled.
The port H2.5 has to be configured as normal, out, low for
operation as polling output.
Select fPC (POL.CLK) and fPP (POL.PER). Enable input and
output (POL.ENA, POL.OE) and load the delay counter
reload register (POL.DEL) with a non-zero value.
Fig. 8–6: Polling Timing
8.5.3. Stop
Disable all inputs and outputs (POL.ENA=0, POL.OE=0).
8.5.4. Restart
Set POL.ENA and POL.OE to one. Initialize the delay
counter reload register (POL.DEL).
8.6. Operation of Port Wake Module
Fig. 8–7: Edge/Level Trigger Logic
fpp
wake out
Poll Clk
strobe
0123
poll period
poll delay
00 3
&
strobe
&
&
Delay &
&
&
≥1
WPMx.MOD 210
rising
falling
high
low
strobe
strobe
delay
delay
wake in
trigger
wake out
(from port)
(from Polling Module) ≥1
WSC.AST alt. strobe
wake in
wake in
CDC16xxF-E ADVANCE INFORMATION
72 March 31, 2003; 6251-606-2AI Micronas
8.6.1. Reset
Neither the wake-up source register (WUS) nor the Wake-
Port Mode registers (WPMx) are reset to a defined value by
any reset source.
8.6.2. Initialization
The corresponding ports must be configured as inputs.
For every Wake Port which is to generate a wake-up signal,
the trigger mode in the WPMx register has to be pro-
grammed. For every Wake Port which must not generate a
wake-up signal, the trigger mode in the WPMx register has to
be disabled. The source of the strobe signal has to be
selected with WSC.AST if a level triggered mode must be
used. The whole register WUS has to be cleared.
8.6.3. Operation
The Port Wake Module can be operated by polling. It can
generate an interrupt (IR_WAPI) or a wake-up signal to leave
the Power-Saving Mode.
To use a level-triggered mode of a Wake Port the RTC Mod-
ule and the Polling Module have to be configured to provide
the necessary Wake Out signal. The signal Wake Out is nec-
essary for a strobe pulse at the falling edge of Wake Out.
If no RTC-/Polling module is available or is not to be used, an
alternative strobe signal can be used by setting flag
WSC.AST to one. As long as WSC.AST is set and the pro-
grammed trigger level is applied to the corresponding pin,
the flag WUS.WPx is set and can’t be reset by the SW. If
more than one Wake Port is operated with the alternative
strobe signal, WPMx.MOD has to be disabled before
WUS.WPx can be cleared. Only clearing WSC.AST during
the WUS clearing procedure does not help in this case. Inter-
rupts of other Wake Ports can be lost if the high or low time is
too short. The selected strobe signal source is valid for all
Wake Ports. Mixing of the strobe signals sources (polling and
alternative) is not possible.
If only the edge-triggered mode is used, the RTC Module
and the Polling Module are not necessary for correct opera-
tion of the Port Wake Module.
8.6.4. Wake-up from Power-Saving Mode
Following the initialization described above, it is necessary to
enable the Port Wake Module as wake-up source by register
WSC flag P.
After wake-up the reason can be read in register WUS. It
should be cleared after reading, as otherwise neither wake-
up nor Wake Port interrupt via IR_WAPI is possible.
8.6.5. Wake Port Interrupt
Following the initialization described above, the IR has to be
initialized.
All Wake Ports are directed to an interrupt vector. After an
interrupt the source can be read in register WUS. It should
be cleared after reading, as otherwise no further Wake Port
interrupts via WAPI are possible.
Precautions
Parallel usage of a P-Port as analog and Wake Port input is
possible but not recommended. In this case the Schmitt Trig-
ger input circuit is enabled. This is the reason why input lev-
els other than AVSS and AVDD may cause quiescent currents
in the Schmitt Trigger circuit and thus lead to higher power
consumption.
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 73
9. Memory Patch Module
The Memory Patch Module allows the user to modify up to
ten hardwired ROM locations by external means. This func-
tion is useful if faulty parts of software or data are detected
after the ROM code has been cast into mask ROM.
Software loads addresses and the corrected code, e.g., from
external non-volatile memory into the respective registers of
the module. The module will then replace faulty code upon
address match.
Single ROM locations are directly replaced. Longer faulty
sequences may be repaired by introducing a jump to a new
subroutine in RAM (e.g. opcode JSR requires 3 consecutive
bytes to be patched). The RAM subroutine then may consist
of any number of instructions, ending with a return to the
next correct instruction in ROM. Thus it is possible to also
include complex software modules.
ICs which are derived from the Emulator IC may have less
patch cells. In this case the upper patch cells are not avail-
able.
Features
– patching of read data from up to 10 different ROM loca-
tions (24-bit physical address)
– automatic insertion of 1 CPU wait state for each patched
access
Fig. 9–1: Block Diagram
9.1. Principle of operation
9.1.1. General Remarks
The logic contains ten patch cells (see Fig. 9–1 on page 73),
each consisting of a 24-bit compare register (Patch Address
Register, PARn), a 24-bit address comparator, a Patch
Enable Register (PERn) bit and an 8-bit Patch Data Register
(PDR).
The current address information for a ROM access is fed to a
bank of ten patch cells. In case of a match in one patch cell,
and provided that the corresponding Patch Enable Register
bit is set, a wait cycle for CPU is included by pulling down the
RDY input of CPU for one cycle (see Fig. 9–2 on page 74). In
the meantime, the module’s logic disables the ROM data bus
drivers, and instead places the data information from the cor-
responding Patch Data Register on the data bus.
9.1.2. Initialization
After reset, as bit PER0.PMEN is reset to 0, all patch cell
registers are in write mode and patch operation is disabled.
To initialize a patch cell, first set the corresponding PSEL bit
in register PER0 or PER1 as a pointer. Then enter the 24-bit
address to registers PAR2 (high byte), PAR1 (middle byte)
and PAR0 (low byte) and the desired patch code to register
PDR.
If desired, repeat the above sequence for other patch cells.
Only set one PSEL pointer bit at a time in registers PER0
and PER1.
PA[23:16]PA[7:0]
ADB[23:0]
DB[7:0]
≥1
PA[15:8]
Patch Address Register
Patch Enable Register
Output Enable
Patch Data
Patch Cell 0
Patch Cells 1...9
Register
Enable
Write/Compare
Sequencer
PSEL9...0 PMEN
DBP[7:0]
RDY
ROMEN
RWQ ROMACC
PH2
PATOE
&
≡
CDC16xxF-E ADVANCE INFORMATION
74 March 31, 2003; 6251-606-2AI Micronas
9.1.3. Patch Operation
To activate a number of properly initialized patch cells for
ROM code patching, set all the corresponding PSEL bits in
registers PER1, then PER0, setting bit PER0.PMEN to 1.
The Memory Patch Module will immediately start comparing
the current address with the setting of the enabled patch
cells. In case of a match, the ROM data will be replaced by
the corresponding patch cell data register setting.
9.1.4. Reconfiguration
To reconfigure the Memory Patch Module, first set
PER0.PMEN to 0. The module will immediately terminate
patch operation.
Then proceed as described in “Initialization” on page 73.
Fig. 9–2: Timing
9.2. Registers
ADB[23:0]
RDY
PATOE
DB[7:0]
ROMEN
PH2
A2 A2 A3
A1 A3
D2D1 PD2 D3 PD3
PAR0 Patch Address Register 0
76543210
Note
PAR1 Patch Address Register 1
76543210
Note
wPA7 PA6 PA5 PA4 PA3 PA2 PA1 PA0
11111111Res
wPA15 PA14 PA13 PA12 PA11 PA10 PA9 PA8
11111111Res
PAR2 Patch Address Register 2
76543210
Note
PDR Patch Data Register
76543210
Note
wPA23 PA22 PA21 PA20 PA19 PA18 PA17 PA16
11111111Res
wPD7 PD6 PD5 PD4 PD3 PD2 PD1 PD0
00000000Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 75
PA23 to 0 Patch Address
Upon occurrence of this address the patch cell replaces
ROM data with data from PDR.
PD7 to 0 Patch Data
Data to replace false ROM data at certain address.
PSEL0 to 9 Select Patch Cell
w1: select cell for write or enable for patch
w0: disable patch cell
Before writing compare address or replace data of a patch
cell, only one cell must be selected. In compare mode one or
more patch cells can be selected.
PMEN Patch Mode Enable
w1: enable patch mode of all cells
w0: enable write mode of all cells
PER0 Patch Enable Register 0
76543210
Note
PER1 Patch Enable Register 1
76543210
Note
wPSEL6 PSEL5 PSEL4 PSEL3 PSEL2 PSEL1 PSEL0 PMEN
00000000Res
wx x x x x PSEL9 PSEL8 PSEL7
xxxxx000Res
CDC16xxF-E ADVANCE INFORMATION
76 March 31, 2003; 6251-606-2AI Micronas
10. Interrupt Controller (IR)
The Interrupt Controller has 16 input channels. Each input
has its own interrupt vector pointing to an interrupt service
routine (ISR). One of 15 priority levels can be assigned to
each input or the input can be disabled. The Interrupt Con-
troller is connected to the NMI input of the CPU. However,
despite the non-maskable interrupt input, it is possible to dis-
able all interrupt sources together in the Interrupt Controller.
Features
– 16 interrupt inputs.
– 16 interrupt vectors.
– 15 individual priority levels.
– Global/individual disable of interrupts.
– Single interrupt service mode.
10.1. Principle of Operation
10.1.1. General Remarks
Interrupt requests are served in the order of their pro-
grammed priority level. Interrupt requests of the same priority
level are served in descending order of interrupt input num-
ber.
Each of the 16 interrupt inputs clears a flag in the interrupt
pending register (IRRET and IRP), which can be read by the
user. A pending interrupt enables the output of the corre-
sponding priority register (IRPRI10 to IRPRIFE) which is
connected to a parallel priority decoder together with the
other priority registers. The decoder outputs the highest pri-
ority and its input number to a latch. The latched priority is
compared with the top entry of the priority stack. The top
entry of the priority stack contains the priority of the currently
served interrupt. Lower entries contain interrupts with lower
priority whose interrupt service routines were started but
interrupted by the higher priority interrupts above. If the
latched priority is lower or equal than the top of stack priority,
nothing happens. If the latched priority is higher than the top
of stack priority, a NMI is sent to the CPU and the latched pri-
ority is pushed on the stack.
The Interrupt Controller signals an interrupt to the CPU via
NMI input . After the current instruction is finished the CPU
starts an interrupt sequence. First it puts the program bank
register, the program counter high byte, the program counter
low byte and the program status register into the stack. Then
the CPU writes the vector address low byte (FFFA in 6502
mode, FFEA in 816 mode) to the bus. The Interrupt Control-
ler recognizes this address and stops the CPU by the RDY
signal. Now the Interrupt Controller writes the vector address
low and high byte of the corresponding interrupt number to
the bus and releases the CPU by releasing RDY. The CPU
now operates with the new vector of the interrupt service rou-
tine.
When the Interrupt Controller writes the new vector to the
address bus, the interrupt pending flag of this vector is set,
indicating that no interrupt is pending.
The software must pull the top entry from the priority stack at
the end of an interrupt service routine. This happens with the
write access to the interrupt return register IRRET. Then the
next entry (with lower priority) is visible at top of stack and is
compared with the priority latch.
The Interrupt Controller and related circuitry is clocked by the
CPU clock and participates in CPU FAST and SLOW mode.
10.1.2. Hardware settings
According to Section 10.3. some of the Interrupt Controller
inputs allow selection of a source by HW option (cf. Table
10–3). This configuration has to be done prior to operation.
Refer to “HW Options” for setting them.
10.1.3. Initialization
After reset, all internal registers are cleared but the Interrupt
Controller is active. When an interrupt request arrives, it will
be stored in the respective pending register IRP/IRRET. But
it will not trigger an interrupt as long as its interrupt priority
register IRPRIxy is set to zero.
The interrupt sources in peripheral modules have to be prop-
erly SW configurated prior to operation.
Before enabling individual inputs, make sure that no previ-
ously received signal on that input has cleared its pending
flag which may trigger the Interrupt Controller. Clear all
pending interrupts with the flag IRC.CLEAR to avoid such an
effect.
10.1.4. Operation
Activation of an interrupt input is done by writing a priority
value ranging from 1h to Fh to the respective IRPRIxy regis-
ter. Upon an interrupt request, pending or fresh, the Interrupt
Controller will immediately generate an interrupt.
During operation, changes in the priority register setting may
be made to obtain varying interrupt servicing strategies.
Flags IRC.DAINT, IRC.DINT and IRC.A1INT allow some
variation in the Interrupt Controller response behavior.
10.1.5. Inactivation
There are two possibilities to disable an interrupt within the
Interrupt Controller. Changing the priority of an interrupt input
to zero disables this interrupt locally. Interrupts are globally
disabled by writing a zero to flag IRC.DINT of register IRC.
During the evaluation period (see also Section 10.4.: Inter-
rupt Timing) it is not possible to suppress an interrupt by
changing priority.
A zero in the flag IRC.DINT of register IRC prevents the
Interrupt Controller from pulling the signal NMI low. However,
if this flag is set after the falling edge of NMI, the correspond-
ing interrupt cannot be cancelled.
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 77
Pending
Register Priority
Registers
IRPRI10
IRPRI32
IRPRIFE
Fig. 10–1:
Block Diagram
R
S
Q
Int-Input 1
Int-Input 2
Int-Input 3
Int-Input 4
Int-Input 15
Int-Input 16
4
4
4
4
4
4
16
Parallel
Priority
Decoder
A>B
A
B
Interrupt
Vector
Tabl e
16
Ctrl
Priority
Stack
15 x 4
RDY
A0...A15
NMI
clke
Priority
Latch
enable
FFFA
clke
push
IRRET write
pull
Clear Request
input #
prio
4
priority
4 4
A=B
A
B
clke
&clkeDMAE
Ph2
DMAE
or
FFEA
R
S
Q
R
S
Q
R
S
Q
R
S
Q
R
S
Q
PATCH
CDC16xxF-E ADVANCE INFORMATION
78 March 31, 2003; 6251-606-2AI Micronas
10.1.6. Precautions
10.1.6.1. Return from Interrupt
The write access to the IRRET must be performed just
before the RTI command at the end of the interrupt service
routine. After a write access to this location it is guaranteed
that the next command (should be RTI) will be processed
completely before a new interrupt request is signalled to the
CPU. If the RTI command does not immediately follow the
write to IRRET, an interrupt with the same priority may be
detected before the corresponding RTI is processed. A stack
underflow may occur because this may happen several
times.
An interrupt with a higher priority than the one actually
served, may interrupt between the write access to IRRET
and the belonging RTI command. Now an interrupt request
from the interrupted low priority interrupt may occur during
service of the high priority interrupt. This one will be served
after the RTI command of the high priority interrupt and
before the RTI command of the first interrupted low priority
interrupt. In this case, the return address and the PSW of the
same interrupt are stored twice on the stack. This may hap-
pen several times and can cause a problem if the stack size
is calculated without sufficient buffer or if the interrupt load is
too high, which means that there is no time for the back-
ground loop.
10.1.6.2. Disable Interrupt
If an opcode fetch of a disable interrupt instruction (DI) hap-
pens one clock cycle after the falling edge of NMI (see Sec-
tion 10.4.1. on page 83), it is possible, that an interrupt ser-
vice routine (ISR) is active, though the corresponding
interrupt is disabled. That is why after disabling an interrupt,
and before accessing critical data, at least one uncritical
instruction is necessary. This guarantees that the ISR is fin-
ished before critical data access and no further ISR can
interrupt it.
As it is now possible that an ISR (Interrupt Service Routine)
can lengthen the time between the disable interrupt instruc-
tion (DI) and the enable interrupt instruction (EI) indefinitely,
it is necessary that an ISR first saves registers and enable
interrupt flags and then enables interrupts. After interrupt
execution enable flags and registers must be restored. This
guarantees that other interrupts are not locked out during
interrupt execution.
Fig. 10–2: Interrupt Service Routine
10.2. Registers
RESET Reset
w1: No action.
w0: Momentary reset of the Interrupt Controller,
all internal registers are cleared.
The reset of the Interrupt Controller happens with writing
zero to this Flag. It is not necessary to write a one to finish
the reset.
The standard Interrupt Controller function is performed by
setting all flags to one. A hardware reset of the Interrupt Con-
troller is performed by setting RESET low and the other flags
to high.
DAINT Disable after interrupt
r1: Don’t disable after interrupt.
r0: Disable Interrupt Controller after interrupt.
w1: Cancel this feature.
w0: Disable Interrupt Controller after interrupt.
This is the enable flag for the flag A1INT function.
DINT Disable interrupt
r1: Interrupts are enabled.
r0: All interrupts are disabled.
w1: Enable interrupts according to priority setting.
w0: Disable all interrupts.
A1INT Allow one interrupt
w1: No action.
w0: Serve one interrupt.
This is a momentary signal. With DAINT = 0, only one inter-
rupt (with the highest priority) will be served.
The Flags DAINT and A1INT must be considered in com-
mon. They provide the possibility to serve interrupts one by
one, only when the main program has enough time (see
Table 10–1 on page 79).
Save Registers
Save Interrupt Enable Flags
Enable Interrupts
Execute Interrupt
Restore Interrupt Enable Flags
Restore Registers
Write to IRRET
RTI
IRC Interrupt Control Register
76543210
rx x x x DAINT DINT x x
wx x x RESET DAINT DINT A1INT CLEAR
x11xxRes
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 79
CLEAR Clear all requests
w1: No action.
w0: Momentarily clears all interrupt requests.
IRE Enable Interrupts
A write access to this memory location enables interrupts
according to priority setting (same effect as setting
IRC.DINT).
Enabling interrupts using this register take effect not before
the execution of the command, following the write access to
IRE.
IPF0 to 7 Interrupt Pending Flag of Input 0 to 7
r1: No interrupt is pending.
r0: Interrupt is pending.
w: Current request is finished.
For interrupt pending flags 8 to 15, please refer to the
description of register IRP.
A write access to this memory location signals to the Inter-
rupt Controller that the current request has been served.
PRIOn Priority of input number n
r: Priority of the corresponding interrupt input.
w: Priority of the corresponding interrupt input.
Priority zero prevents the Interrupt Controller from being trig-
gered, but the pending register is not affected. All incoming
requests are stored in the pending registers. Of two inputs
Table 10–1: Single Interrupt Service
DAINT A1INT Resulting function
0 1 Disable after current interrupt.
0 0 Serve one interrupt request.
1 x Normal interrupt mode.
IRE Interrupts Enable Register
76543210
IRRET Interrupt Return Register
76543210
IRPRI10 Interrupt Priority, Inputs 0 and 1
76543210
wA write access enables interrupts according to priority setting (same
effect as setting IRC.DINT)
00000000Res
rIPF7 IPF6 IPF5 IPF4 IPF3 IPF2 IPF1 IPF0
wA write access signals to the Interrupt Controller that the current
request has been served
00000000Res
r/w PRIO1 PRIO0
00000000Res
IRPRI32 Interrupt Priority, Inputs 2 and 3
76543210
IRPRI54 Interrupt Priority, Inputs 4 and 5
76543210
IRPRI76 Interrupt Priority, Inputs 6 and 7
76543210
IRPRI98 Interrupt Priority, Inputs 8 and 9
76543210
IRPRIBA Interrupt Priority, Inputs 10 and 11
76543210
IRPRIDC Interrupt Priority, Inputs 12 and 13
76543210
IRPRIFE Interrupt Priority, Inputs 14 and 15
76543210
r/w PRIO3 PRIO2
00000000Res
r/w PRIO5 PRIO4
00000000Res
r/w PRIO7 PRIO6
00000000Res
r/w PRIO9 PRIO8
00000000Res
r/w PRIO11 PRIO10
00000000Res
r/w PRIO13 PRIO12
00000000Res
r/w PRIO15 PRIO14
00000000Res
CDC16xxF-E ADVANCE INFORMATION
80 March 31, 2003; 6251-606-2AI Micronas
with the same PRIO setting, the input with the higher number
has priority (see Table 10–2 on page 80).
IPF8 to 15 Interrupt Pending Flag of Input 8 to 15
r1: No interrupt is pending.
r0: Interrupt is pending.
For interrupt pending flags 0 to 7, please refer to description
of register IRRET.
10.3. Interrupt Assignment
While most interrupt assignments are hard-wired, some are
configured via HW option (see Fig. 10–3 on page 82).
Table 10–2: PRIOn usage
PRIOn resulting function
0h Interrupt input is disabled
1h Interrupt input is enabled with lowest priority
::
Fh Interrupt input is enabled with highest priority
IRP Interrupt Pending Register
76543210
rIPF15 IPF14 IPF13 IPF12 IPF11 IPF10 IPF9 IPF8
00000000Res
Table 10–3: Interrupt assignment
Inter-
rupt
Input
Interrupt
Vector
Address
Interrupt
Source HW
Option
0 00FFE2-E3 INT-MUX 7 FFC1h
1 00FFE0-E1 INT-MUX 8 FFC1h
2 00FFDE-DF Timer 0
3 00FFDC-DD PINT0
4 00FFDA-DB PINT1
5 00FFD8-D9 INT-MUX 1 FFC0h
6 00FFD6-D7 INT-MUX 2 FFC0h
7 00FFD4-D5 INT-MUX 3 FFC0h
8 00FFD2-D3 CC2OR
9 00FFD0-D1 CAN 0
10 00FFCE-CF INT-MUX 9 FFC2h
11 00FFCC-CD UART 0 TX/RX
12 00FFCA-CB RESET/ALARM
13 00FFC8-C9 INT-MUX 4 FFC0h
14 00FFC6-C7 INT-MUX 5 FFC1h
15 00FFC4-C5 INT-MUX 6 FFC1h
Table 10–4: INT-MUX 1 = HW Option addr. FFC0H
bit 1 bit 0 selects
0 0 CC0 COMP
01Timer 2
1 0 CAN 2
11Timer 1
Table 10–5: INT-MUX 2 = HW Option addr. FFC0H
bit 3 bit 2 selects
00UART 2
01P06 COMP
1 0 SPI 0
11Timer 1
Table 10–6: INT-MUX 3 = HW Option addr. FFC0H
bit 5 bit 4 selects
00PINT3-IN
0 1 SPI 1
10UART 1
1 1 CC1 COMP
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 81
Table 10–7: INT-MUX 4 = HW Option addr. FFC0H
bit 7 bit 6 selects
0 0 CAN 2
0 1 SPI 0
10DMA
11PINT3-IN
Table 10–8: INT-MUX 5 = HW Option addr. FFC1H
bit 1 bit 0 selects
00Timer 2
01UART 1
1 0 SPI 1
11DMA
Table 10–9: INT-MUX 6= HW Option addr. FFC1H
bit 3 bit 2 selects
00Timer 2
01DIGITbus
10UART 2
11PINT2-IN
Table 10–10: INT-MUX 7 = HW Option addr. FFC1H
bit 5 bit 4 selects
0 0 CC0OR
01UART 1
10IR-RTC
1 1 IR-WAPI
Table 10–11: INT-MUX 8 = HW Option addr. FFC1H
bit 7 bit 6 selects
0 0 CC1OR
01PINT2-IN
10IR-RTC
1 1 IR-WAPI
Table 10–12: INT-MUX 9= HW Option addr. FFC2H
bit 1 bit 0 selects
0 0 CAN 1
01UART 1
10IR-RTC
1 1 IR-WAPI
CDC16xxF-E ADVANCE INFORMATION
82 March 31, 2003; 6251-606-2AI Micronas
10.3.1. Interrupt Multiplexer
Ten interrupt inputs are directly connected to the respective
module’s interrupt output. Six interrupt inputs, 4 to 6 and 13
to 15, allow source selection via multiplexers. The multiplex-
ers are configured by HW Option. Please refer to section HW
Options for details.
Fig. 10–3: Interrupt Assignment and Multiplexer
Mux
1
Mux
2
Mux
3
Mux
4
Mux
5
Mux
6
Interrupt
Controller
Programming
of multiplexer
in HW Option
field
16 Interrupt
inputs
Interrupt sources
of peripheral
modules
Assignment
is
implementation
specific
Assignment
is
implementation
specific
Mux
3
Mux
7
Mux
8
Mux
9
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 83
10.4. Interrupt Timing
10.4.1. Interrupt Response Time
The interrupt response time is calculated from the interrupt
event up to the first interrupt vector on the address bus.
After an interrupt event, the Interrupt Controller starts evalua-
tion with the first falling edge of Phi2. Evaluation requires one
clock cycle until the Interrupt Controller pulls the signal NMI
low.
After the falling edge of NMI the CPU finishes the actual
command. If the falling edge of NMI occurs one clock cycle
before an opcode fetch, the following command will be fin-
ished too. Then PC and status will be saved on stack before
the low byte of the interrupt vector is written to the address
bus.
CDC16xxF-E ADVANCE INFORMATION
84 March 31, 2003; 6251-606-2AI Micronas
Fig. 10–4:
Timing Diagram
Opcode ISRSecond ByteVector first ByteDMA AccessFFFA or FFEA 1)
Interrupt
Request
NMI
A0...A15
RDY
Clear
Request
Ph2
Interrupt
DMAE
Interrupts
Finish actual command and save status.
(Save status = 5 clocks (+ 1 clock if in Native Mode)).
enabled
FFFA in 6502 mode
FFEA in 816 mode
1)
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 85
10.5. Port Interrupt Module
Port interrupts are the interface of the Interrupt Controller to
the external world. Four U-Port pins are connected to the
module via their special input lines. Port interrupt 0 and 1 can
scale down the interrupt load by prescalers. Port interrupt 2
and 3 share the interrupt input with signals from other
sources. HW Option programmable multiplexers define
which signal is actually connected to the Interrupt Controller.
Fig. 10–5: Port Interrupts
The user can define the trigger mode for each port interrupt
by the interrupt port mode register.
The Port interrupt prescaler can be switched by the interrupt
port prescaler register. The pulse duty factor of the prescaler
output is 50%.
The Trigger Mode defines on which edge of the interrupt
source signal the Interrupt Controller is triggered. The trig-
gering of the Interrupt Controller is shown in figure 10–6 and
10–7 for port prescaler active (P1INT32 or P0INT4 = 1).
0
1
SI
U6.1
PINT1-IN
IRPP.P1INT32
1/32
0
1
SI
U6.0
PINT0-IN
1/4
U6.2
PINT2-IN
U5.6
PINT3-IN
IRPP.P0INT4
HW Option
HW Option
HW Option
U5.7
PINT0-OUT
Mux6
Mux4
Mux3
U6.6
PINT1-OUT
PINT0
INT-MUX 6
Trigger
Mode
SI
SI
SO
SO
IRPM0
0
1
U5.7
PINT3-IN
SI
HW Option
Interrupt
Source
PINT1
Interrupt
Source
Interrupt
Source
INT-MUX 4
Interrupt
Source
INT-MUX 3
Interrupt
Source
Table 10–13: Module-specific settings
Module
Name HW Options Initialization
Item Address Item Setting
PINT0 PINT0-IN U6.0 special in
PINT0-OUT U5.7 special out
PINT1 PINT1-IN U6.1 special in
PINT1-OUT U6.6 special out
PINT2 Interrupt multiplexer 6 FFC1h PINT2-IN U6.2 special in
PINT3 Input pin selection FFC2h PINT3-IN U5.6 or U5.7 special in
Interrupt multiplexer 4 FFC0h
Interrupt multiplexer 3 FFC0h
CDC16xxF-E ADVANCE INFORMATION
86 March 31, 2003; 6251-606-2AI Micronas
PITn Port interrupt trigger number n
This field defines the trigger behavior of the associated port
interrupt (Table 10–14).
P1INT32 Port 1 interrupt prescaler
w1: Indirect mode, 1:32 prescaler
w0: Direct mode, bypass prescaler
P0INT4 Port 0 interrupt prescaler
w1: Indirect mode, 1:4 prescaler
w0: Direct mode, bypass prescaler
Fig. 10–6: Interrupt Timing (1/4 Prescaler On)
Fig. 10–7: Interrupt Timing (1/32 Prescaler On)
IRPM0 Interrupt Port Mode Register 0
76543210
IRPP Interrupt Port Prescaler Register
76543210
wPit3 Pit2 PIT1 PIT0
00000000Res
wxxxxxxP1INT32P0INT4
00Res
Table 10–14: PITn usage
PITn Trigger Mode
0h Interrupt source is disabled
1h Rising edge
2h Falling edge
3h Rising and falling edges
123412341234
Port U6.0
Falling edge
Falling and rising
Independent
Interrupt
(low active)
Interrupt
(low active)
1/4 prescaler output of trigger mode
edge trigger mode
Rising edge
Interrupt
(low active)
32 1 2 15 16 31 32
Port U6.1
Falling edge trigger
Falling and rising
Independent
Interrupt
(low active)
Interrupt
(low active)
1/32
of trigger mode
edge trigger mode
prescaler output
17 18
Rising edge trigger
Interrupt
(low active)
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 87
11. Ports
Three kinds of ports exist. The analog input port, Port 0,
serves as input for the analog-to-digital converter. The uni-
versal ports, U1 to U7, serve as digital I/O and can be config-
ured as LCD drivers. The high current ports, H0 to H3, serve
as digital I/O and can be configured as stepper motor drivers.
11.1. Analog Input Port 0
The 9-bit-wide analog input port is called Port 0. Five of
these analog input pins can additionally be configured as
digital inputs. One pin is connected to a comparator, which
can be selected as interrupt source.
Features
– 9-bit analog input multiplexer.
– 5 bits additionally usable as digital input ports.
Fig. 11–1: Port 0 with Input Multiplexer and Undervoltage Alarm Comparator
The nine analog input lines are connected to a multiplexer.
The output of this multiplexer is connected to a 10-bit A/D
converter.
Port P0.6 is input of an undervoltage alarm comparator
which is described in the A/D converter section.
Five of the analog input pins (P0.1 to P0.5) may be used as
digital inputs if enabled by setting the flag P0DIN in the
Standby Register SR1. The digital value of the input pins can
be read in the Port 0 Data register P0D. These ports should
either be used as analog or digital inputs.
D1 to 5 Port 0 Digital Input 1 to 5
r1: High level at input pin.
r0: Low level at input pin.
P0.1
P0.2
P0.3
P0.4
P0.5
P0.6
P0.7
P0.8
P0.9
To A/D converter
SR1.P0DIN
P0D
12
3
4
5
6
789
01234567
rd xxx
To Alarm Comparator
P0D Port 0 Data Register
76543210
rx x D5 D4 D3 D2 D1 x
xxxxxxxxRes
CDC16xxF-E ADVANCE INFORMATION
88 March 31, 2003; 6251-606-2AI Micronas
11.2. Universal Ports U1 to U7
There are 52 Universal Port pins. The universal ports U1 to
U6 are 8 bits wide. The universal port U7 is 4 bits wide. Features
– SW selectable as digital I/O or LCD driver.
– LCD mode: 1:4 multiplex, 5 V supply.
– I/O mode: Tristate output, current limited.
– Individually programmable to deliver constant current
(slew rate).
– Schmitt hysteresis input buffer.
Fig. 11–2: Universal Port Circuit Diagram
Universal ports can be operated in different modes:
Table 11–1: Universal Ports Operating Modes
Modes Function
Port Mode Normal Input
and
Special Input
The SW uses the ports as digital input.
The port input is additionally connected to specific hardware modules.
Normal Output The SW uses the ports as latched digital tristate output.
Special Output The output signals of specific hardware modules are directly port output source.
LCD Mode The port bit serves as backplane/segment driver for an LC Display.
D
DATA
Q
D
N/S
Q
D
TRI
Q
0
1
Ux.y
DBy
Special In
Special Out
UxSEG.TRIy
UxSEG.N/Sy
UxD write
UxD read
x: Port number 1 to 7
y: Port pin number 0 to 7
slow/fast
From LCD module
UxM.PMODE
0
1
VDD
VSS
UVDD
UVSS
0
1
OR
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 89
The Universal Port bits can be switched to LCD or Port mode
in groups of two. The flag PMODE must be cleared to use
port pins in LCD mode. If PMODE is set, they serve as I/O
ports.
In both LCD and Port mode, the Port Slow mode may be
defined for each individual Universal port bit. It reduces the
current drive capability of the output stage.
After reset, all Universal Ports are in port, tristate condition.
11.2.1. Port Mode
Each port bit can be individually configured to several port
modes. The output driver of each pin can be disabled
(tristated) by setting the flag TRI. Set the flag N/S to select
the source of the output value.
For Port mode, the UxM registers have to be set for mode
selection, the UxSEG registers have to be properly set for
individual port bit configuration and the UxD registers serve
as I/O registers.
Table 11–2 shows configurations of flags if the corresponding
flag PMODE is true (Port mode)
In Port Mode, Special Input mode is always active. This
allows manipulating the input signal to the special hardware
through Normal Output operations by software.
As the Special Output mode allows reading the pin levels,
the output state of the special hardware may be read by the
CPU.
11.2.2. LCD Mode
For LCD Mode, the UxM registers have to be cleared for
mode selection and the UxSEG registers serve as segment
output data registers.
The output sequence timing on backplane and segment out-
put ports in LCD Mode is controlled by the LCD module.
Please refer to section LCD Module for information about
operation of this module.
As generation of the backplane port output sequence is fully
done by the LCD module, no segment setting is necessary
for these ports.
Port bits in LCD mode will always read as logical 0 as the
port input buffer is turned off.
11.2.3. Port Slow Mode
The output drivers of all port pins together can be configured
to operate in Fast or Slow mode by the Port Slow mode flag
PSLW in register SR1.
All U-Ports exhibit two operating regions in the DC output
characteristic (see Fig. 11–3). Near zero output voltage the
internal driver transistors operate non-limited, to offer a low
on-resistance. With larger output voltages, however, a limit is
imposed on the output current. This measure helps to fight
supply current transients and related EMI noise during port
switching.
Fig. 11–3: Typical U-Port pull-down DC output characteris-
tic (pull-up characteristic is complementary).
In this limited operating region, Port Fast mode and Port
Slow mode select two different current limits Ishf and Ishs.
Port Slow mode reduces the output current to a value where
the output may even be shorted continuously to either supply
rail. Thus, wired or-configurations can be put into practice.
The external load resistance should be greater than
5 kOhms in Port Slow mode. Please note that in Port Normal
Output mode, a READ of register UxD returns the UxD regis-
ter setting, not the pin levels.
Table 11–2: Port Mode Register Settings
Mode N/S TRI DFunction
Normal
Input x 1 x READ of register
UxD returns port
pin input levels to
data bus.
Normal
Output 0 0 Data WRITE to register
UxD changes level
of port pin output
drivers.
READ of register
UxD returns the
UxD register setting
to the data bus.
Special
Input x x x Port pin input level
is presented to spe-
cial hardware.
Special
Output 1 0 x Special hardware
drives port pin.
READ of register
UxD returns port
pin input levels to
data bus.
1V 2V 3V 4V 5V Vol
Port Fast Mode
Port Slow Mode
Non- Limited
region
Io
Ishs
Ishf
0
limited
region
CDC16xxF-E ADVANCE INFORMATION
90 March 31, 2003; 6251-606-2AI Micronas
With PSLW = 0 all ports are in Fast Mode. Only port bits that
are enabled via HW option will switch to Port Slow Mode with
PSLW = 1. A port pin is in Fast Mode all the time if a zero is
programmed to the appropriate HW option bit. It is not possi-
ble to switch this pin to Slow Mode. If a one is programmed,
the pin can be toggled between Fast and Slow Mode by the
flag PSLW. Please refer to section HW Options for informa-
tion on port/option assignment.
It is recommended to place all LCD ports in the Port Slow
mode.
11.3. Universal Port Registers
Universal Port Data Registers UxD contain input/output data
of the corresponding port.
The “x” in UxD means the number of the port. Thus UxD
stands for U1D to U7D. Remember that port U7 is only 4 bits
wide. For this reason not all of the described registers and
flags are available for U7.
D0 to 7 Universal Port Data Input/Output
r: Read pin level resp. data latch.
w: Write data to data latch.
To use a port pin as software output, the appropriate driver
must be activated and the N/S flag must be programmed to
Normal Mode.
Universal Port Segment registers UxSEG together with the
Universal Port Mode registers UxM are used to configure the
appropriate ports in Port Mode. In LCD Mode the registers
UxSEG contain the data for two segments each. For
instance, register U2SEG76 and U2M76 control port U2, bits
7 and 6.
Registers U1SEG10, U1SEG32, U1M10, U1M32 and
U3SEG32 differ from the corresponding control registers of
other ports in LCD mode. Please refer to the special descrip-
tion of these registers at the end of this section.
UxD Universal Port x Data Register
76543210
UxSEG10 Universal Port x Segment Register of
Ux.1 and Ux.0
76543210
UxM10 Universal Port x Mode Register of
Ux.1 and Ux.0
76543210
r/w D7 D6 D5 D4 D3 D2 D1 D0
00000000Res
wx N/S1 TRI1 x x N/S0 TRI0 x Port
wSEG1.3 SEG1.2 SEG1.1 SEG1.0 SEG0.3 SEG0.2 SEG0.1 SEG0.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
UxSEG32 Universal Port x Segment Register of
Ux.3 and Ux.2
76543210
UxM32 Universal Port x Mode Register of
Ux.3 and Ux.2
76543210
UxSEG54 Universal Port x Segment Register of
Ux.5 and Ux.4
76543210
UxM54 Universal Port x Mode Register of
Ux.5 and Ux.4
76543210
UxSEG76 Universal Port x Segment Register of
Ux.7 and Ux.6
76543210
wxN/S3TRI3x xN/S2TRI2xPort
wSEG3.3 SEG3.2 SEG3.1 SEG3.0 SEG2.3 SEG2.2 SEG2.1 SEG2.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S5TRI5x xN/S4TRI4xPort
wSEG5.3 SEG5.2 SEG5.1 SEG5.0 SEG4.3 SEG4.2 SEG4.1 SEG4.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S7TRI7x xN/S6TRI6xPort
wSEG7.3 SEG7.2 SEG7.1 SEG7.0 SEG6.3 SEG6.2 SEG6.1 SEG6.0 LCD
00100010Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 91
N/S0 to 7 Normal/Special Mode Flag 0 to 7
w1: Special Mode. Special hardware drives pin.
w0: Normal Mode. Data latch drives pin.
The corresponding Port Mode flag PMODE must be true to
make this flag valid. The N/S flag defines from which source
the pin is driven if TRI is false.
TRI0 to 7 Tristate Flag 0 to 7
w1: Output driver is tristate
w0: Output driver is active
The corresponding Port Mode flag PMODE must be true to
make this flag valid.
SEGh.0 to .3 LCD Segment Driver High, Bits 0 to 3
SEGl.0 to .3 LCD Segment Driver Low, Bits 0 to 3
In LCD Mode each port pin is controlled by a field of four
LCD Segment bits. Each segment register UxSEG contains
two fields of segment data, each four bit wide. “h” stands for
the high, “l” for the low pin number. For instance,
U2SEG76.SEG7.3 to U2SEG76.SEG7.0 control LCD seg-
ments driven by port U2.7.
Please refer to Pin Assignment and Description for segment/
pin number assignment. Information about the usage of the
LCD Segment field will be found at the functional description
of the LCD Module.
PMODE Port Mode Flag
Select the mode of the corresponding port pins.
w1: The two port pins are in Port mode.
w0: The two port pins are in LCD mode.
11.3.1. Special Register Layout of Universal Port 1
Universal Port 1 pins U1.0 to U1.3 provide backplane signals
in LCD Mode. To operate any ports as LCD segment driver it
is necessary to switch these ports to LCD mode. All four pins
will be switched together (not in groups by two) to LCD Mode
by clearing the flag PMODE in register U1M30.
Thus, U1M30 replaces U1M10 and U1M32.
As backplane ports U1.0 to U1.3 require no segment data
setting, SEG bits are not available in register U1SEG10 and
U1SEG32.
LCDSLV LCD Module is Slave
Select the mode of the LCD module.
w1: LCD module is slave.
w0: LCD module is master.
A write access to this memory location simultaneously loads
all segment information of all universal ports in LCD mode
into the display. The flag LCDSLV is available only in LCD
mode. This flag is not available in other universal port regis-
ters.
11.3.2. Special Register Layout of Universal Port
3.2
U3.2, in Port Special Output mode, provides the DIGITbus
connection. For this purpose it can be switched into a Double
Pull-down Mode (DPM) by setting U3SEG32.DPM2, where
– the short circuit current Ishs is doubled (with Port Slow
Mode enabled for U3.2 by HW Option, and SR1.PSLW set
to 1)
– the output configuration is pull-down, not the standard
push-pull.
By these means, this port may be configured to operate as
connection to the wired-or, single-wire DIGITbus with exter-
nal pull-up resistor.
DPM Double Pull-Down Mode
w1: Output driver is pull-down,
Ishs (Port Slow mode) doubled.
w0: Standard.
UxM76 Universal Port x Mode Register of
Ux.7 and Ux.6
76543210
U1M30 Universal Port 1 Mode Register of
U1.3 to U1.0
76543210
wxxxxxxxPMODE
00000001Res
wxxxxxxxPMODE
00000001Res
U1SEG10 Universal Port 1Segment Register of
U1.1 and U1.0
76543210
U1SEG32 Universal Port 1Segment Register of
U1.3 and U1.2
76543210
U3SEG32 Universal Port 3 Segment Register of
U3.3 and U3.2
76543210
wxN/S1TRI1x xN/S0TRI0xPort
wLCDSLV LCD
00100010Res
wxN/S3TRI3x xN/S2TRI2xPort
00100010Res
wx N/S3 TRI3 x DPM2 N/S2 TRI2 x Port
wSEG3.3 SEG3.2 SEG3.1 SEG3.0 SEG2.3 SEG2.2 SEG2.1 SEG2.0 LCD
00100010Res
CDC16xxF-E ADVANCE INFORMATION
92 March 31, 2003; 6251-606-2AI Micronas
11.4. High Current Ports H0.0 to H3.5
The High Current Ports 0, 1, 2 and 3 are used to drive coils
of stepper motors. Each port is 6 bits wide. They are similar
to universal ports, but as the name says, they can drive
higher currents. H-Ports can be operated via software like
Universal Ports (Port Mode). Their Special Out connections
are connected with the stepper motor module, or with the
PWM output.
Features
– Tristate output.
– ±30 mA output current.
– Schmitt hysteresis input buffers.
– Reduced slew rate of current and voltage for driving
resistive, capacitive or inductive loads.
Fig. 11–4: High Current Port Circuit Diagram
The H-Ports H0 and H1 are supplied by the power supply
pins HVDD1 and HVSS1. The H-Ports H2 and H3 are sup-
plied by the power supply pins HVDD2 and HVSS2.
Flag TRI is used to switch the output driver on and off, and
the flag N/S defines the mode of each port pin. Table 11–3
shows the various selectable modes.
The Special Outputs of high current ports are connected to
the Stepper Motor module or some PWMs. Please refer to
Pin Assignment and Description for information on assign-
ment of PWMs to H-Port pins.
D
DATA
Q
D
N/S
Q
D
TRI
Q
0
1
Hx.y
DBy
Special In
Special Out
HxTRI.TRIy
HxNS.N/Sy
HxD write
HxD read
x: Port number 0 to 3
y: Port pin number 0 to 5
High
HVDD
HVSS
VDD
VSS
Table 11–3: Register Settings
Mode N/S TRI DFunction
Normal
Input x 1 x READ of register
HxD returns port
pin input levels to
data bus.
Normal
Output 0 0 Data WRITE to HxD
changes level of
output, READ of
HxD reads pin level
Special
Input x x x Port pin input level
is presented to spe-
cial hardware.
Special
Output 1 0 x Special hardware
drives port pin.
READ of register
HxD returns port
pin input levels to
data bus.
Table 11–3: Register Settings
Mode N/S TRI DFunction
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 93
Twenty of the twenty-four high-current ports are connected to
the stepper motor control module. Two high-current ports,
together with a coil, form an H-Bridge. Two H-Bridges are
necessary to operate a stepper motor. The twenty stepper
motor outputs can thus drive five stepper motors.
The N-channel and the P-channel transistor of the output
driver are controlled separately. Thus crossover currents are
eliminated.
The output levels of the ports during and after reset are low,
to avoid floating coils.
11.5. High Current Port Registers
High Current Port Data registers are used to input/output dig-
ital values. The “x” means the number of the port. Thus HxD
stands for H0D, H1D, H2D or H3D.
D0 to 5 H-Port Data Input/Output
r: Read pin level .
w: Write data to data latch.
To use a port pin as output, the appropriate driver must be
activated and N/S mode flag must be set to normal mode.
The High Current Tristate and Normal/Special registers
(HxTRI and HxN/S) are used to configure the corresponding
port.
N/S0 to 5 Normal/Special Mode Flag 0 to 5
w1: Special Mode. Special hardware drives pin.
w0: Normal Mode. Data latch drives pin.
The N/S flag defines from which source the pin is driven if
TRI is false.
TRI0 to 5 Tristate Flag 0 to 5
w1: Output driver is tristate
w0: Output driver is active
HxD High Current Port x Data Register
76543210
HxTRI High Current Port x Tristate Register
76543210
HxNS High Current Port x Normal/Special
Register
76543210
r/w x x D5 D4 D3 D2 D1 D0
00000000Res
wx x TRI5 TRI4 TRI3 TRI2 TRI1 TRI0
00000000Res
wx x N/S5 N/S4 N/S3 N/S2 N/S1 N/S0
00000000Res
CDC16xxF-E ADVANCE INFORMATION
94 March 31, 2003; 6251-606-2AI Micronas
12. A/D Converter (ADC)
This 10-bit analog-to-digital converter allows the conversion
of an analog voltage in the range of 0 to URef, into a digital
value. A multiplexer connects the ADC to one of 9 analog
input ports. A sample and hold circuit holds the analog volt-
age during conversion. The duration of the sampling time is
programmable. The A/D conversion is done by a charge bal-
ance A/D converter using successive approximation.
Features
– A/D converter with 10-bit resolution.
– Successive approximation, charge balance type.
– Input multiplexer with 9 analog channels.
– Sample and hold circuit.
– 4/8/16/32 µs conversion selectable for optimum through-
put/accuracy balance.
– 2.5 V to 5 V external reference input.
– Zero standby current, 300 µA active current.
Fig. 12–1: Block Diagram
P0.1
P0.2
P0.3
P0.4
P0.5
P0.6
P0.7
P0.8
P0.9
12
3
4
5
6
789
A
D
S&H
xxxx
TSAMP CHANNEL
AD0
AD1
01234567 01234567
987 432 10
AVDD
0
1
VREF
AVSS
4
SR1.ADC
EOC
CMPO CMPO
2
10
r
w
r
56
interrupt
source
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 95
12.1. Operation
12.1.1. ADC
After reset, the module is off (zero standby current). The
module is enabled by the flag SR1.ADC. The user must be
sure that the flag End of Conversion (EOC) in register AD0 is
true, before he starts to operate the module.
A write access to register AD0 indicating sample time and
channel number starts the conversion. The flag EOC signals
the end of conversion. The 10-bit result is stored in the regis-
ters AD1 (8 MSB) and AD0. The conversion rate depends on
the software, the oscillator frequency, and the programmed
sample time.
The module may be operated in CPU FAST or SLOW mode.
12.1.1.1. Conversion Law
The result of A/D conversion is described by the following
formula:
Fig. 12–2: Characteristic Curve
The voltage on the reference input pin VREF can be set to
any level in the range from 2.56 V to AVDD.
12.1.1.2. Measurement Errors
The result of the conversion mirrors the voltage potential of
the sampling capacitance (typically 15 pF) at the end of the
sampling time. This capacitance has to be charged by the
source through the source impedance within the sampling
time period. To avoid measurement errors, system design
has to make sure that at the end of the sampling period, the
potential error on the sampling capacitance is less than ±0.1
LSB.
Measurement errors can occur, when the voltage of high
impedance sources has to be measured:
– To reduce these errors, the sampling time may be
increased by programming the field TSAMP in register
AD1.
– In cases where high impedance sources are only rarely
sampled, a 100 nF capacitor from the input to AVSS is a
sufficient measure to ensure that the potential on the sam-
pling capacitance reaches the full source potential, even
with the shortest sampling time.
– In some high impedance applications a charge pumping
effect may noticeably influence the measurement result:
Charge pumping from a high potential to a low potential
source will occur when such two sources are measured
alternatingly. It results in a DC current that appears as
flowing from the high potential source through the IC into
the low potential source. This current is explained by the
fact that during the sampling periods the high potential
source always up-charges the sampling capacitance
while the low potential source always discharges it.
12.1.2. P0.6 Comparator
In addition to the A/D converter the module contains a com-
parator. The level on port P0.6 is compared to AVDD/2. The
state of the comparator output can be read at flag CMPO in
register AD0.
The interrupt source output of this module is routed to the
Interrupt Controller logic. But this does not necessarily select
it as input to the Interrupt Controller. Check section “Interrupt
Controller” for the actually selectable sources and how to
select them.
The CMPO interrupt source is gated with an internal clock.
This is the reason why interrupts are generated as long as
the level at P0.6 is lower than the internal reference.
DV INT UIn
1LSB
--------------
=1LSB
URef
1024
------------=
where
DV = Digital Value; INT = Integer part of the result
3FF
3FE
3FD
02
01
00 123 1021 1023
DV
UIn [LSB]
03
CDC16xxF-E ADVANCE INFORMATION
96 March 31, 2003; 6251-606-2AI Micronas
12.2. Registers
A write access to register AD0 starts the A/D conversion of
the written channel number and sampling duration. The flag
EOC signals the end of conversion. The result is stored in
register AD1 (bit 9 to 2) and in register AD0 (bit 1 and 0).
EOC End of Conversion
r1: End of conversion
r0: Busy
EOC is reset by a write access to the register AD0. EOC
must be true before starting the first conversion after
enabling the module by setting SR1.ADC.
CMPO Comparator Output
r1: P0.6 is lower than reference.
r0: P0.6 is higher than reference.
Zero means that the voltage at P0.6 is higher than the com-
parator reference voltage.
TSAMP Sampling Time
TSAMP adjusts the sample time and the conversion time.
The total conversion time is 20 clock cycles longer than the
sample time.
Sampling starts one clock cycle after completion of the write
access to AD0.
CHANNEL Channel of Input Multiplexer
CHANNEL selects from which pin of port P0 the conversion
is done. The MSB of CHANNEL is bit 3. No port pin is con-
nected to the ADC, if values are selected which are not rec-
ommended by table 12–2. In this case the result is not
defined because the input of the A/D converter is open. After
reset, CHANNEL is set to zero. No channel is selected in this
case.
AN 9 to 0 Analog Value Bit 9 to 0
The 10-bit analog value is in the range of 0 to 1023. The 8
MSB can be read from register AD1. The two LSB can be
read from register AD0. The result is available until a new
conversion is started.
AD0 ADC Register 0
76543210
AD1 ADC Register 1
76543210
Table 12–1: Sampling Time Adjustment
TSAMP tSample tConversion
0H 20 TOSC 40 TOSC
1H 60 TOSC 80 TOSC
2H 140 TOSC 160 TOSC
3H 300 TOSC 320 TOSC
rEOC CMPO x x x x AN1 AN0
wTSAMP CHANNEL
00xx0000Res
rAN9 AN8 AN7 AN6 AN5 AN4 AN3 AN2
Table 12–2: ADC Input Multiplexer
CHANNEL Port Pin
1H P0.1
2H P0.2
3H P0.3
4H P0.4
5H P0.5
6H P0.6
7H P0.7
8H P0.8
9H P0.9
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 97
13. Timers (TIMER)
Three general-purpose timers are implemented. T0 is a 16-
bit timer, T1 and T2 are 8-bit timers.
13.1. Timer T0
Timer T0 is a 16-bit auto-reload down counter. It serves to
deliver a timing reference signal to the IRC, to output a fre-
quency signal or to produce time stamps.
Features
– 16-bit auto-reload counter
– Time value readable
– Interrupt source output
– Frequency output
Fig. 13–1: Timer T0 Block Diagram
13.1.1. Principle of Operation
13.1.1.1. General Remarks
The timer’s 16-bit down-counter is clocked by the input clock
and counts down. Falling below zero, it generates an output
pulse (underflow) to get reloaded with the value in its reload
register which is counted down subsequently.
T0 is not affected by CPU SLOW mode.
13.1.1.2. Operation
The clock input frequency can be set via HW option (see
Table 13–1 on page 98).
Prior to entering active mode, the U-Ports assigned to func-
tion as T0-OUT outputs have to be properly SW initialized
(Table 13–1). The ports have to be configured Special Out.
Refer to “Ports” for details.
T0 is always active (no standby mode). After reset, the timer
starts counting with reload value FFFFh generating a maxi-
mum period output signal.
A new time value is loaded by writing to the 16-bit register
TIM0, high byte first. Upon writing the low byte, the reload
register is set to the new 16-bit value, the counter is reset,
and immediately starts down-counting with the new value.
Falling below zero, the counter generates a reload signal,
which can be used to trigger an interrupt. The same signal is
connected to a divide-by-two scaler to generate the output
signal T0-OUT with a pulse duty factor of 50%.
The interrupt source output of this module is routed to the
Interrupt Controller logic. But this does not necessarily select
it as input to the Interrupt Controller. Check section “Interrupt
Controller” for the actually selectable sources and how to
select them.
The state of the down-counter is readable by reading the 16-
bit register TIM0, low byte first. Upon reading the low byte,
the high byte is saved to a temporary latch, which is then
accessed during the subsequent high byte read. Thus, for
time stamp applications, read consistency between low and
high byte is guaranteed.
13.1.1.3. Precautions
16-bit CPU commands do not generally keep to a certain
order in addressing high and low bytes of a register. Make
sure that the command used performs reading a 16-bit value
low byte first and writing high byte first.
In case of uncertainty use 8-bit commands.
16
TIM0w
T0
Interrupt
Source
HW Option
T0-OUT
underflow
12
tclk
clk ≤1/2 tclk
&
Reload-Reg.
16 bit Auto-reload
Down counter
clk res
r
CDC16xxF-E ADVANCE INFORMATION
98 March 31, 2003; 6251-606-2AI Micronas
13.1.2. Registers
TIM0 has to be read low byte first and written high byte first.
Table 13–1: Module specific settings
Module
Name HW Options Initialization Enable Bit
Item Address Item Setting
T0 Input clock FFA0h T0-OUT output U3.4 special out
TIM0L T0 low byte
76543210
TIM0H T0 high byte
76543210
Table 13–2: Reload Register Programming
Reload
value Output interrupt
source frequency is
divided by
Output T0-OUT is
divided by
0000h 1 2
0001h 2 4
0002h 3 6
:: :
FFFFh 65536 131072
rRead low byte of down-counter and latch high byte
wWrite low byte of reload value and reload down-counter
11111111Res
rLatched high byte of down-counter
wHigh byte of reload value
11111111Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 99
13.2. Timer T1 and T2
Timer T1 and T2 are 8-bit auto-reload down counters. They
serve to deliver timing reference signals to the IR or to output
frequency signals.
Table 13–3 describes implementation-specific HW Option
addresses and enable flags of T1 and T2.
Features
– 8-bit auto-reload counter
– Interrupt source output
– Frequency output
Fig. 13–2: Timer T1 and T2 Block Diagram
13.2.1. Principle of Operation
13.2.1.1. General Remarks
The timer’s 8-bit down-counter is clocked by the input clock
and counts down. Falling below zero, it generates an output
pulse (underflow) to get reloaded with the value in its reload
register which is counted down subsequently.
Tx is not affected by CPU SLOW mode.
13.2.1.2. Operation
The clock input frequencies can be set via HW options (see
Table 13–3 on page 99). After reset, the 8-bit timer is in
standby (inactive) mode.
Prior to entering active mode, proper SW initialization of the
U-Ports assigned to function as Tx-OUT outputs has to be
made (Table 13–3). The ports have to be configured Special
Out. Refer to “Ports” for details.
To initialize a timer, reload register TIMx has to be set to the
desired time value, still in standby mode.
To enter active mode, set the corresponding enable bit in the
standby registers (see Table 13–3 on page 99). The timer
will immediately start counting down from the time value
present in register TIMx.
During active mode, a new time value is loaded by simply
writing to register TIMx. Upon writing, the counter is reset,
and immediately starts counting down from the new time
value.
Falling below zero, the counter generates a reload signal,
which can be used to trigger an interrupt. The same signal is
connected to a divide-by-two scaler to generate the output
signal Tx-OUT with a pulse duty factor of 50%.
The interrupt source output of this module can be, but need
not be, connected to the interrupt controller directly or via
multiplexer. This is a HW option which is done by the factory
only. Please refer to section Interrupt Controller.
Returning Tx to standby mode by resetting its respective
enable bit will halt its counter and will set its outputs LOW.
The register TIMx remains unchanged.
The state of the down-counter is not readable.
13.2.2. Registers
8
TIMx
w
Tx
Interrupt
Source
HW Option
Tx-OUT
enable
underflow
12
tclk
clk ≤1/2 tclk
&
Reload-Reg.
8 bit Auto-reload
Down counter
clk &
res
Table 13–3: Module-specific settings
Module
Name HW Options Initialization Enable Bit
Item Address Item Setting
T1 Input clock FFA7h T1-OUT output U6.2 or U6.5 special out SR1.TIM1
T2 Input clock FFA8h T2-OUT output U3.7 special out SR2.TIM2
TIM1 Timer 1
76543210
wReload value
00000000Res
TIM2 Timer 2
76543210
wReload value
00000000Res
CDC16xxF-E ADVANCE INFORMATION
100 March 31, 2003; 6251-606-2AI Micronas
Table 13–4: Reload Register Programming
Reload
value Output interrupt
source frequency is
divided by
Output Tn-OUT is
divided by
00h 1 2
01h 2 4
02h 3 6
:: :
FFh 256 512
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 101
14. Pulse Width Modulator (PWM)
A PWM is an 8-bit reload down-counter with fixed reload
interval. It serves to generate a frequency signal with vari-
able pulse width or, with an external low pass filter, as a digi-
tal-to-analog converter.
The number of PWMs implemented is given in table 14–1.
The “x” in register names distinguishes the module number.
Fig. 14–1: PWM Block Diagram
14.1. Principle of Operation
14.1.1. General Remarks
A PWM’s 8-bit down-counter is clocked by its input clock and
counts down to zero. Reaching zero, it stops and sets the
output to LOW. A period input pulse reloads the counter with
the content of the PWM register, restarts it and sets the out-
put to HIGH.
A PWM is not affected by CPU SLOW mode.
14.1.2. Hardware settings
The clock and period input frequencies can be set via HW
option (Table 14–1). For full resolution a clock-to-period fre-
quency ratio of 256 is recommended. Should other ratios be
used, make sure that the combination of clock, period and
pulse width setting allow the PWM to generate an output sig-
nal with a LOW transition.
8
zero
Pulse Width Register
PWMx
w
clk
HW Option
SQ
R
1
0
HW Option
load
8 bit down counter
1
0
SR.PWMx
PWMx
period
clock
1
0
Table 14–1: Module specific settings
Module Name HW Options Initialization Enable Bit
Item Address Item Setting
PWM0 Clock and period FFA1h
FFA2h PWM0
output H1.0 or H2.4
special out SR0.PWM0
H1.0 SME/PWM0 output multiplexer FFC2h
PWM1 Clock and period FFA3h
FFA4h PWM1
output H3.0
special out SR0.PWM1
PWM2 Clock and period FFA5h
FFA6h PWM2
output H1.1 or U5.6
special out SR2.PWM2
H1.1 SME/PWM2 output multiplexer FFC2h
CDC16xxF-E ADVANCE INFORMATION
102 March 31, 2003; 6251-606-2AI Micronas
Some of the PWM outputs share pins with outputs of other
modules. The output multiplexer is controlled by HW option
(Table 14–1).
14.1.3. Initialization
Prior to entering active mode, proper SW initialization of the
H-Ports and U-Ports assigned to function as PWMx outputs
has to be made (Table 14–1). The ports have to be config-
ured Special Out. Refer to “Ports” for details.
14.1.4. Operation
After reset, a PWM is in standby mode (inactive) and the out-
put signal PWMx is LOW.
For entering active mode, set the respective enable bit
(Table 14–1). Then write the desired pulse width value to
register PWMx. Each PWM will start producing its output sig-
nal immediately after the next subsequent input pulse on its
period input.
During active mode, a new pulse width value is set by simply
writing to the register PWMx. Upon the next subsequent
input pulse on its period input the PWM will start producing
an output signal with the new pulse width value, starting with
a HIGH level.
Returning a PWM to standby mode by resetting its respec-
tive enable flag will immediately set its output LOW.
The state of the down-counters is not readable.
14.2. Registers
PWM3 Clock and period FFA1h
FFA2h PWM3
output H3.1
special out SR2.PWM3
PWM4 Clock and period FFA3h
FFA4h PWM4
output H2.5
special out SR2.PWM4
Table 14–1: Module specific settings
Module Name HW Options Initialization Enable Bit
Item Address Item Setting
PWMx PWMx Register
76543210
Table 14–2: Pulse Width Programming
Pulse width
value Pulse duty factor
00h 0% (Output is permanently low)
01h 1/256
02h 2/256
::
FEh 254/256
FFh 100% (Output is permanently high) 1)
1) Pulse duty factor 255/256 is not selectable.
wPulse width value
00000000Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 103
15. Capture Compare Module (CAPCOM)
The Capture Compare Module (CAPCOM) is a complex rela-
tive timer. It comprises a free-running 16-bit Capture Com-
pare Counter (CCC) and a number of Subunits (SU). The
timer value can be read by SW.
A SU is able to capture the relative time of an external event
input and to generate an output signal when the CCC
exceeds a predefined timer value. Three types of interrupts
enable interaction with SW. Special functionality provides an
interface to the asynchronous external world.
– 16-bit free running counter with read out.
– 16-bit capture register.
– 16-bit compare register.
– Input trigger on rising, falling or both edges.
– Output action: toggle, low or high level.
– Three different interrupt sources: overflow, input, compare
– Designed for interface to asynchronous external events
Fig. 15–1: CAPCOM Module Block Diagram
A
B=
&
>1
0
0
1
1
0
1
0
1
0
0
1
1
0
1
0
1
TOGGLE
LOW
76543210 76543210
2
MSK MSK MSK FOLCAP CMP OFL LAC RCR XXX
16-Bit Compare-Register
16-Bit Capture-Register r
w
16
16
32
01
2
ofl
CC0MCC0I
CC0
>1
load
CC0COMP
CC0OR
IAMOAM
&
&
reset
Output Action Logic
Input Action Logic Interrupt
Source
Interrupt
Source
16
Subunit 0
CCC
clk
fclk
CC0-IN
CC0-OUT
16
Timer Value
Subunit 1 CC1COMP
CC1ORCC1-IN
CC1-OUT
Subunit 2 CC2COMP
CC2ORCC2-IN
CC2-OUT
16
SR0.CCC
oflTimer Value
oflTimer Value
CCCOFL
Interrupt
Source
HW Option
CDC16xxF-E ADVANCE INFORMATION
104 March 31, 2003; 6251-606-2AI Micronas
15.1. Principle of Operation
15.1.1. General Remarks
The Capture Compare Module (CAPCOM, Fig. 15–1) con-
tains one common free-running 16-bit counter (CCC) and a
number of capture and compare subunits (SU). More details
are given in Table 15–1. The timer value can be read by SW
from 16-bit register CCC. The CCC provides an interrupt on
overflow.
Each SU is able to capture the CCC value at a point of time
given by an external input event processed by an Input
Action Logic.
A SU can also change an output line level via an Output
Action Logic at a point of time given by the CCC value.
Thus, a SU contains a 16-bit capture register CCx to store
the input event CCC value, a 16-bit compare register CCx to
program the Output Action CCC value, an 8-bit interrupt reg-
ister CCxI and an 8-bit mode register CCxM. Two types of
interrupts per SU enable interaction with SW.
For limitations on operating the CAPCOM module in CPU
SLOW mode, see section 15.1.5.3.
15.1.2. Hardware Settings
The CCC clock frequency must be set via HW option (Table
15–1). Refer to “HW Options” for setting them.
15.1.3. Initialization
After system reset the CCC and all SUs are in standby mode
(inactive).
In standby mode, the CCC is reset to value 0000h. Capture
and compare registers CCx are reset. No information pro-
cessing will take place, e.g., update of interrupt flags. How-
ever, the values of registers CCxI and CCxM are only reset
by system reset, not by standby mode. Thus it is possible to
program all mode bits in standby mode and a predetermined
startup out of standby mode is guaranteed.
Prior to entering active mode, proper SW configuration of the
U-Ports assigned to function as Input Capture inputs and
Output Action outputs has to be made (Table 15–1). The
Output Action ports have to be configured as special out and
the Input Capture ports as special in. Refer to “Ports” for
details.
15.1.3.1. Subunit
For a proper setup the SW has to program the following SU
control bits in registers CCxI and CCxM: Interrupt Mask
(MSK), Force Output Logic (FOL, 0 recommended), Output
Action Mode (OAM), Input Action Mode (IAM), Reset Cap-
ture Register (RCR, 0 recommended), and Lock After Cap-
ture (LAC). Refer to section 15.2. for details.
Please note that the compare register CCx is reset in
standby mode. It can only be programmed in active mode.
15.1.4. Operation of CCC
For entering active mode of the entire CAPCOM module set
the enable bit (Table 15–1).
The CCC will immediately start up-counting with the selected
clock frequency and will deliver this 16-bit value to the SUs.
The state of the counter is readable by reading the 16-bit
register CCC, low byte first. Upon reading the low byte, the
high byte is saved to a temporary latch, which is then
accessed during the subsequent high byte read. Thus, for
time stamp applications, read consistency between low and
high byte is guaranteed.
The CCC is free-running and will overflow from time to time.
This will cause generation of an overflow interrupt event. The
interrupt (CCCOFL) is fed directly to the Interrupt Controller
and also to all SUs where further processing takes place.
15.1.5. Operation of Subunit
15.1.5.1. Compare and Output Action
To activate a SUs compare logic the respective 16-bit com-
pare register CCx has to be programmed, low byte first. The
compare action will be locked until the high byte write is com-
pleted. As soon as CCx setting and CCC value match, the
following actions are triggered:
– The flag CMP in the CCxI register is set.
– The CCxCOMP interrupt source is triggered.
– The CCxOR interrupt source is triggered when activated.
– The Output Action logic is triggered.
Four different reactions are selectable for the Output
Action signal: according to field CCxM.OAM (Table 15–2)
the equal state will lead to a high or low level, or toggling
or inactivity on this output.
Another means to control the Output Action is bit
CCxM.FOL. E.g. rise-mode and force will set the output
pin to high level, fall-mode and force to low level. This
forcing is static, i.e., it will be permanently active and may
override compare events. Thus it is recommended to set
and reset shortly after that, i.e., to pulse the bit with SW.
Toggle mode of the Output Action logic and forcing leads
to a burst with clock-frequency and is not recommended.
Table 15–1: Unit specific settings
Sub-
unit HW Options Initialization Enable
Bit
Item Address Item Setting
SU0 Input
clock FFA9h CC0-
OUT U5.1
special out SR0.
CCC
CC0-IN U5.0
special in
SU1 CC1-
OUT U4.6 & U3.6
special out
CC1-IN U4.7
special in
SU2 CC2-
OUT U3.3
special out
CC2-IN U4.6
special in
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 105
15.1.5.2. Capture and Input Action
The Input Action logic operates independently from the Out-
put Action logic and is triggered by an external input in a way
defined by field CCxM.IAM. As shown in Table 15–3 it can
completely ignore events, trigger on rising or falling edge or
on both edges. When triggered, the following actions take
place:
– Flag CCxI.CAP is set.
– The CCxOR interrupt source is triggered when activated.
– The 16-bit capture register CCx stores the current CCC
value, i.e., the “time” of the external event. Read CCx low
byte first. Further compare action will be locked until the
subsequent high byte read is completed. Thus a coherent
result is ensured, no matter how much time has elapsed
between the two reads.
Some applications suffer from fast input bursts and a lot of
capture events and interrupts in consequence. If the SW
cannot handle such a rate of interrupts, this could evoke
stack overflow and system crash. To prevent such fatal situa-
tions the Lock After Capture (LAC) mode is implemented. If
bit CCxI.LAC is set, only one capture event will pass. After
this event has triggered a capture, the Input Action logic will
lock until it is unlocked again by writing an arbitrary value to
register CCxM. Please make sure that this write only
restores the desired setting of this register.
Programming the Input Action logic while an input transition
occurs may result in an unexpected triggering. This may
overwrite the capture register, lock the Input Action logic, if in
LAC mode, and generate an interrupt. Please ensure that
SW is prepared to handle such a situation.
For testing purposes, a permanent reset (FFFFh) may be
forced on capture register CCx by setting bit CCxI.RCR.
Please make sure that the reset is only temporary.
15.1.5.3. Interrupts
Each SU supplies two internal interrupt events:
1. Input Capture event and
2. Comparator equal state.
As previously explained, interrupt events will set the corre-
sponding flags in register CCxI. In addition to the above
mentioned two, the CCC Overflow interrupt event sets flag
CCxI.OFL in each SU. Thus, three interrupt events are avail-
able in each SU. The corresponding flags are masked with
their mask bits in register CCxM and passed to a logical or.
The result (CCxOR) is fed to the interrupt controller as a first
interrupt source. In addition, the Comparator equal (CCx-
COMP) interrupt is directly passed to the interrupt controller
as second interrupt source. Thus a SU offers four types of
interrupts: CCC overflow (maskable ored), input capture
event (maskable ored) and comparator equal state
(maskable ored and non-maskable direct).
All interrupt sources act independently, parallel interrupts are
possible. The interrupt flags enable SW to determine the
interrupt source and to take appropriate action. Before
returning from the interrupt routine, the corresponding inter-
rupt flag should thus be cleared by writing a 1 to the corre-
sponding bit location in register CCxI.
The interrupts generated by internal logic (CCC Overflow
and Comparator equal) will trigger in a predetermined and
known way. However, as explained in 15.1.5.2. erroneous
input signals may cause some difficulties concerning the
Input Capture input, as well as interrupt handling. To over-
come possible problems, the Input Capture Interrupt flag
CCxI.CAP is double-buffered. If a second or even more input
capture interrupt events occur before the interrupt flag is
cleared (i.e. SW was not able to keep track), the flag goes to
a third state. Two consecutive writes to this bit in register
CCxI are then necessary to clear the flag. This enables SW
to detect such a multiple interrupt situation and eventually to
discard the capture register value, which always relates to
the latest input capture event and interrupt.
The internal CAPCOM module control logic always runs on
the oscillator frequency, regardless of CPU SLOW mode.
Avoid write accesses to the CCxI register in CPU SLOW
mode since the logic would interpret one CPU access as
many consecutive accesses. This may yield unexpected
results concerning the functionality of the interrupt flags. The
following procedure should be followed to handle the capture
interrupt flag CAP:
1. SW responds to a CAPCOM interrupt, switching to CPU
FAST mode if necessary and determining that the source is a
capture interrupt (CAP flag =1).
2. The interrupt service routine is processed.
3. Just before returning to main program, the service routine
acknowledges the interrupt by writing a 1 to flag CAP.
4. The service routine reads CAP again. If it is reset, the rou-
tine can return to main program as usual. If it is still set an
external capture event overrun has happened. Appropriate
actions may be taken (i.e. discarding the capture register
value etc.).
5. go to 3.
15.1.6. Inactivation
The CAPCOM module is inactivated and returned to standby
mode (power down mode) by setting the enable bit to 0. Sec-
tion 15.1.3. applies.
CCxI and CCxM are only reset by system reset, not by
standby mode.
15.1.7. Precautions
16-bit CPU commands do not generally keep to a certain
order in addressing high and low bytes of a register. Make
sure that the command used performs reading and writing a
16-bit value low byte first.
In case of uncertainty use 8-bit commands.
CDC16xxF-E ADVANCE INFORMATION
106 March 31, 2003; 6251-606-2AI Micronas
15.2. Registers
The CAPCOM module counter has to be read low byte first
to avoid inconsistencies.
MSK Mask Flag
r/w1: Enable.
r/w0: Disable.
These mask flags refer to the corresponding event flags in
CAPCOM interrupt register.
FOL Force Output Action Logic
r/w1: Force Output Action logic.
r/w0: Release Output Action logic.
This flag is static. As long as FOL is true neither comparator
can trigger nor SW can force, by writing another “one”, the
Output Action logic. After forcing it is recommended to clear
FOL unless Output Action logic should not be locked.
OAM Output Action Mode
r/w: Defines behavior of Output Action logic.
IAM Input Action Mode
r/w: Defines behavior of Input Action logic.
CAP Capture Event
r1: Event.
r0: No Event.
w1: Clear flag.
w0: No change.
CMP Compare Event
r1: Event.
r0: No Event.
w1: Clear flag.
w0: No change.
OVL Overflow Event
r1: Event.
r0: No Event.
w1: Clear flag.
w0: No change.
LAC Lock After Capture
r/w1: Enable.
r/w0: Disable.
Refer to section 7.1.5.2
RCR Reset Capture Register
r/w1: Reset capture register permanently to FFFFh.
r/w0: Release capture register.
CCC CAPCOM Counter low byte
76543210
CCC CAPCOM Counter high byte
76543210
CCxM CAPCOM x Mode Register
76543210
Table 15–2: OAM usage
Bit
3 2 Output Action Logic Modes
0 0 Disabled, ignore trigger, output low level.
0 1 Toggle output.
1 0 Output low level.
1 1 Output high level.
rRead low byte and lock CCC
00000000Res
rRead high byte and unlock CCC
00000000Res
r/w MSK MSK MSK FOL OAM IAM
00000000Res
Table 15–3: IAM usage
Bit
1 0 Input Action Logic Modes
0 0 Disabled, don’t trigger.
0 1 Trigger on rising edge.
1 0 Trigger on falling edge.
1 1 Trigger on rising and falling edge.
CCxI CAPCOM x Interrupt Register
76543210
r/w CAP CMP OFL LAC RCR x x x
00000000Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 107
CCx CAPCOM x Capture/Compare Regis-
ter low byte
76543210
CCx CAPCOM x Capture/Compare Regis-
ter high byte
76543210
rRead low byte of capture register and lock it.
wWrite low byte of compare register and lock it.
11111111Res
rRead high byte of capture register and unlock it.
wWrite high byte of compare register and unlock it.
11111111Res
CDC16xxF-E ADVANCE INFORMATION
108 March 31, 2003; 6251-606-2AI Micronas
16. Stepper Motor Module (SMM)
The SMM serves to control air-cored movements or stepper
motors that are directly coupled in H-bridge formation to H-
Ports. Upon CPU programming it creates all waveforms nec-
essary to position the drive pointer as desired.
The number of motors that are controllable by subunits (con-
trol units) of the module is given in Table 16–4.
Features
– Multichannel pulse width modulated output
– Outputs offset for improved EMC properties
– Four quadrant operation
– 8-bit resolution
– HW Option selectable output cycle frequency
16.1. Functional Description
An 8-bit, free-running counter FRC (see Fig. 16–2) operates
on the fSM input clock (generally 4MHz) and creates an 8-bit
counter word that is fed to a number of control units SMx.
A control unit (Fig. 16–2) contains 8-bit sine and cosine com-
pare registers. One comparator each is associated with
these registers and creates a compare signal when register
content and FRC word are equal. An output flip-flop associ-
ated with each comparator is set when the FRC word is zero
and reset by the respective compare signal. A delay stage
associated with each control unit delays the flip-flop output
signals by a fixed number of fSM cycles to achieve non-syn-
chronism between the output signals of the various control
units, thus achieving an improved EMC behavior of the SMM
(cf. Fig. 16–1). According to the setting of a quadrant register
associated with each control unit, each of a unit’s two output
signals is multiplexed to signals SMxn+ and SMxn− so as to
properly control 2 individual H-Ports that form an H-bridge
together with the connected motor coil. By these means, a
control unit supplies two H-bridges with signals SMx1+,
SMx1−, SMx2+ and SMx2− to function as variable pulse
width modulator outputs with selectable polarity.
Summing up: when the compare registers are set to the sine
and cosine value of a desired rotor angle and the quadrant
register is set to the desired quadrant, an air-cored move-
ment or a stepper motor connected to the unit’s 4 H-Ports
will carry the proper average coil currents of proper polarity
so that its rotor will assume the desired rotary angle.
Three registers control readjustment of a rotor to a new
angle. Sine, cosine and unit/quadrant registers serve as tem-
porary storage of new sine, cosine, related quadrant and unit
selection values. A scheduler logic times the synchronous
downloading of the three buffered words to the respective
unit’s sine, cosine and quadrant registers, so as to avoid
inconsistencies among them. A busy bit may be read out,
signalling completion of the downloading.
Fig. 16–1: Timing Diagram of Output Signals
ftrig =fSM /28
SMA1+
SMA1−
SMA2+
SMA2−
SMB1+
SMB1−
SMB2+
SMB2−
Example:
SMA in
1st quadrant
Example:
SMB in
4th quadrant
1/ftrig
tdA = 0/ fSM
tdB = 1/ fSM
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 109
Fig. 16–2: Block Diagram of Output Generation Circuit
x x x x x x x B
RS
Q
RS
QDELAY
0 / fSM
Comparator sin A “=”
sin A comp. latch
(reload register)
Comparator cos A “=”
cos A comp. latch
(reload register)
8bit FRC
Sine registerCosine register
SMVSIN
SMVCOS
SR0.SM
fSM /2E8
PWM0-OUT
PWM2-OUT
SME1-
SME1+
HW Option
sin E / cos E
sin D / cos D
sin C / cos C
sin B / cos B
sin
Load
cos
Quadrant
register
and
decoder
Scheduler
SMA
DELAY
1 / fSM
DELAY
0 / fSM
SMB
DELAY
2 / fSM
DELAY
3 / fSM
DELAY
4 / fSM
fCLK
ftrig
fSM
Load A
SMC
SMD
SME
Busy
5
4
3
2
1
Load E
SMA1+
SMA1-
SMA2+
SMA2-
SMB1+
SMB1-
SMB2+
SMB2-
SMC1+
SMC1-
SMC2+
SMC2-
SMD1+
SMD1-
SMD2+
SMD2-
SME2+
SME2-
Load B
Load D
OVFL
Load C
ww
SMVC
w
r
3
SEL QUADxxx
CDC16xxF-E ADVANCE INFORMATION
110 March 31, 2003; 6251-606-2AI Micronas
16.2. Registers
SEL Control unit Selection field (Table 16–1)
QUAD Quadrant selection field (Table 16–2)
BUSY Scheduler Busy Flag
r0: Scheduler not busy
r1: Scheduler busy, do not write registers
SMVC, SMVCOS, SMVSIN
16.3. Principle of Operation
The SMM may only be operated in CPU FAST mode.
16.3.1. Hardware settings
Prior to entering active mode, the fSM input clock has to be
set by HW Option (see Table 16–4 on page 111). A fre-
quency value of 4 MHz is recommended, resulting in a pulse
width modulator cycle frequency of 4 MHz/256.
Some H-Ports may receive the output signals either of the
SMM or of PWM modules as an alternative. Refer to Table
16–4 for the necessary settings.
Refer to section “HW Options” for details.
16.3.2. Initialization
Prior to entering active mode, proper SW initialization of the
H-Ports assigned to function as H-bridge outputs SMxn+ and
SMxn- has to be made (Table 16–4). The H-Ports have to be
configured Special Out. Refer to “Ports” for details.
16.3.3. Operation
After reset, the SMM is in standby mode (inactive). The out-
put lines to the H-Ports are low.
SMVC SMM Control Register
76543210
Table 16–1: SEL usage
SEL selected control unit
000 SMA
001 SMB
010 SMC
011 SMD
100 SME
No other values are permitted.
Table 16–2: QUAD setting and resulting control unit output
signal function
QUAD Control unit output signal function
SMx1+ SMx1- SMx2+ SMx2-
00 sine VSS cosine VSS
01 sine VSS VSS cosine
10 VSS sine VSS cosine
11 VSS sine cosine VSS
wx x SEL x QUAD
xx000x00Res
SMVSIN SMM Sine Register
76543210
SMVCOS SMM Cosine Register
76543210
Table 16–3: Usage of SMVSIN and SMVCOS registers
Value Duty factor Pulse Diagram
00h 0/256 (continu-
ously low)
01h 1/256
02h 2/256
::
FEh 254/256
FFh 255/256 1)
1) 256/256 (continuously high) is not available.
rxxxxxxxBUSY
w8bit Sine Value
00000000Res
w8bit Cosine Value
00000000Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 111
For entering active mode, set bit SR0.SM. The FRC will
immediately start counting but the control units’ output lines
will still be low.
16.3.3.1. Generating Output
After entering active mode, the SMM’s control units are
ready to receive sine, cosine and quadrant values.
First load the unit/quadrant information to register SMVC,
then the cosine value to register SMVCOS and at last the
sine value to register SMVSIN. Upon writing SMVSIN, the
scheduler logic will set flag SMVSIN.BUSY and load the buff-
ered values to the respective unit’s sine, cosine and quad-
rant registers on the next zero transition of the FRC, after a
maximum of 256 fSM input clock cycles. After completing the
download, flag BUSY is reset and the respective unit will
immediately start producing the output signals with the
desired timing (see Table 16–3) on the proper pins (see
Table 16–2).
The above procedure for loading values to a first unit is
repeated for all others. Make sure that the BUSY flag is 0
before rewriting registers SMVC, SMVCOS and SMVSIN.
16.3.4. Inactivation
Returning the SMM to standby mode by resetting bit
SR0.SM will immediately halt the FRC, return all output sig-
nals to 0, reset all internal registers.
Table 16–4: Unit specific settings
Contr.
Unit HW Options Initialization Enable Bit
Item Address Item Setting
SMA SMAn+/- outputs H0.0 to H0.3 special out SR0.SM
SMB SMBn+/- outputs H1.2 to H1.5 special out
SMC SMCn+/- outputs H2.0 to H2.3 special out
SMD SMDn+/- outputs H3.2 to H3.5 special out
SME SME/PWM selection FFC2h SMEn+/- outputs H0.4 to H1.1 special out
All Input clock selection FFAEh
CDC16xxF-E ADVANCE INFORMATION
112 March 31, 2003; 6251-606-2AI Micronas
16.4. Rotor Zero Position Detection (RZPD)
In addition to above descriptions this module supports the
Rotor Zero Position Detection by supplying motor blockage
information.
The Rotor Zero Position Detection capability is protected by
a patent from Siemens VDO Automotive (SV) and may only
be used with SV’s prior approval.
Fig. 16–3: Block Diagram of Rotor Zero Position Detection Circuit
16.4.1. Functional Description
Each control unit contains circuitry to detect an induced volt-
age resulting from the rotation of the connected motor’s rotor
(Fig. 16–3). A comparator compares the input voltage from
one of the unit’s H-Ports to 1/9th of the supply voltage. A
capture logic opens a capture window and samples the com-
parator output. The capture result signal supplies a rotor
blockage information necessary for the Rotor Zero Position
Detection in all cases where the CPU has lost track of the
display angle of a pointer that is driven by the motor via a
mechanical transmission.
16.4.2. Registers
ACRA to E Analog Comparator Control and Result
for SMA to SME
r0: Capture result: no induced voltage detected
r1: Capture result: induced voltage detected
w0: Stop capture and clear result flag
w1: Start capture
Debouncer and
measurement
window
+
−Result
latch
SMA
+
−
SMB
HVDD1 HVDD2
+
−
SMC
+
−
SMD
+
−
SME
S
R
fCPU
8R
R
SR0.SM
SMA-COMP
SMB-COMP
SMC-COMP
SMD-COMP
SME-COMP
8R
R
ABCDExxx SMVCMPr/w
0
1
2
4
5
01245
SMVCMP SMM Comparator Register
76543210
r/w x x ACRD ACRB x ACRE ACRC ACRA
xx00x000Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 113
16.4.3. Principle of Operation
The RZPD can only be operated together with the Stepper
Motor module. Switching SR0.SM connects/disconnects the
comparators from supply and resets all registers.
During Rotor Zero Position Detection one of a unit’s H-Ports
(Table 16–5) temporarily has to be operated as an input to an
internal analog comparator. Reconfigure this port as Special
Input. Refer to “Ports” for details.
Reading of the induced voltage at the measured motor wind-
ing is started by setting the questioned unit’s control bit
SMVCMP.ACRx to 1. The respective analog comparator’s
output will now be sampled. Once three consecutive ’1’ sam-
ples (spaced 1/fCPU) - indicating a sufficient analog compara-
tor input voltage - are received, a ’1’ may be read from the
questioned unit’s result flag SMVCMP.ACRx, indicating that
the Rotor Zero Position Detection is under way.
Resetting the questioned unit’s control bit SMVCMP.ACRx to
0 stops the sampling and resets the result flag. When, after a
restart of the above sampling procedure and after a suffi-
ciently long capture period, a ’1’ was still not read from the
questioned unit’s result flag SMVCMP.ACRx, this indicates
that the Rotor Zero Position Detection is complete.
Parallel Rotor Zero Position Detection on all control units is
permitted.
After completion of Rotor Zero Position Detection, reconfig-
ure the comparator input port as Special Out.
Table 16–5: RZPD input ports
Contr.
Unit Initialization
Item Setting
SMA SMA-COMP input H0.0 special in
SMB SMB-COMP input H1.2 special in
SMC SMC-COMP input H2.0 special in
SMD SMD-COMP input H3.2 special in
SME SME-COMP input H0.4 special in
CDC16xxF-E ADVANCE INFORMATION
114 March 31, 2003; 6251-606-2AI Micronas
17. LCD Module
The Liquid Crystal Display (LCD) Module is designed to
directly drive a 1:4 multiplexed liquid crystal display. It gener-
ates all signals necessary to drive 4 backplane and 48 seg-
ment lines which are output via U-Ports in LCD mode. Up to
192 segments or pixels can be controlled if all U-Ports are
designated as segment outputs.
In addition, the module provides functions that enable the
user to cascade it with external expansion ICs providing
more segment lines. It can be operated as master or slave in
such an extended system.
Features
– 1:4 multiplex
– 5 V supply
– Maximum of 192 segments
– Cascadable with external expansion ICs
– 0.3 mA buffered 1/3 and 2/3 voltage divider
– Zero standby current
–200µA no load active current
– Frame frequency HW Option selectable
17.1. Principle of Operation
17.1.1. General Remarks
Each LCD pixel or segment which is controlled by the LCD
module is located at the crossing point of a segment line and
a backplane line. The LCD module co-ordinates the output
sequences of backplane and segment lines (see Fig. 17–3
on page 117).
Fig. 17–1: Segments and Backplanes
A segment pin can drive 4 different voltage levels (UVDD, 1/
3 VDD, 2/3 VDD, VDD) in LCD mode. The output of each
segment pin is controlled by the segment field of the corre-
sponding UxSEGy register. Each such register contains the
segment fields of two neighboring segment pins. A segment
field is 4 bits wide. Each flag (0 to 3) of a segment field corre-
sponds to a backplane line (BP0 to BP3). If the segment flag,
corresponding with the backplane line BPx is true, the seg-
ment at the crossing of the two lines is on (black).
The LCD module does not contain a display ROM translating
character information into segment code. The advantage is
that arbitrary characters or displays can be generated just by
changing the program code. Segment information is directly
entered by writing to the corresponding UxSEGy register. It
is validated (loaded to all corresponding slave registers) for
all segment U-Ports simultaneously by a write access to
register U1SEG10.
Two internal voltage sources provide the U-Port circuits and
the backplane generator with the voltage levels 1/3 UVDD
and 2/3 UVDD. These levels are generated by a buffered
resistor divider.
17.1.2. Hardware settings
The LCD frame frequency is settable by HW option FFADh.
The resulting frame frequency is the selected input fre-
quency, divided by 120. It should be in the range from 50 Hz
to 200 Hz.
For best electromagnetic interference results it is recom-
mended to operate all segment and backplane U-Ports in
Port Slow mode. Refer to “Ports” for more details and to
“HW-Options” for setting the corresponding HW options. Set
flag PFST in register SR1 to HIGH to enable Port Slow
mode.
17.1.3. Initialization
After reset, the LCD module is in standby mode (inactive)
and all U-Ports are in Port mode, non-conducting.
All U-Ports designated to function as backplane or segment
outputs are to be set to LCD mode. Refer to “Ports” for more
details. This will set these U-Ports to output LOW state.
After reset the content of the segment registers is undefined.
It must be set by writing the desired segment information to
registers UxSEGy and by validating it by a write access to
register U1SEG10 (write 00h for master mode, FFh for slave
mode), before the LCD module is enabled.
17.1.4. Operation
For entering active mode, set flag LCD in register SR1. Each
segment and backplane U-Port will immediately start produc-
ing its LCD output signal according to the segment informa-
tion provided during initialization.
During active mode, a new segment information is entered
by simply writing the desired segment information to regis-
ters UxSEGy and by validating it by a write access to register
U1SEG10 (write 00h while in master mode, FFh while in
slave mode). Each segment and backplane U-Port will
immediately start producing an LCD output signal according
to the new segment information.
Returning the LCD module to standby mode by resetting flag
LCD in register SR1 will immediately return all segment and
backplane U-Ports to the output LOW state.
BP3 BP2 BP1 BP0
SEGn
SEGn+1
SEGn-1
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 115
Fig. 17–2: Block Diagramm
1/1
1/1,5
1/15
1/2,5
1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1
LCD CLK IN
LCD CLK OUT
LCD SYNC OUT
LCD SYNC IN
1
0
1
0
U1SEG10.LCDSLV
SR1.LCD
UV
DD
2
/
3
UV
DD
1
/
3
UV
DD
UV
SS
8 x frame frequency
+
-
+
-
UV
DD
2
/
3
UV
DD
1
/
3
UV
DD
UV
SS
213 03210
wr U1SEG10 load
U1M54.
PMODE U7M32.
PMODE U1M30.
PMODE
3210
HW Option
UV
DD
reset
overflow
U1.5 U1.4 U7.3 U7.2 U1.3 U1.2 U1.1 U1.0
U1SEG54 U7SEG32
Analog Switch
and
Segment Driver
Analog Switch
and
Segment Driver
Analog Switch
and
Segment Driver
Analog Switch
and
Segment Driver
213 03210
3
HW Option
SR1.LCD
1010 10 10 10 101010
Back Plane Generator
3 33
8 State
Counter
R
R
R
Analog
Switch
and
Backplane
Driver
Analog
Switch
and
Backplane
Driver
Analog
Switch
and
Backplane
Driver
Analog
Switch
and
Backplane
Driver
f
clk
CDC16xxF-E ADVANCE INFORMATION
116 March 31, 2003; 6251-606-2AI Micronas
The state of the segment registers is not readable.
All LCD operations are not affected by CPU SLOW mode.
17.1.5. Cascading of LCD Driver Modules
For expansion purposes, the LCD module may be cascaded
with external LCD driver ICs. Master or slave mode is select-
able for the LCD module while in standby mode. Special sig-
nals provide phase and frequency synchronism for the LCD
frame among the cascaded ICs.
For master mode, set flag LCDSLV in register U1SEG10
LOW. The module always directs signal LCD-SYNC-OUT to
pins U1.4 and U6.0 and LCD-CLK-OUT to pins U1.5 and
U6.1. They connect to external slave ICs’ SYNC-IN and
CLK-IN inputs for synchronization.
For slave mode, set flag LCDSLV in register U1SEG10
HIGH. Configure pins U5.2 and U5.3 to receive signals LCD-
SYNC-IN and LCD-CLK-IN from an external master IC’s
SYNC-OUT and CLK-OUT outputs. These signals will then
substitute the LCD module’s own HW option frame fre-
quency settings.
Starting up and shutting down such an expanded system is
described in section 17.3.
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 117
Fig. 17–3: Frame Timing Diagram
A segment at a crossing of backplane and segment lines is
turned black when at the same time the backplane driver out-
puts a full swing and the segment driver outputs a full swing
of opposite polarity.
VDD
2/3
1/3
0
1 Frame
Back plane BP0
Back plane BP1
Back plane BP2
Back plane BP3
SEG6.3
SEG6.4
SEG6.5
SEG6.6
SEG6.7
3 2 1 0
0 0 0 0
1 0 0 0
0 1 1 0
1 0 1 0
1 1 1 1
SEG63_
SEG64_
SEG65_
SEG66_
SEG67_
Pin-No.Bit-No.Seg-No.
Segment off
Segment on
CDC16xxF-E ADVANCE INFORMATION
118 March 31, 2003; 6251-606-2AI Micronas
17.2. Registers
U-Port registers U1SEG10, U1SEG32 and U1M30 play spe-
cial roles during operation of the LCD module.
LCDSLV LCD Module is Slave
w1: validate all segment information and select
slave mode
w0: validate all segment information and select
master mode
Register U1SEG54 is an example for any Universal Port
Segment Register with exception of U1SEG10 and
U1SEG32. Refer to section Ports for detailed description.
SEG15.3 Segment # 3 of Pin U1.5
This bit defines the segment at the crossing of the lines
SEG1.5 (Pin U1.5) and BP3 (U1.3).
w1: Segment is on (black)
w0: Segment is off (white)
SEG15.2 Segment # 2 of Pin U1.5
This bit defines the segment at the crossing of the lines
SEG1.5 (Pin U1.5) and BP2 (U1.2).
SEG15.1 Segment # 1 of Pin U1.5
This bit defines the segment at the crossing of the lines
SEG1.5 (Pin U1.5) and BP1 (U1.1).
SEG15.0 Segment # 0 of Pin U1.5
This bit defines the segment at the crossing of the lines
SEG1.5 (Pin U1.5) and BP0 (U1.0).
17.3. Software Hints for Cascading LCD Modules
17.3.1. Power-On and Start-Up Procedure
1. The SW in master and slave configures the corresponding
IC.
2. Optionally the slave signals to the master via handshake
link or an inter processor interface (IPI) that it is ready to dis-
play.
3. The slave continuously scans the inputs LCD-CLK-IN and
LCD-SYNC-IN for the bit combination “01” (SW debouncing
required).
4. The master LCD module is switched on. LCD-CLK-OUT
and LCD-SYNC-OUT switch to “01”.
5. The slave CPU detects the bit combination “01” and
immediately switches on the slave LCD module. The slave
LCD now generates a display.
Note: During the time that the slave needs to detect the bit
combination “01”, master and slave operate asynchronously.
Suggestion: limit time to approximately 100 to 200 ms.
6. The LCD modules now operate in controlled synchroniza-
tion.
17.3.2. Operation
In order to obtain optimum synchronization of LCD switch-
over, a change of display must be coordinated between mas-
ter and slave (preferably via IPI) so that the time lag between
write accesses to U1SEG10 of the master and of the slave is
kept as small as possible. Suggestion: Lower ms range or
customer specification.
17.3.3. Power-Off Procedure
1. (Optional) The processor which decides that the display is
to be switched off signals this to the other via IPI.
2. The slave continuously scans the inputs LCD-CLK-IN and
LCD-SYNC-IN for the bit combination “11” (SW debouncing
required).
3. The master LCD module is switched off. LCD-CLK-OUT
and LCD-SYNC-OUT switch to “11”.
4. The slave CPU detects the bit combination “11” and imme-
diately switches off the slave LCD module.
Note: Keep time delay as short as when switching on.
5. All LCD ports then output a low signal. The LCD display is
now inactive.
U1SEG10 Universal Port x Segment Register of
Ux.1 and Ux.0
76543210
U1SEG54 Universal Port 1 Segment Register of
Pin U1.5 and U1.4
76543210
wLCDSLVxxxxxxxLCD
00100010Res
wSEG15.3 SEG15.2 SEG15.1 SEG15.0 SEG14.3 SEG14.2 SEG14.1 SEG14.0 LCD
00100010Res
Table 17–1: Suggested sequence
Master Slave
Load LCD display register. Load LCD display register.
Clear flag LCDSLV. Set flag LCDSLV.
LCD-CLK-OUT, and
LCD-SYNC-OUT:
Configure universal ports
as Special Out Ports.
LCD-CLK-IN, and
LCD-SYNC-IN:
Configure universal ports
as Special In Ports.
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 119
18. DMA
The DMA module allows read and write access to an exter-
nal IC with minimum CPU interaction.
The module is intended to support the operation of external
LCD driver ICs (e.g. SED1560 by Epson):
The DMA module copies 8 bit pixel data bytes by direct
memory access (DMA) to the external IC’s graphic RAM with
the help of that IC’s internal autoincrement address counter,
and without CPU interaction. Other off-chip registers, allow-
ing control of the display behavior (blinking, scrolling, etc.),
are written and read by CPU operations supported and timed
by the DMA module.
The CPU programs the DMA module to copy an array of
data from the internal to the external IC’s memory. The CPU
writes to or reads from external registers while the DMA
module generates the necessary timing and control signals.
Features
– 3 operating modes:
direct memory write access (DMA) without CPU interac-
tion,
support and timing of CPU write access to ext. registers,
support and timing of CPU read access to ext. registers
– 16 MB maximum DMA block size
– one byte DMA block alignment, no confinement to banks
– DMA sequence interruptible by CPU or interrupt controller
– CPU cycles are stolen only during CPU bus access
– flag for CPU-DMA conflict
– Interrupt on DMA transfer finished
– External bus cycle time selectable from Fxtal/2 to Fxtal/
256
– Automatic generation of CPU wait states to support read
access to external registers
Fig. 18–1: Block Diagram
18.1. Principle of Operation
18.1.1. DMA Write Mode
To select the desired write timing, the corresponding bits
have to be set in the DMA Initial Condition Register (DIC).
To obtain the 8-bit pixel data GDB7 .. 0 on the U2.7 .. U2.0
pins, all U2 bits have to be configured as port, normal, out-
puts. To obtain the GWRQ control output on U7.3, this port
has to be configured as port, special, output. Other signals
necessary to control the external IC have to be realized by
software using other ports.
The range of internal addresses to be transmitted is finally
set by first writing the 3 (lower) physical start address bytes
to registers DSA2, DSA1, DSA0, then the 3 (upper) physical
end address bytes to registers DEA2, DEA1 and DEA0.
The module generates the physical 24-bit address, as it is
presented to the memory, even with alternative banking
mode. Refer to the Memory Banking section for translation of
logical addresses as generated by the CPU to physical
addresses.
CPU
Mem
DMA
LCD
driver
BE*
RDY
DB
ADB
GWRQ
GADB
GBus
8
8
24 CS
GRDQ
control
Ifc
GDB
n
fix
U2
U7.2
VPA*, VDA*Logic
U-Port
U7.3
* chip internal signals
CDC16xxF-E ADVANCE INFORMATION
120 March 31, 2003; 6251-606-2AI Micronas
Upon writing DEA0 the DMA module will immediately begin
presenting data on U2 and corresponding control signals on
U7.3.
During the necessary DMA access cycles to the internal
addresses a wait cycle is generated for the CPU whenever it
tries to perform a bus cycle itself.
During operation, a DMA transfer active status bit
(DCS.DTA) is readable from the DMA Control and Status
register DCS to indicate that a requested transfer is not fin-
ished. A busy status bit (DCS.BSY) is readable to indicate
that the DMA is currently active and not halted by the Inter-
rupt Controller or by software. Other bits in the DCS are writ-
able to immediately stop a running DMA sequence, and to
restart it, or to allow the Interrupt Controller, when active, to
stop and when inactive, to restart it.
Upon reaching the end address setting the DMA module fin-
ishes operation. At this time the DMA interrupt source output
is triggered.
The interrupt source output of this module is routed to the
Interrupt Controller logic. But this does not necessarily select
it as input to the Interrupt Controller. Check section “Interrupt
Controller” for the actually selectable sources and how to
select them.
18.1.2. CPU Write Mode
To select the desired write timing and to use U2 data writes
as cycle trigger, the corresponding bits have to be set in the
DMA Initial Condition Register (DIC).
To obtain the GWRQ control output on U7.3, this port has to
be configured as port, special, output. Other signals neces-
sary to control the external IC have to be realized by soft-
ware using other ports.
The CPU configures and operates U2 as port, normal, out-
put. Upon the CPU writing the U2 Data Register (U2D), the
DMA module will present corresponding control signals for
one write cycle to the external IC on U7.3.
If the CPU tries to rewrite U2D before the previous cycle is
finished, the DMA module halts the CPU by generating the
appropriate number of wait cycles.
During operation a busy status bit (BSY) in the DMA Control
and Status Register (DCS) is readable to indicate the busy
condition.
18.1.3. CPU Read Mode
To select the desired read timing and to use U2 data reads
as cycle trigger the corresponding bits have to be set in the
DMA Initial Condition Register (DIC).
To obtain the GRDQ control output on U7.2, this port has to
be configured as port, special, output. Other signals neces-
sary to control the external IC have to be realized by soft-
ware using other ports.
The CPU configures and operates U2 as port, normal, input.
Upon the CPU reading the U2 Data Input (U2D), the DMA
module will present corresponding control signals for one
read cycle from the external IC on U7.2.
The DMA module halts the CPU by generating the appropri-
ate number of wait cycles, until the data read cycle from the
external IC is complete. Thus one CPU read of U2D is suffi-
cient to read data from external IC.
18.2. Registers
The DMA control registers are located in the I/O area.
The registers DSAx (start address) and DEAx (end address)
are writable by the CPU. Write the address of the first byte to
be transferred into the start address register. After the DMA
has finished it points to the first byte after the transferred
block. This makes it easy to transfer consecutive blocks with-
out rewriting the start address register every single time.
Only the end address has to be updated in this case. The
end address has to be initialized with the address pointing to
the first byte after the block to be transferred. Writing the low
byte of the end address (DEA0) starts the DMA sequence.
The first transfer starts after the next opcode fetch. The com-
pare signal stops the DMA sequence if both pointers, DSAx
and DEAx, point at the same memory location. A start or
restart is not possible in this case.
The DMA Control and Status Register (DCS) shows the CPU
whether a requested DMA transfer is not yet finished (DTA
flag), whether a DMA transfer is active and not halted, or a
CPU access is not yet finished (BSY flag), or if a DMA CPU
conflict (DCC flag) has occurred. DCC is set true, if the data,
mode or tristate register of U-Port 2 is addressed by the CPU
though the DMA logic is active. DCC will not be set, if the
respective registers of U-Port 7 are addressed. It allows the
CPU to stop (interrupt) and start (continue) a DMA
sequence. Furthermore, it lets the CPU define whether an
interrupt may stall a DMA sequence.
Table 18–1: DMA Control Registers
Mnemonic Name
DCS DMA Control/Status
DIC DMA Initial Configuration
DSA0 Start Address 0 (low byte)
DSA1 Start Address 1
DSA2 Start Address 2
DEA0 End Address. 0 (low byte)
DEA1 End Address. 1
DEA2 End Address. 2
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 121
Fig. 18–2: DMA Address Generator
DTA DMA Transfer Active
r1: DMA transfer active.
r0: No DMA transfer active.
DSI DMA Stopped by Interrupt
r/w1: DMA stops during interrupt.
r/w0: DMA continues during interrupt.
DSI should only be modified when DTA is zero.
DCC DMA CPU Conflict
r1: DMA CPU conflict.
r0: No DMA CPU conflict.
w0: Clears DCC flag.
STP Stop DMA sequence
w1: Stops the DMA sequence.
w0: Nothing happens.
BSY Busy
r1: DMA or CPU sequence is active.
r0: DMA or CPU sequence is not active.
w1: Starts the DMA sequence.
w0: Nothing happens.
The two flags BSY and STP should be used in common.
Table 18–2 shows the possible bit combinations.
With BSY true, the CPU must neither access the ports or the
DMA address registers, nor change the DMA Initial Configu-
ration register (DIC). Even after interrupting a DMA
sequence by setting the STP flag, it is necessary to guaran-
tee that the BSY flag was cleared by the DMA logic, before
changing those registers.
With DSI active a DMA sequence is stopped while an inter-
rupt is served. In this case the DMA sequence doesn’t
lengthen an interrupt service routine. The start of an interrupt
sends a stop signal to the DMA logic. Only the first of the
nested interrupts sends this stop signal. The end of an inter-
rupt (the last interrupt if they are nested) sends a start signal
to the DMA logic. From this reason, each interrupt can start a
DMA sequence if DSI is true. The only way to avoid a wrong
start is to guarantee that the two address registers DSAx and
DEAx point to the same memory location.
DSI active implies other restrictions to the user:
You can stop a DMA sequence within an interrupt, but it is
not continued until the end of the last of nested interrupts.
You can start a DMA sequence within an interrupt and it is
not stopped by nested interrupts, until the main program is
interrupted again.
Normally, a DMA sequence will be started by writing the end
address to the appropriate registers. Writing to DEA0 starts
the DMA sequence. So this byte has to be written last. If a
DMA sequence is interrupted by SW, it has to be continued
by rewriting DEA0 or by writing a one to BSY. A new DMA
sequence may be started too by setting BSY, if the end
address hasn’t changed. In this case the start address has to
be rewritten because it points at a position after the last
transferred byte.
The DMA Initial Configuration Register (DIC) contains wait
state stuff like the duration of external access cycles or
whether wait states should be generated. If the flag WSA
(Wait States Active) is cleared, the U-Port 2 may be used as
normal I/O or LCD port. In this case no wait states are gener-
ated with CPU access but with DMA access. The bits WSA,
WS0, WS1 and WS2 must not be modified if DTA or BSY are
true.
Start Address (0...23)
ADB (0...23)
0...23
07
8
15
24
16
23
Upcounter
DMAE
End Address Register (0...23)
07
8
15
16
23
DTA
(End Address + 1)
=
DCS DMA Control and Status Register
76543210
Note
Table 18–2: BSY and STP Usage
STP BSY
00
01
10
11
No action
Start DMA sequence
Stop DMA sequence
Not allowed
r/w x x x DTA DSI DCC STP BSY
00000000Res
CDC16xxF-E ADVANCE INFORMATION
122 March 31, 2003; 6251-606-2AI Micronas
WS2 to 0 Wait States Bit 2 to 0
Write only field for programming the number of wait states.
WSA Wait States Active
w1: Wait states at CPU access to U2 and genera-
tion of GWRQ and GRDQ.
w0: No wait states (RDY) and no control signals
(GWRQ, GRDQ) at CPU access.
At DMA wait states are always generated.
The wait state logic controls the duration of external bus
cycles (see Table 18–3 on page 122). It defines the maxi-
mum number of cycles the CPU is stopped. Not each kind of
access stops the CPU the maximum number of cycles. Only
a read from the external memory causes the wait state logic
to generate the (programmable) maximum number of wait
states. A write stops the CPU only if the previous access is
not finished. A DMA sequence causes the CPU to stop for
one Phi2 clock at each byte transfer. The next DMA happens
after n-1 Phi2 clocks if n is the programmed number of wait
states.
The timing in table 18–4 relate on a system clock of 10MHz.
The external address has to be written by SW to the data
latch. So the SW must guarantee the required address setup
time. The GWRQ high to GWRQ low time ratio is symmetri-
cal at DMA accesses. The SW has to guarantee the required
GWRQ high time at CPU accesses. In this case the pro-
grammer can not rely on self timing with the RDY signal. At
consecutive write accesses the GWRQ high time is exactly 2
Phi2 clocks.
DIC DMA Initial Configuration Register
76543210
Note
Table 18–3: Wait States
WS2
to 0 #WS Ext. Bus Fre-
quency @ 10 MHz
0 2 Phi2/2 5 M (Reset)
14Phi2/4 2.5M
28Phi2/8 1.25M
3 16 Phi2/16 625 k
4 32 Phi2/32 312.5 k
5 64 Phi2/64 156.25 k
6 128 Phi2/128 78 k
7 256 Phi2/256 39 k
wx WS2 WS1 WS0 x x x WSA
00000000Res
Table 18–4: Bus timing at 10 MHz
#WS 24816 32 64 128 256 Units
tWDS min. 125 325 725 1525 3125 6325 12725 25525 ns
tACC max. 150 350 750 1550 3150 6350 12750 25550 ns
tDWH max. 100 200 400 800 1600 3200 6400 12800 ns
tDDH max. 50 (always constant 1/2 Phi2) ns
tCWH max. 200 (always constant 2 Phi2) ns
tWDS: Write data setup time
tACC: Read access time
tDWH: DMA write high time
tDDH: DMA write data hold time
tCWH: CPU write high time
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 123
18.3. Ports
The access to external memory is done by universal ports.
The assignment to external signals is shown in table 18–5. Table 18–6 shows the settings of the port configuration regis-
ters in different modes.
18.4. SW Application Hints
CPU must run in CPU FAST mode when accessing external
memory. Don’t try to access (CPU or DMA) in CPU SLOW
mode. Port fast mode is recommended, to guarantee the tim-
ing between control signals and data.
Don’t try to access the external memory or the involved ports
while a DMA sequence is running (DCS.DTA=1).
The ports must be initialized depending on the kind of access
(DMA, write, read). Don’t reconfigure ports or addresses
(DSAx, DEAx) as long as the flags DCS.DTA or DCS.BSY
are true. This may cause problems, particularly if more than
two wait states are programmed. The flag BSY will be set if
the CPU accesses external addresses too. BSY will be
cleared after the transfer cycle has finished.
All involved ports and external addresses may be read or
written as long as there is no DMA working (DTA=0, BSY
=0). If a DMA is working (DTA=1) or a CPU access to an
external address is not finished (BSY=1), it is neither allowed
to access the involved ports and the DMA start and end
address pointers, nor to change the configuration of the wait
state logic. If there is an access, this is a programming error.
Both accesses (DMA and CPU) are disturbed. A flag signals
the occurrence of this conflict for debugging purposes.
There are two different modes to operate the DMA logic. In
the DMA high priority mode a DMA sequence is not affected
by an interrupt. DMAs take place during an interrupt service
routine. Each DMA steals one cycle of the ISR. In the DMA
low priority mode, a DMA sequence is stalled by an interrupt
and continues after the end of the ISR. The ISR is not length-
ened by an active DMA sequence.
Table 18–5: Port Assignment
Port Name
U2.0 GDB0 External data bus
::
U2.7 GDB7
1) GADB External address bus
U7.2 GRDQ External read signal
U7.3 GWRQ External write signal
1) Any universal port may be used as address out-
put port.
Table 18–6: Port Configurations
Mode Register Setting
DMA Write,
CPU Write U2SEG10, 32, 54, 76 00H Normal, out
U2M10, 32, 54, 76 01H Port mode
U7SEG32 44H Special, out
U7M32 01H Port mode
CPU Read U2SEG10, 32, 54, 76 22H Normal, in
U2M10, 32, 54, 76 01H Port mode
U7SEG32 44H Special, out
U7M32 01H Port mode
CDC16xxF-E ADVANCE INFORMATION
124 March 31, 2003; 6251-606-2AI Micronas
18.4.1. DMA High Priority Mode
The flag DSI is, and remains, set to zero. A DMA may inter-
rupt an interrupt service routine. A DMA sequence may be
started, stopped, the registers and ports may be modified
everywhere if you make sure that the flags BSY/DTA are
false.
The overall initialization has to be done once after each
reset.
Overall initialization:
– Configure WSA, WS0, WS1 and WS2. Set DSI to zero.
– Switch data port U2.0 to U2.7 to normal mode.
– Switch address port to normal mode and activate output
driver.
– Switch control port U7.2 and U7.3 to special mode and
activate output driver.
DMA write:
– Activate output driver of data port U2.0 to U2.7.
– Write destination address to address port.
– Write start address to registers DSA2 to DSA0.
– Write end address + 1 to registers DEA2 to DEA0. Writing
to register DEA0 starts the DMA sequence.
CPU write:
– Activate output driver of data port U2.0 to U2.7.
– Write external address to address port.
– Write to data port U2.
CPU read:
– Deactivate output driver of data port U2.0 to U2.7.
– Write external address to address port.
– Read from data port U2.
18.4.2. DMA Low Priority Mode
With the flag DSI set to one, any interrupt will stop a DMA
sequence. The user may modify the DMA logic setting within
an ISR as long as flag DTA is zero. If you want to stop a run-
ning DMA sequence by setting flag STP, it is advisable to
clear flag DSI with the same write access to DSC. Otherwise
the next interrupt would restart this interrupted DMA
sequence.
The overall initialization has to be done once after each
reset.
Overall initialization:
– Configure WSA, WS0, WS1 and WS2. Set DSI to one.
– Switch data port U2.0 to U2.7 to normal mode.
– Switch address port to normal mode and activate output
driver.
– Switch control port U7.2 and U7.3 to special mode and
activate output driver.
DMA write:
– Activate output driver of data port U2.0 to U2.7.
– Write destination address to address port.
– Clear DSI.
– Write start address to registers DSA2 to DSA0.
– Write end address + 1 to registers DEA2 to DEA0. Writing
to register DEA0 starts the DMA sequence.
– Set DSI.
CPU write:
– Activate output driver of data port U2.0 to U2.7.
– Write external address to address port.
– Write to data port U2.
CPU read:
– Deactivate output driver of data port U2.0 to U2.7.
– Write external address to address port.
– Read from data port U2.
18.4.2.1. Unwanted DMA Interrupts in Low Priority Mode
On leaving interrupt service in Low Priority Mode (DSI=1),
the DMA-HW generates a compare of start and end address
register. If they are not equal, a halted DMA sequence is
continued. But if they are equal, a DMA interrupt is gener-
ated instantly. DMA interrupts are generated at every com-
pare, if the result is equal. Thus, an unwanted DMA interrupt
is generated every time the DSI function tries to restart a
DMA that has no transfers pending.
To work around these unwanted interrupts, the following
measures should be taken:
– No special measures have to be taken if a new DMA
sequence is initiated within the DMA ISR (Interrupt Ser-
vice Routine).
– If no new DMA sequence has to be initiated within a DMA
ISR, it is necessary to either clear flag DSI (High Priority
Mode), or to disable the DMA interrupt at the Interrupt
Controller, within the DMA ISR.
– After starting a DMA sequence from within the main rou-
tine, DSI has to be set and/or DMA interrupt has to be
enabled again.
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 125
18.5. Timings
Fig. 18–3: DMA Write, WS = 2
Every second cycle is stolen from the CPU. The external
DMA transfer lasts 2 cycles. The CPU computes an internal
operation at the same time with the third DMA. In this case
the CPU is not stopped. CPU3 is not visible on the buses
ADB or DB.
Phi2
ADB
DB
GADB
GDB
GWRQ
RDY *
BE *
DMA-Src-1 CPU1
DMA1 CPU1
DMA-Src-2 CPU2
DMA2 CPU2
DMA-Src-3 CPU4
DMA3 CPU4
CPU
CPU
DMA-Dest.-Adr.
DMA internal DMA external
DMA1 DMA2 DMA3
DMAE *
Count *
Busy *
V
PA *
VDA *
tDDH
* chip internal signals
CDC16xxF-E ADVANCE INFORMATION
126 March 31, 2003; 6251-606-2AI Micronas
Fig. 18–4: DMA Write, WS = 4
Fig. 18–5: CPU Write, WS = 2
Phi2
ADB
DB
GADB
GDB
GWRQ
RDY *
BE *
DMA-Src-1 CPU1
DMA1 CPU1
CPU2
CPU2
DMA-Src-2 CPU4
DMA2 CPU4
CPU
CPU
DMA-Dest-Adr
DMA internal DMA external
CPU3
CPU3
DMA1 DMA2
DMAx
DMAE *
Count *
Busy *
tWDS
tDWH
tDDH
* chip internal signals
Phi2
ADB
DB
GADB
GDB
GWRQ
CPU1
CPU1
CPU internal CPU external
CPU1
CPU1
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 127
Fig. 18–6: CPU Write, WS = 4, with RDY because CPU rewrites to fast.
Fig. 18–7: CPU Read, WS = 2
Phi2
ADB
DB
GADB
GDB
GWRQ
CPU1
CPU1
CPU internal CPU external
CPU1
CPU1
CPU2
CPU2
CPU2
RDY * tCWH
* chip internal signals
Phi2
ADB
DB
GADB
GDB
GRDQ
CPU1
CPU1
CPU internal
CPU1
RDY *
CPU1
tACC
* chip internal signals
CDC16xxF-E ADVANCE INFORMATION
128 March 31, 2003; 6251-606-2AI Micronas
Fig. 18–8: CPU Read, WS = 4
Phi2
ADB
DB
GADB
GDB
GRDQ
CPU1
CPU1
CPU internal
CPU1
RDY *
CPU1
* chip internal signals
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 129
19. Serial Synchronous Peripheral Interface (SPI)
An SPI module provides a serial input and output link to
external hardware. An 8 or 9-bit data frame can be transmit-
ted synchronized to an internally or externally generated
clock.
The number of SPIs implemented is given in Table 19–1. The
“x” in register names distinguishes the module number.
Features
– 8 or 9-bit frames
– Internal or external clock
– Programmable data valid edge
– Three internal clock sources programmable
– Input deglitcher for clock and data
Fig. 19–1: Block Diagram
SI
Deglitcher
1
1
0
SPIx-D-IN
1 03 25 47 6
1
0
1
0
1
0
1
0
SO
1
SPIx-D-OUT
SO
SI
1/1
1/1,5
1/2,5
3:1
MUX
Deglitcher
1
0
1
1
0
1
Scheduler
SPIx-CLK-IN
SPIx-CLK-OUT
01
shift in
shift out
LEN9
HW Option
76543210
HW Option
INTERN
RXSEL
0
1
INTERN
2
HW Option
F0SPI
F1SPI
F2SPI
intclk
SR0.SPIx
SPIx
Interrupt
Source
HW Option
3xTosc
extclk
clk
clkout
Deglitcher
2
CSF0/1
F0SPI
F1SPI
F2SPI
D1 D0
00
0
1
1
x
NEDGE
BIT8
SPIxM
SPIxD
8
CDC16xxF-E ADVANCE INFORMATION
130 March 31, 2003; 6251-606-2AI Micronas
19.1. Principle of Operation
19.1.1. General Remarks
A SPI serves as an 8 or 9 bit wide input/output shift register.
Either an internally or an externally generated clock can be
used to shift data in and out.
The input SPIx-D-IN is connected to the LSB of the shift reg-
ister. The output of the shift register is connected to output
signal SPIx-D-OUT. Thus each time a frame is transmitted by
shifting bits out, bits are shifted in simultaneously and vice
versa. Deglitchers in the data and clock input paths are
active only in external clock mode. The input and output can
be inverted by HW Option.
If the deglitcher is active, input changes polarity after three
consecutive samples have shown the same new polarity.
Thus, a delay of three oscillator clock cycles is introduced.
This feature imposes a limit on the maximum transmission
frequency.
The interrupt source output of this module is routed to the
Interrupt Controller logic. But this does not necessarily select
it as input to the Interrupt Controller. Check section “Interrupt
Controller” for the actually selectable sources and how to
select them.
SPIs are not affected by CPU SLOW mode.
19.1.2. Hardware settings
Clock frequency settings and the polarity of the data connec-
tions of the SPIs can be set via HW Options (Table 19–1).
Refer to “HW Options” for setting them.
19.1.3. Initialization
After reset, a SPI is in standby mode (inactive).
Prior to entering active mode, proper SW configuration of the
U-Ports assigned to function as data in- or outputs and clock
in- or outputs has to be made (Table 19–1). Refer to “Ports”
for details.
For entering active mode of a SPI, set the respective enable
bit (Table 19–1).
Prior to operation, the desired clock frequency and telegram
length have to be selected.
19.1.3.1. Clock Source
The SPI can be operated as clock master, using an internally
generated clock, or as clock slave, using an externally gen-
erated clock.
The flag INTERN must be set in the SPIxM Mode register to
operate the SPI as clock master. There are several options
for selection of the internal clock. Each input of a 3 to 1 multi-
plexer can be programmed by HW Options to a different fre-
quency. These three input frequencies F0SPI, F1SPI and
F2SPI are used for all SPIs. The output of the 3 to 1 multi-
plexer is programmed by way of clock selection field (CSF) in
register SPIxM. This clock can be used as shift clock directly
or divided by 1.5 or 2.5. This selection is done by HW Option
too. The shift clock is output by signal SPIx-CLK-OUT which
can be inverted by the flag NEDGE of register SPIxM.
If flag INTERN is zero, the SPI operates as clock slave and
an externally generated clock is used. The external clock is
input by signal SPIx-CLK-IN and can be inverted by the flag
NEDGE.
The data valid edge of the clock is defined by flag NEDGE in
register SPIxM.
19.1.3.2. Telegram Length
Flag LEN9 in register SPIxM defines the length of a trans-
ferred frame. The ninth bit of the shift register is read or writ-
ten at the location of flag BIT8 in register SPIxM.
Table 19–1: Module specific settings
Module
Name HW Options Initialization Enable
Bit
Item Address Item Setting
All SPIs F0SPI
clock
FFAFh
F1SPI
clock
FFB0h
F2SPI
clock
FFB1h
SPI0 D in
inver-
sion
FFAFh SPI0-D-
IN
input
U6.5
special
in
SR0.
SPI0
D out
inver-
sion
FFAFh SPI0-D-
OUT
output
U6.4
special
out
Pres-
caler FFAEh SPI0-
CLK-IN
input
U6.3
special
in
SPI0-
CLK-
OUT
output
U6.3
special
out
SPI1 D in
inver-
sion
FFB0h SPI1-D-
IN
input
U3.1
special
in
SR0.
SPI1
D out
inver-
sion
FFB0h SPI1-D-
OUT
output
U3.0
special
out
Pres-
caler FFAEh SPI1-
CLK-IN
input
U3.5
special
in
SPI1-
CLK-
OUT
output
U3.5
special
out
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 131
19.1.4. Operation
19.1.4.1. Transmit Mode
Transmission is initiated by a write access to data register
SPIxD. The SPI will immediately begin transmitting the
selected number of data bits out from its shift register, in syn-
chronism with the selected clock. A write access during a
transmission is ignored. The frame is transmitted MSB first.
In nine-bit mode flag BIT8 is MSB of the shift register (Fig.
19–2, 19–3). At the end of the frame, an interrupt source sig-
nal is generated which may be selected to trigger an inter-
rupt.
19.1.4.2. Receive Mode
The receive mode must be activated by a write access to
register SPIxD. The SPI will immediately begin clocking in
the selected number of data bits into its shift register, in syn-
chronism with the selected clock. At the end of the frame, an
interrupt source signal is generated which may be selected
to trigger an interrupt.
19.1.5. Inactivation
Returning an SPI module to standby mode by resetting its
respective enable bit (Table 19–1) will immediately terminate
any running receive or transmit operation and will reset all
internal registers.
19.1.6. Precautions
A single-wire bus is easiest implemented by a wired-or con-
figuration of the SPIx-D-OUT output port and the open drain
output of the external transmitter:
simply configure the SPIx-D-OUT output port in Port Slow
mode, always operate it in Port Special Output mode and
connect it directly to the external open drain output. An exter-
nal pull-up resistor is not necessary in this configuration
because the SPIx-D-OUT output port supplies the necessary
pull-up drive.
If the SPIx-D-OUT output port has to be operated in Port
Fast mode, this simple scheme is not possible, because the
pull-down action of the external open drain output may
exceed the absolute maximum current rating of the SPIx-D-
OUT output port. A discrete external wired-or is recom-
mended for this situation.
During operation, please make sure that the external clock
does not start until after SPIxD has been written, otherwise
correct data transfer is not be guaranteed.
Attention must be paid to the neutral level of the external
clock. Neutral level is defined high when data are valid on
the rising clock edge. Neutral level is low otherwise.
19.2. Registers
The following registers are available once for SPI0 and SPI1
each.
BIT8 Bit 8 of Rx/Tx Data
r/w: Rx/Tx data bit.
LEN9 Frame Length 9 Bit Selection
r/w0: 8 bit mode.
r/w1: 9 bit mode.
RXSEL Receive Selection
r/w0: Input active.
r/w1: Low level at input.
INTERN Internal/External Clock Selection
r/w0: Use external clock.
r/w1: Use internal clock.
NEDGE Negative Edge Selection
r/w0: Data valid at rising edge.
r/w1: Data valid at falling edge.
CSF Clock Selection Field
wr: Source of internal clock (Table 19–2)
SPIxD SPI x Data Register
76543210
SPIxM SPI x Mode Register
76543210
r/w Bit 7 to 0 of Rx/Tx Data
00000000Res
r/w BIT8 LEN9 RXSEL INTERN NEDGE x CSF
00000000Res
Table 19–2: CSF usage
Bits
1 0 Source of internal clock
0 0 F0SPI
0 1 F1SPI
1 x F2SPI
CDC16xxF-E ADVANCE INFORMATION
132 March 31, 2003; 6251-606-2AI Micronas
19.3. Timing
Fig. 19–2: Nine bit frame. Data valid at rising edge.
Fig. 19–3: Eight bit frame. Data valid at falling edge.
D7 D6 D5 D4 D3 D2 D1 D0
D8 D7 D6 D5 D4 D3 D2 D1 D0
D8
wr SPIxD
clk out
SPIx-D-OUT
SPIx-D-IN
SPIx Int. Src.
D6 D5 D4 D3 D2 D1 D0
D7 D6 D5 D4 D3 D2 D1 D0
D7
wr SPIxD
clk out
SPIx-D-OUT
SPIx-D-IN
SPIx Int. Src.
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 133
20. Universal Asynchronous Receiver Transmitter (UART)
A UART provides a serial Receiver/Transmitter. A 7-bit or 8-
bit telegram can be transferred asynchronously with or with-
out a parity bit and with one or two stop bits. A 13-bit baud
rate generator allows a wide variety of baud rates. A two-
word receive FIFO unburdens the SW. Incoming telegrams
are compared with a register value. Interrupts can be trig-
gered on transmission complete, reception complete, com-
pare and break.
The number of UARTs implemented is given in Table 20–1.
The “x” in register names distinguishes the module number.
Features
– 7-bit or 8-bit frames.
– Parity: None, odd or even.
– One or two stop bits.
– Receive compare register.
– Two-word receive FIFO.
– 13-bit baud rate generator.
Fig. 20–1: Block Diagram
compare address register
rx FIFO r
1
=
rx shift register 2 of 3
rx control
UAxCAUAxD
5 347 2 01
23 01
tx control
6 457
6
tx tx
tx data register
tx shift register
LEN
ODD
PAR
STPB
2 012 01
>1
8
8
8
&
&
&
UAxIMUAxIF
UART
Interrupt
Source
5 bit down cnt
UAxBR1
8 bit down counter
UAxBR0 1/8
zeroclk zeroclk
RCVD
BRK
ADR
RCVD
BRK
ADR
clk
fBR
fsample
rx
w
3
4
r
w
BRKD
FRER
PAER
OVRR
EMPTY
FULL
TBUSY
4
break
received
UAxD
UAxC
w
rw
RBUSY
CDC16xxF-E ADVANCE INFORMATION
134 March 31, 2003; 6251-606-2AI Micronas
20.1. Principle of Operation
20.1.1. General Remarks
A UART module contains a receive shift register that serves
to receive a telegram via its RX input. A FIFO is affixed to it
that stores two previously received telegrams.
A transmit shift register serves to transmit a telegram via its
TX output.
Other features include a receive compare function, flexible
interrupt generation and handling, and a set of control, error
and status flags that facilitate management of the UART by
SW.
The interrupt source output of this module is routed to the
Interrupt Controller logic. But this does not necessarily select
it as input to the Interrupt Controller. Check section “Interrupt
Controller” for the actually selectable sources and how to
select them.
A programmable baud rate generator generates the required
bit clock frequency.
A UART module is not affected by CPU SLOW mode.
A UART module is only capable of receiving telegrams that
differ by no more than ± 2.5% from its own baud rate setting.
20.1.2. Hardware settings
The polarity of most RX and TX connections of the UART
can be set via HW Options (See table 20–1 and figure 20–2).
Refer to “HW Options” for setting them.
Fig. 20–2: Context Diagram
20.1.3. Initialization
After reset, a UART is in standby mode (inactive).
Prior to entering active mode, the U-Ports assigned to func-
tion as RX input and TX output have to be properly SW con-
figurated (Table 20–1). The RX port has to be configured
Special In and the TX port has to be configured Special Out.
Refer to “Ports” for details.
For entering active mode of a UART, set the respective
enable bit (Table 20–1).
Prior to operation, the desired baud rate, telegram format,
compare address and interrupt source configuration have to
be done.
20.1.3.1. Baud Rate Generator
The receive and transmit baud rate is internally generated.
The Baud Rate registers UAxBR0 (low byte) and UAxBR1
(high byte) serve to enter the desired 13bit setting. Write
UAxBR0 first, UAxBR1 last.
The baud rate generator is a 13-bit down-counter which is
clocked by fOSC. It generates the sample frequency:
Its output frequency fsample is divided by eight to generate
the baud rate (bit/second).
20.1.3.2. Telegram Format
The format of a telegram is configured in the Control and
Status register UAxC. A telegram starts with a start bit fol-
lowed by the data field. The data field consists of 7 or 8 data
bit. There can be a parity bit after the data field. The telegram
is finished by one or two stop bits (see Table 20–3 on
page 138).
1
0
SI
1
UARTx-RX
UARTx 1
0
SO
1
UARTx-TX
tx
rx
HW Option
HW Option
fOSC
SR.UARTx
clk
fsample
fOSC
Value of Baud Rate Registers 1+
--------------------------------------------------------------------------------=
BR fOSC
Value of Baud Rate Registers 1+()8×
---------------------------------------------------------------------------------------------- fsample
8
---------------==
Value of Baud Rate Registers fOSC
BR 8×
-----------------1–=
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 135
Fig. 20–3: Examples of Telegram Formats
The level of the start bit is always opposite to the neutral
level. The level of the stop bits is always the same as the
neutral level. If a parity bit is programmed, odd or even parity
can be selected.
As a general rule, the parity bit completes the number of
ones in the data field to the selected parity.
20.1.3.3. Compare Address
The content of the Compare Address register UAxCA is
compared with each received telegram. If they match, the
interrupt flag ADR is set and the interrupt source signal is
triggered.
The MSB of register UAxCA must be set to zero if transmis-
sion of a seven bit data field is configured in register UAxC.
20.1.3.4. Interrupt
Four signals can trigger the UART interrupt source output.
Three of them set their own flags in the Interrupt Flag regis-
ter UAxIF and can be enabled by setting bits in the Interrupt
Mask register UAxIM.
1. When the flag TBUSY in register UAxC is set to zero, the
interrupt source output is triggered. This indicates that a
transmission is finished and the transmit buffer is empty.
There is neither an interrupt flag to indicate this event, nor a
mask flag to disable this interrupt.
2. RCVD is generated by the receive control logic at the end
of each received telegram even if the FIFO is full. This signal
is enabled by setting the corresponding bit in register UAxIM.
3. BRK is generated by the receive control logic each time a
break is detected. This signal is enabled by setting the corre-
sponding bit in register UAxIM.
4. ADR is generated by the address comparator. This signal
is enabled by setting the corresponding bit in register UAxIM.
BRK and ADR also set flags in the Interrupt Flag register
UAxIF when enabled. The first RCVD interrupt, when the
FIFO has been empty before, sets a flag in UAxIF too. Even
if all interrupts are enabled in register UAxIM, the interrupt
source output is triggered only once within a telegram. UAxIF
flags remain valid until the end of the next telegram. ADR is
not generated and the ADR flag is not set if a frame or parity
error was detected in the corresponding telegram.
20.1.4. Operation
With proper HW configuration and SW initialization, a UART
module is ready to transmit and receive telegrams in the
selected format.
20.1.4.1. Transmit
A write access to UART Data register UAxD immediately
loads the transmit shift register and starts transmission by
sending the start bit. The flag TBUSY in register UAxD is set.
At the end of transmission the interrupt source signal is trig-
gered and the flag TBUSY is reset.
To avoid data corruption, ensure that flag TBUSY is LOW
before writing to UAxD.
Table 20–1: Module-specific settings
Module
Name HW Options Initialization Enable Bit
Item Address Item Setting
UART0 RX inversion FFB4h UART0-RX input U4.4 special in SR1.UART0
TX inversion FFB4h UART0-TX output U4.5 special out
UART1 RX inversion FFB5h UART1-RX input U5.7 special in SR2.UART1
TX inversion FFB5h UART1-TX output U5.4 special out
UART2 RX inversion FFB4h UART2-RX input U4.2 special in SR0.UART2
TX inversion FFB4h UART2-TX output U4.3 special out
Table 20–2: Definition of Parity Bit
Parity Flag Number of Ones Parity Bit
odd odd 0
odd even 1
even odd 1
even even 0
01234567PST0T1
01234567ST0T1
0123456ST0
01234567PST0
S = Start bit
P = Parity bit T0 = 1. stop bit
T1 = 2. stop bit
CDC16xxF-E ADVANCE INFORMATION
136 March 31, 2003; 6251-606-2AI Micronas
20.1.4.2. Receive
A first negative edge of a telegram on the RX line of a UART
starts a receive cycle and sets the flag RBUSY in UAxC.
After reception of the last bit of the telegram, the telegram
content, together with its status information, is transferred to
the receive FIFO and an interrupt is generated. RBUSY is
reset. Telegram data are available in register UAxD, tele-
gram status in register UAxC.
During reception, the following checks are performed
according to the register UAxC setting:
1. A parity error is detected if the parity of the received tele-
gram does not match the programmed parity. The flag PAER
in register UAxC is set in this case. Differing telegram length
settings in register UAxC and receiver may also cause parity
errors.
2. A frame error is detected if the level of start or stop bits
violate the transmission rule. The flag FRER in register
UAxC is set in this case.
3. A break condition is detected if the receive input remains
low for one complete telegram duration. When a break starts
during telegram, this condition must extend over another
telegram length to be properly detected. This event sets the
flag BRKD in register UAxC and can trigger the interrupt
source output if enabled. After a break, the receive input
must be high for at least 1/4 of the bit length before a new
telegram can be received.
Telegrams of an external RS232 interface are correctly
received, even if they are transmitted without gaps (the start
bit immediately follows the stop bit of the preceding tele-
gram).
20.1.4.3. Receive FIFO
The receive FIFO is able to buffer the data fields of two con-
secutive telegrams. But not only the data field of a telegram
is double-buffered, the related information is double-buffered
too. The flags PAER, FRER and BRKD in register UAxC
apply to a certain telegram and are thus double-buffered.
The receive FIFO is full if two telegrams have been received
but the SW has not yet read register UAxD. If there is a third
telegram, it is not written to the FIFO and its data are lost.
The flags EMPTY, FULL and OVRR show the status of the
FIFO. EMPTY indicates that there is no entry in the FIFO.
FULL will be set with the second entry in the receive FIFO
and indicates that there is no more entry free. OVRR indi-
cates that there was a third telegram which could not be writ-
ten to the FIFO.
Status flags are readable as long as the corresponding data
field was not read from register UAxD. As soon as a FIFO
entry is read out, the status flags of this entry are lost. They
are overwritten by the flags of the second entry. SW first has
to read the flags and then the corresponding FIFO entry.
The flags PAER, FRER and BRKD apply to a certain tele-
gram and are only valid if there is at least one entry in the
FIFO (EMPTY = 0). The flags EMPTY, FULL and OVRR
apply to the FIFO and are valid all the time.
20.1.5. Inactivation
Returning a UART module to standby mode by resetting its
respective enable bit (Table 20–1) will immediately terminate
any running receive or transmit operation and will reset all
internal registers.
20.2. Timing
The duration of a telegram results from the total telegram
length in bits (LTG) (see Table 20–3 on page 138) and the
baud rate (BR).
The incoming signal is sampled with the sample frequency
and filtered by a 2 of 3 majority filter. A falling edge at the out-
put of the majority filter starts the receive timing frame for the
telegram. An individual bit is sampled with the fifth sample
clock pulse within that timing frame (cf. Fig. 20–4 and 20–5).
If a bit was the last bit of its telegram, reception of a new tele-
gram can start immediately after this sample. With a receive
telegram, interrupt source is triggered and flags are set just
after the sample of the last stop bit. With a transmit telegram,
interrupt source is triggered and BUSY reset after the nomi-
nal end of the last stop bit.
tTG
LTG
BR
----------=
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 137
Fig. 20–4: Start of Telegram
Fig. 20–5: End of Telegram
1 5432 6 7 8
fsample
sample clock
data
start
1. sample
2. sample
3. sample
rxdat asynchron bit 0 bit 1
startbit
indicates the recognition of the low level of the filtered input signal
fsample
sample clock
data
start
1. sample
2. sample
3. sample
rxdat asynchron
BUSY
Tx Interrupt
Rx Interrupts
1 5432 6 7 8
Flags are set
2. stopbit startbit1. stopbit
CDC16xxF-E ADVANCE INFORMATION
138 March 31, 2003; 6251-606-2AI Micronas
20.3. Registers
RBUSY Receiver Busy
r0: Not busy.
r1: Busy.
BRKD Break Detected
r0: No break.
r1: Break.
FRER Frame Error Detected
r0: No error.
r1: Error.
OVRR Overrun Detected
r0: No overrun.
r1: Overrun.
PAER Parity Error Detected
r0: No error.
r1: Error.
EMPTY Rx FIFO Empty
r0: Not empty.
r1: Empty.
There is at least one entry present if EMPTY is zero. PAER,
FRER and BRKD are not valid if EMPTY is set.
FULL Rx FIFO Full
r0: Not full.
r1: Full.
TBUSY Transmitter Busy
r0: Not busy.
r1: Busy.
Do not write to register UAxD as long as BUSY is true.
STPB Stop Bits
w0: One stop bit.
w1: Two stop bits.
ODD Odd Parity
w0: Even parity.
w1: Odd parity.
PAR Parity On
w0: No parity.
w1: Parity on.
LEN Length of Frame
w0: 7-bit frame.
w1: 8-bit frame.
The Baud Rate Registers UAxBR0 and UAxBR1 have to be
written low byte first to avoid inconsistencies. UAxBR0 is the
low byte.
Valid entries in the Baud Rate Registers range from 1 to
8191. Don’t operate the baud rate generator with its reset
value zero.
UAxD UART x Data Register
76543210
UAxC UART x Control and Status Register
76543210
r Receive register
wTransmit register
xxxxxxxxRes
rRBUSY BRKD FRER OVRR PAER EMPTY FULL TBUSY
0xx0x100Res
wxxxxSTPBODDPARLEN
xxxx0000Res
Table 20–3: Telegram Format and Length
LEN PAR STPB Format LTG
000S, 7D, T0 9
001S, 7D, T0, T110
010S, 7D, P, T0 10
011S, 7D, P, T0, T111
100S, 8D, T0 10
101S, 8D, T0, T111
110S, 8D, P, T0 11
111S, 8D, P, T0, T112
UAxBR0 UART x Baud Rate Register low byte
76543210
UAxBR1 UART x Baud Rate Register high byte
76543210
UAxCA UART x Compare Address Register
76543210
UAxIM UART x Interrupt Mask Register
76543210
wBit 7 to 0 of Baud Rate
00000000Res
wx x x Bit 12 to 8 of Baud Rate
- - -00000Res
wBit 7 to 0 of address
00000000Res
wx x x x x ADR BRK RCVD
-----000Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 139
ADR Mask Compare Address Detected
w0: Disable interrupt.
w1: Enable interrupt.
BRK Mask Break Detected
w0: Disable interrupt.
w1: Enable interrupt.
RCVD Mask Received a Telegram
w0: Disable interrupt.
w1: Enable interrupt.
Test Reserved for test (do not use)
ADR Compare Address Detected
r0: No Interrupt.
r1: Interrupt pending.
BRK Break Detected
r0: No Interrupt.
r1: Interrupt pending.
RCVD Received a Telegram
r0: No Interrupt.
r1: Interrupt pending.
UAxIF UART x Interrupt Flag Register
76543210
rTest Test Test Test Test ADR BRK RCVD
-----x00Res
CDC16xxF-E ADVANCE INFORMATION
140 March 31, 2003; 6251-606-2AI Micronas
21. CAN Manual
This manual describes the user interface of the CAN module.
For further information about the CAN bus, please refer to
the CAN specification 2.0B from Bosch.
Features
– Bus controller according to CAN Licence Specification
1992 2.0B
– Supports standard and extended telegrams
– FullCAN: at least 16 Rx and Tx telegrams
– Variable number of receive buffers
– Programmable acceptance filter
Single, group or all telegrams received.
– Time stamp for each telegram
– Overwrite mode programmable for each telegram
– Programmable baud rate. Max. 1 MBd @ 8 MHz
– Sleep mode
The CAN interface is a VLSI module which enables coupling
to a serial bus in compliance with CAN specification 2.0B. It
controls the receiving and sending of telegrams, searches for
Tx telegrams and interrupts and carries out acceptance filter-
ing. It supports transmission of telegrams with standard (11
bit) and extended (29 bit) addresses.
The CAN interface can be configured as BasicCAN or Full-
CAN. It enables several active receive and transmit tele-
grams and supports the remote transmission request. The
number of telegrams which can be handled depends mainly
on the size of the communication RAM (16 byte per tele-
gram), the system clock and the transmission speed. A max-
imum of 254 telegrams can be handled.
A mask register makes it possible to receive different groups
of telegram addresses with different receive telegrams.
Transmitting or receiving of a telegram as well as the occur-
rence of an error can trigger an interrupt.
Fig. 21–1: Block diagram of the CAN bus interface
Address
Data
Interrupt
CAN
Bus
Protocol
Manager
Interface
Managm.
Logic
Rx/Tx-
Buffer
Global Con-
trol and Sta-
tus Register
Error
Managem.
Logic
Bit Timing
Logic
(Com.
Area)
Tx. Obj.
Rx. Obj.
CAN RAM
CPU
Source
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 141
21.1. Abbreviations
BI CAN Bus Interface
BTL Bit Timing Logic
CAN Controller Area Network
CA Communication Area
CO Communication Object
CM Communication Mode
CRC Cyclic Redundancy Code
DLC Data Length Code
EoCA End of CA
Ext. ID Extended Identifier
Ext. Tg Extended Telegram
GCS Global Control and Status Register
ID Identifier
IML Interface Management Logic
Rx. Obj. Receive Object
RxTg Receive Telegram
Std. ID Standard Identifier
Std. Tg Standard Telegram
TD Telegram Descriptor
Tg Telegram
TQ Time Quantum
Tx. Obj. Transmit Object
TxTg Transmit Telegram
21.2. Functional Description
21.2.1. HW Description
The CAN bus interface consists of the following components:
Bit Timing Logic: Scans the bus and synchronizes the CAN
bus controller to the bus signal.
Protocol Manager: The PM monitors or generates the com-
position of a telegram and performs the arbitration, the CRC
and the bit stuffing. It controls the data flow between Rx/Tx
buffer and CAN bus. It also drives the Error Management
Logic.
Error Management Logic: Adds up the error messages
received from the protocol manager and generates error
messages when particular values are exceeded. Guarantees
the error limitation as per the CAN Spec. V2.0B.
Interface Management Logic: The IML scans the Commu-
nication Area (CA) in the CAN-RAM for transmit telegrams.
As soon as it finds one, it enters it into the Rx/Tx buffer and
reports it to the protocol manager as ready for transmission.
If a telegram is received, the IML carries out the acceptance
filtering, i.e. scans the CA, taking into account the Identifier
Mask Register in the GCS, for a Tg with the appropriate
address. After correct reception, it copies the Tg from the Rx/
Tx buffer to the CA. The IML also reports to the CPU the
valid transfer of a telegram or given errors per interrupt.
The interrupt source output of this module is routed to the
Interrupt Controller logic. But this does not necessarily select
it as input to the Interrupt Controller. Check section “Interrupt
Controller” for the actually selectable sources and how to
select them.
Rx/Tx buffer: This is used to buffer a full telegram (ID, DLC,
data) during sending and receiving.
Global Control and Status Register: The GCS contains
registers for the configuration of the BI. It also contains error
and status flags and an identifier mask. The Error Counter
and the Capture Timer can be read from the GCS.
Receive Object: The BI enters received telegrams into a
matching Rx-Object. It can be retrieved from the application.
Transmit Object: The application enters data into the Tx-
Object and reports it ready for transmission. The BI sends
the telegram as soon as the bus traffic allows.
For the effect of CPU clock modes on the operation of this
module, refer to section “CPU and Clock System” (see
Table 4–2 on page 35).
21.2.2. Memory Map
From the CAN bus interface the user sees two storage areas
in the user RAM area. The BI is configured with the Global
Control and Status Registers (GCS). It also indicates the sta-
tus here. The communication area (CA) contains the Rx and
Tx telegrams.
The communication area lies in the CAN-RAM. The end of
the Com. Area is fixed by the first control byte of an object
whose 3 MSBs contain only ones (Communication Mode = 7
= EoCA). The area after this is available to the user.
The CA consists of communication objects (COs). A CO con-
sists of 6 bytes telegram descriptor (TD), 8 data bytes and
the Time Stamp which is 2 bytes long. The TD contains the
address (ID) and the length of a telegram (DLC) as well as
control bits which are needed for access to the CO and for
the transmission of a telegram.
In the BasicCAN and the FullCAN versions, all the communi-
cation objects have the same, maximum size of 16 byte.
Unassigned storage locations in the data area of a CO can
be freely used.
The maximum number of COs is limited by the time which
the CAN interface has to search for an identifier in the Com.
Area.
CDC16xxF-E ADVANCE INFORMATION
142 March 31, 2003; 6251-606-2AI Micronas
21.2.3. Global Control and Status Registers (GCS)
The GCS registers can be used to determine the behavior of
the CAN interface. As well as flags for the interrupts, halt and
sleep modes, they also contain interrupt index, ID mask, bus
timing, error status, output control registers, baud rate pre-
scalers, Tx and Rx error counters as well as the capture
timer.
Fig. 21–2: Memory allocation
Access modes:
r: read
w: write
i: init (BI halted)
w0: clear
w1: set
HLT Halt
r/w0: Run.
r/w1: Halt.
Switches the CAN interface into the halt mode. Transmis-
sions which have been started are brought to an end. The
halt acknowledge is indicated in the status register (HACK).
Re-initialization can be carried out in the halt mode (HACK is
set). After this, the halt flag must be deleted again. After a
reset, HLT is set.
If HLT is set during a Tx-Tg and this has to be repeated
(error or no acknowledge), the BI stops yet. The correspond-
ing TxCO is still reserved, however, and can no longer be
operated from BI. Therefore, when HLT is set, the CA should
always be re-initialized if the last Tx-Tg has not been cor-
rectly transmitted (Status Transfer Flag is still deleted).
If HLT is set during the BI is in Bus-Off mode, the BI stops
after Bus-Off mode is finished. Flag BOFF is cleared then
and receive and transmit error counters are reset to zero.
SLP Sleep
r/w0: Run.
r/w1: Sleep.
The BI goes into the sleep mode when the sleep flag is set
and a started Tg is terminated. The sleep mode is finished as
0
15
0
1
2
16
15
31
(n+1)*16
32
47
n*16
n*16+15
16
Control
Status
Error Status
Interrupt Index
ID Mask 28 ... 21
ID Mask 20 ... 13
ID Mask 12 ... 5
ID Mask 4 ... 0
Bit Timing 1
Bit Timing 2
Bit Timing 3
Input Control
Output Control
Transmit Error Counter
Receive Error Counter
Capture Timer low
Capture Timer high
Global Control and Status Communication Area
Control
ID 28 ... 21
ID 20 ...13
ID 12 ... 5
ID 4 ... 0 and Control
DLC and Control
Data 0
Data 1
Data 2
Data 3
Data 4
Data 5
Data 6
Data 7
Time Stamp low
Time Stamp high
TD
Data and Time Stamp
TD
Data and Time Stamp
TD
Data and Time Stamp
Control: CM = 7
Com.-Obj. 1
Telegram
Descriptor
TD
Com.-Obj. 2
Com.-Obj. 3
Com.-Obj. n
End of Com. Area
STR
CTR
IDX
ESTR
IDM
BT2
BT1
ICR
BT3
TEC
OCR
CTIM
REC
Error Status Mask
ESM
17
CANxCTR Control Register
76543210
r/w HLT SLP GRSC EIE GRIE GTIE BOST rsvd
1000000xRes
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 143
soon as a dominant bus level is detected, or the sleep flag is
deleted.
GRSC Global Rescan
r0: Don’t rescan.
r/w1: Rescan.
The microprocessor can set this flag in order to initiate a
transmit telegram search at the beginning of the Com. Area.
The BI resets the bit. The BI also sets the GRSC flag if the
flag RSC has been set in a telegram descriptor of a Tx-Tg
just operated, and thereby initiates a rescan. If the micropro-
cessor writes a zero, nothing happens.
EIE Error-Interrupt-Enable
r/w0: Disabled.
r/w1: Enabled.
GRIE Global Rx-Interrupt-Enable
r/w0: Disabled.
r/w1: Enabled.
GTIE Global Tx-Interrupt-Enable
r/w0: Disabled.
r/w1: Enabled.
BOST Bus-Off Stop Select
r/w0: Don’t stop when leaving Bus-Off mode.
r/w1: Stop when leaving Bus-Off mode.
The flag HLT is set by the BI after leaving the Bus-Off recov-
ery sequence. The SW has to restart the CAN module in this
case after re-initialisation. Consider the flag HACK even in
this case.
HACK Halt-Acknowledge
r0: Running.
r1: Halted.
Is set by the BI when it enters the halt mode. It is deleted
again when the halt mode is exited.
BOFF Bus-Off
r0: Bus active.
r1: Bus off.
With this flag the BI indicates whether the node is still
actively participating in the bus. If the transmit error counter
reaches a value of > 255 (overflow), the node is separated
from the bus and the flag is set. The Bus-Off mode is left
after the Bus-Off recovery sequence. The flag CANx-
CTR.BOST defines the behavior after leaving Bus-Off mode.
EPAS Error-Passive
r0: Error active.
r1: Error passive.
With this flag the BI indicates whether the node is still partici-
pating in the bus with active Error Frames. If an error counter
has reached a value > 127, the node only transmits passive
error frames and the flag is set.
ERS Error-Status
r0: No Errors.
r1: Errors.
This flag is set when the BI detects an error and the apropri-
ate error flag is not masked in the error status mask register.
It is set even if an error counter is greater than 96. It means
that a bit has been set in the error status register. As soon as
all the flags in the error status register are either deleted or
masked, ERS is also deleted.
As long as a bit is set in the CANxESTR and not masked, the
ERS bit is also set in the status register. If EIE has been set
in the control register, an interrupt is triggered too; i.e. the
value 254 is entered in the register CANxIDX as soon as it is
free, and the interrupt source output is triggered.
To erase a bit in the CANxESTR the user must write a one at
the appropriate place. Places at which he writes a zero will
not be changed. Because it makes sense to erase only those
bits which have previously been read, only the value which
has been read has to be re-written.
Read-Modify-Write operations on single flags of this register
must be avoided. Unwanted clearing of other flags of this
register may be the result otherwise.
GDM Good Morning
r0: No wake-up.
r1: Wake-up.
w0: Unaffected.
w1: Clear.
Is set by the BI when it is aroused from the sleep mode by a
dominant bus level. The user must delete it.
CTOV Capture Time Overflow
r0: No overflow.
r1: Overflow.
w0: Unaffected.
w1: Clear.
Is set by the BI when the capture timer (CTIM) overflows.
The user must delete it.
ECNT Error Counter Level
r0: No error counter.
r1: Error counter.
w0: Unaffected.
w1: Clear.
Is set by the BI as soon as the transmit error counter or the
receive error counter exceeds a limit value. The user must
delete it.
BIT Bit Error
r0: No bit error.
r1: Bit error.
w0: Unaffected.
w1: Clear.
Is set by the BI when a transmitted bit is not the same as the
bit received. The user must delete the flag.
STF Stuff Error
r0: No stuff error.
r1: Stuff error.
w0: Unaffected.
w1: Clear.
Is set by the BI when 6 identical bits are received succes-
sively in one Tg. The user must delete it.
CRC CRC Error
r0: No stuff error.
r1: Stuff error.
w0: Unaffected.
CANxSTR Status Register
76543210
rHACK BOFF EPAS ERS rsvd rsvd rsvd rsvd
1000xxxxRes
CANxESTR Error Status Register
76543210
r/w GDM CTOV ECNT BIT STF CRC FRM ACK
00000000Res
CDC16xxF-E ADVANCE INFORMATION
144 March 31, 2003; 6251-606-2AI Micronas
w1: Clear.
Is set by the BI when the CRC received does not coincide
with the CRC calculated. The user must delete it.
FRM Form Error
r0: No form error.
r1: Form error.
w0: Unaffected.
w1: Clear.
Is set by the BI when an incorrect bit is received in a field
with specified bit level (start of frame, end of frame, ...). The
user must delete it.
ACK Acknowledge Error
r0: No acknowledge error.
r1: Acknowledge error.
w0: Unaffected.
w1: Clear.
Is set by the BI when there is no acknowledge for a transmit-
ted Tg. The user must delete it.
The interrupt index indicates the source of the interrupt. If a
transmission has been the cause of an interrupt, the interrupt
index points to the corresponding telegram descriptor
(CANxIDX = 0..253). If an error has been responsible for the
interrupt, the interrupt index designates the error status reg-
ister (CANxIDX = 254). After dealing with the interrupt, the
user must eliminate the cause of the interrupt and set the
interrupt index to minus one (255 = EMPTY). As soon as
CANxIDX is empty, the BI can enter a new index and initiate
an interrupt. An interrupt can only be initiated when CANx-
IDX contains the value 255.
r/w0: Don’t care.
r/w1: Compare.
The identifier mask register is 29 bits long; the MSB is in the
MSB position in the lowest byte address. The CANxIDM
defines a mask for the acceptance of address groups. Only
the permitted bits are used for comparison with a received
identifier. Whether the mask is used can be determined indi-
vidually for each receive object.
MSAM Multi Sample
r/w0: Bus level is determined only once per bit.
r/w1: Bus level is determined three times per bit.
SYN Sync On
r/w0: Synchronization with falling edges only.
r/w1: Synchronization with rising edges too.
BPR Baud Rate Pre-scaler
r/w: Pre-scaler value.
The baud rate pre-scaler sets the length of a time quantum
for the bit timing logic.
tQ = tXTAL x (BPR + 1).
With the 6-bit counter it is possible to extend tQ by a factor of
1...64. Values from 0 to 63 are allowed.
0: tQ = tXTAL
1: tQ = tXTAL x 2
2: tQ = tXTAL x 3
3: tQ = tXTAL x 4 etc.
TSEG2 Time Segment 2
r/i: TSEG2 value.
TSEG2 determines the number of time quanta after the sam-
ple point. Permitted entries: 1...7 (result in 2...8 TQ).
TSEG1 Time Segment 1
r/i: TSEG1 value.
TSEG1 determines the number of time quanta before the
sample point. Permitted entries: 2...15 (result in 3...16 TQ).
SJW Synchronization Jump Width
r/i: SJW value.
SJW defines by how many TQs a bit may be lengthened or
shortened because of resynchronization. Permitted entries:
1...4 (result in 1...4 TQ). Values greater than 4 must not be
used.
CANxIDX Interrupt Index Register
76543210
CANxIDM Identifier Mask Register
76543210
r/w Interrupt Index
11111111Res
r/w Identifier Mask Bits 29 to 21 low
r/w Identifier Mask Bits 20 to 13
r/w Identifier Mask Bits 12 to 5
r/w Identifier Mask Bits 4 to 0 x x x high
00000000Res
CANxBT1 Bit Timing Register 1
76543210
CANxBT2 Bit Timing Register 2
76543210
CANxBT3 Bit Timing Register 3
76543210
r/w MSAM SYN BPR BPR BPR BPR BPR BPR
00000000Res
r/w rsvd TSEG2 TSEG2 TSEG2 TSEG1 TSEG1 TSEG1 TSEG1
00000000Res
r/w rsvd rsvd rsvd rsvd rsvd SJW SJW SJW
xxxxx000Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 145
XREF External Reference
r/w0: The internal reference is used.
r/w1: The external reference is used where avail-
able.
REF1 Use Reference for RxD1
r/w0: RxD is used as inverted input signal.
r/w1: Supply voltage is used as inverted input sig-
nal.
REF0 Use Reference for RxD0
r/w0: RxD is used as input signal.
r/w1: Ground is used as input signal.
ITX Inverted transmission
r/w0: Tx output is not inverted.
r/w1: output is inverted.
Every flag of the CANxESTR can be enabled/disabled gen-
erating an interrupt by modifying the corresponding flag in
register CANxESM.
EGDM Enable Good Morning
r/w0: Disable.
r/w1: Enable.
ECTV Enable Capture Time Overflow
r/w0: Disable.
r/w1: Enable.
EECT Enable Error Counter Level
r/w0: Disable.
r/w1: Enable.
EBIT Enable Bit Error
r/w0: Disable.
r/w1: Enable.
ESTF Enable Stuff Error
r/w0: Disable.
r/w1: Enable.
ECRC Enable CRC Error
r/w0: Disable.
r/w1: Enable.
EFRM Enable Form Error
r/w0: Disable.
r/w1: Enable.
EACK Enable Acknowledge Error
r/w0: Disable.
r/w1: Enable.
The Capture Timer is incremented with a clock pulse derived
from the CAN bus. As it can only be read byte-wise, the low
byte must be read first. The corresponding high byte is
latched at the same time. When CANxCTIM overflows, the
flag CTOV in the error status register is set. The Capture
Timer will not be incremented during CAN module sleep
mode (SLP = 1).
CANxICR Input Control Register
76543210
CANxOCR Output Control Register
76543210
CANxTEC Transmit Error Counter
76543210
CANxREC Receive Error Counter
76543210
CANxESM Error Status Mask Register
76543210
r/w rsvd rsvd rsvd rsvd rsvd XREF REF1 REF0
xxxxx000Res
r/w rsvd rsvd rsvd rsvd rsvd rsvd rsvd ITX
xxxxxxx0Res
rCounter Bit 7 to 0
00000000Res
rx Counter Bit 6 to 0
x0000000Res
r/w EGDM ECTV EECT EBIT ESTF ECRC EFRM EACK
11111111Res
CANxCTIM Capture Timer
76543210
rTimer Bit 7 to 0 low
rTimer Bit 15 to 8 high
00000000Res
CDC16xxF-E ADVANCE INFORMATION
146 March 31, 2003; 6251-606-2AI Micronas
21.2.4. Communication Area (CA)
The CA is located in the CAN-RAM. It consists of com.
objects each of which is 16 bytes long. The CA begins at
address 0 of the CAN-RAM with the first byte of a CO. It
ends with the first byte of a CO which contains ones in its 3
MSBs (communication mode = 7 = EoCA). The following
bytes can be used by the application. If the CAN-RAM is
filled completely with COs, there is no place left and no need
to mark the end of CA.
Every telegram which this node is to receive or transmit, is
represented by a CO. As well as the data and the time
stamp, this also contains a header, the telegram descriptor
(TD), in which the attributes of the communication object are
stored.
The COs are entered into the CA in order of priority . This
starts with the highest priority (the lowest identifier). The
identifier defines the priority of a Tg. If the first eleven bits of
an ext. Tg are the same as the identifier of a std. Tg, the Tg
with standard identifier has higher priority.
21.2.4.1. Telegram Descriptor (TD)
The telegram descriptor is 6 bytes (TD0 to TD5) long and
forms the beginning of a CO. Telegrams with std. and ext.
identifiers have different TDs. They differ only in the length of
the identifiers. 18 bits are therefore not allocated in the TD of
a std. Tg. They cannot be used by the application because
they are overwritten by the reception of a Tg.
Fig. 21–3: Extended and Standard TD Map
Forms of access:
r: read
w: write
i: init (BI halted or CM = inactive)
w0: clear
w1: set
CM Communication Mode
r/i: Mode.
CM defines the type of telegram.
0: Inactive Inactive. No participation in the bus traffic.
1: Send Send data.
2: Receive Receive data.
3: Fetch Fetch data via remote frame.
4: Provide Have data fetched via remote frame.
5: Rx-All Receive every telegram.
6: rsvd Don’t use (provis. EoCA).
7: EoCA End of Communication Area.
As long as the CO is inactive (CM = 0) or locked (LCK =
TRUE), the BI accesses the first byte of the CO only by read-
ing. All other bytes are neither read nor written. The inactive
mode is suitable therefore for re-configuration of a CO on-
line; i.e. while the node is taking part in the bus traffic.
RSC Rescan
r/w0: Don’t rescan.
r/w1: Rescan.
If the rescan bit has been set in a transmit object just pro-
cessed, the search for active Tx objects is started at the
beginning of the communication area. Otherwise, the search
continues at this transmit object until the end of the CA is
reached. From there, the system jumps back to the begin-
ning of the CA.
MID Mask Identifier
r/w0: Don’t mask.
r/w1: Mask.
If MID has been deleted, the identifier received is compared
bit-by-bit with the identifier from the telegram descriptor, i.e.
the entire identifier must be the same so that the telegram
received is transferred into this CO. If MID has been set, only
bits which are allowed in the ID mask register of the GCS are
used for the comparison.
OW Overwrite
r/w0: Don’t overwrite.
r/w1: Overwrite.
When OW is set, the com. object may be overwritten even if
the application has not yet fetched the contents (TS set). The
BI must of course obtain right of access (LCK deleted).
LCK Lock
r/w0: BI has right of access.
r/w1: BI does not has right of access.
Lock determines the right of access for the BI.
ID Identifier
r/i: Identifier.
TS
ACC
5
13
CM
EXF
TIE RIE
2128
20
12
04
ID
ID
ID
ID
0
1
2
3
4
5
Extended Addr. Format (EXF is set)
DLC
RSR
SR
MIDRSC OW LCKrsvd
EXF
TIE RIE
2128
18
ID
ID
0
1
2
3
4
5
Standard Addr. Format (EXF is deleted)
DLC SR TS
20 don’t use
ACCRSR
don’t use
don’t use
7123456071234560
CM MIDRSC OW LCKrsvd
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 147
The ID contains the address of the telegram. 11 bits in the
standard mode or 29 bits in the extended mode.
ACC Access
r/w0: CPU does not have right of access.
r/w1: CPU has right of access.
Access determines the right of access for the CPU. The CPU
should not modify this flag after initialization. In operation
mode only the BI modifies it and the CPU reads it.
RSR Remote Send Request
r/w0: Remote telegram received.
r/w1: Corresponding data transmitted.
In the provide mode, RSR signals a send request from out-
side; in the fetch mode it means that a remote Tg is being
sent. It is set by the BI if a remote telegram has been
received. It is deleted as soon as the corresponding data
telegram has been transmitted.
EXF Extended Format
r/w0: Standard.
r/w1: Extended.
In order to send/receive telegrams with extended address
format, this flag must be switched on. For standard tele-
grams it is deleted.
DLC Data Length Code
r/w: Data length.
The DLC defines the number of data bytes transmitted. Only
telegrams with 0 to max. 8 data bytes are transmitted. If the
DLC of a TxTg contains a value >8, the entered DLC and
exactly 8 bytes will be transmitted. In the case of RxTgs the
received DLC, and therefore also values > 8 will be entered
by BI.
TIE Tx Interrupt Enable
r/w0: Disable.
r/w1: Enable.
Masks the Tx interrupt for this com. object.
RIE Rx Interrupt Enable
r/w0: Disable.
r/w1: Enable.
Masks the Rx interrupt for this com. object.
SR Send Request
r0: Successful transmission.
r/w1: Send request.
With SR, the microprocessor issues a send request. Both the
microprocessor and the BI write the SR flag. If the micropro-
cessor writes a one, the telegram is sent. The BI deletes the
SR flag after successful transmission.
TS Transfer Status
r/w0: Ready for Transfer.
r/w1: Successful transfer.
The TS flag is set by BI after a successful transfer and is
deleted by the microprocessor after a com. object has been
processed.
21.2.4.2. Data Field
The data field consists of 8 Byte. They are filled with tele-
gram data according to the DLC. Unused data bytes (DLC
less than 8) can be used by the user.
21.2.4.3. Time Stamp
TIMST Time Stamp
r: Counter value.
The last two bytes in the CO are used for the time stamp.
At each SoF (Start of Frame) the free-running 16-bit counter
CANxCTIM is loaded into a register. When the Tg has been
correctly transmitted, this register is copied to the two time
stamp bytes of the corresponding CO.
Fig. 21–4: Time stamp
21.3. Application Notes
21.3.1. Initialization
After reset, a CAN Module is in standby mode (inactive).
Prior to entering active mode, proper SW configuration of the
U-Ports assigned to function as RX input and TX output has
to be made (Table 21–1). The RX port has to be configured
Special In and the TX port has to be configured Special Out.
Refer to “Ports” for details.
For entering active mode of a CAN, set the respective enable
bit (Table 21–1).
In the initialization phase, a configuration of the CAN node
takes place. The mode of operation of the BTL and the bus
coupling is set. The communication area is created in the
CAN-RAM. The different telegrams are specified in it.
The CAN node must be halted (HACK = TRUE) to carry out
the initialization. After a reset, the flags HLT and HACK are
set and initialization can take place. If initialization is required
on-line, the flag HLT must be set. However, the BI must ter-
15
14
Data 5
Data 6
Data 7
Time Stamp low
Time Stamp high
Table 21–1: Module-specific settings
Module
Name Initialization Enable Bit
Item Setting
CAN0 CAN0-RX input U6.6 special in SR0.CAN0
CAN0-TX output U6.7 special out
CAN1 CAN1-RX input U1.6 special in SR3.CAN1
CAN1-TX output U1.7 special out
CAN2 CAN2-RX input U4.0 special in SR3.CAN2
CAN2-TX output U4.1 special out
CDC16xxF-E ADVANCE INFORMATION
148 March 31, 2003; 6251-606-2AI Micronas
minate any current transmission before it comes to a halt.
For the user this means that he must wait until HACK has
been set. If HLT is deleted after initialization, then BI begins
to participate in the bus traffic and to scan the CA for tasks.
During initialization, the error status register (CANxESTR)
and the interrupt index (CANxIDX) should be deleted, other-
wise no interrupts can be initiated. The error status mask
register default value after reset is not masked.
If telegrams with different identifiers are to be received in a
single CO, the identifier mask register must be initialized.
This defines which bit of the ID received must be the same
as the ID in the CO.
Bit timing registers 1, 2 and 3 and the output control registers
1 and 2 must be initialized in all cases.
The CA must be created in the CAN-RAM. The different COs
are created one after the other starting at the address 0. It is
important at this point that the three MSBs have been set in
the first byte after the last CO, i.e. at an address divisible by
16 (CM = End of CA). This is not necessary if the CAN-RAM
is completely filled with COs.
Communication mode (CM), identifier, data length code,
extended format flag (EXF) and remote send request flag
must be initialized in each CO. Lock flag (LCK) must be
deleted and access flag (ACC) must be set, in order that the
BI may also view this CO. Transfer status flag (TS) must be
deleted so that interrupts are not initiated erroneously.
21.3.2. Handling the COs
21.3.2.1. Principles
If the user wishes to access a CO, then he must lock out the
BI from access to it. Also the BI reserves access for itself to
one CO. In this case the user may not have access. When
scanning the CA, the BI ignores inactive or locked COs; i.e. it
reads only the first byte and then jumps to the next CO.
Reservations Procedure
If the user wants to access a com. object, he must first set
LCK. Then he must read ACC. If it is TRUE, he has right of
access. After the operation he must delete LCK.
LCK = TRUE;
if (ACC == TRUE)
{
/* CPU has right of access */
}
LCK = FALSE;
_______________ or _________________
LCK = TRUE;
while (ACC == FALSE)
{
/* wait until BI is ready */
}
/* CPU has right of access */
LCK = FALSE;
Fig. 21–5: Access to a CO by the user
When the BI is accessing a com. object, it first deletes ACC
and then reads LCK. If LCK is FALSE, it has right of access.
ACC = FALSE;
if (LCK == FALSE)
{
/* BI has right of access */
}
ACC = TRUE;
Fig. 21–6: Access to a CO by the BI
The BI does not wait at a CO until it becomes free.
The BI scans the CA from beginning to end. After a TxTg has
been transmitted, the next TxTg entered is reported ready to
send.
It makes sense to enter the COs in the CA in order of their
priority. The priority is determined by the ID. The lowest ID
has the highest priority. If the first bits of an extended ID are
identical with a standard ID, the standard ID has higher prior-
ity. The CO with the highest priority is at the beginning of the
CA. This ensures that Tx-Tgs with high priority are transmit-
ted first when a rescan is initiated.
21.3.2.2. Configuration
A CO may be configured only in the inactive and/or locked
mode or when HACK has been set. Otherwise it can lead to
access conflicts between the user and BI.
The communication mode (CM) is determined in the configu-
ration phase. The identifiers are also entered. The flag EXF
must not be overlooked. The flag RSR and DLC determine
whether and how many data bytes will be transmitted in the
telegram. The interrupts can be permitted. In case of a
receive telegram it is necessary under certain circumstances
to set the flags MID and OW. In case of a transmit telegram,
the flag RSC must be adjusted.
21.3.2.3. Transmit Telegram
CM = Send
A transmit telegram is used to send data. How many data
bytes will be sent is fixed in the DLC. The data is entered
directly after the TD. Unused data bytes can be freely used
by the user. If after the transmission of this telegram the user
would like the next Tx-Tg in the CA to be sent, he deletes the
RSC flag. If he sets the RSC, then the transmit search starts
again at the beginning of the CA. The RSR flag has to be
deleted.
The set SR flag tells BI that this telegram is to be sent; SR
can be likened to a postage stamp. The TS flag must be
deleted before the CO is released with the deletion of LCK.
If the BI finds a CO whose SR flag has been set, it reserves
this (ACC = FALSE) and reports it ”ready to send”. It will be
transmitted as soon as no higher-priority telegrams occupy
the bus. After successful transmission, it deletes the flag SR
and sets TS. The setting of ACC re-releases the CO.
Whether an interrupt will be triggered depends on whether
CANxIDX in the GCS contains the value minus one (255)
and transmit interrupts are permitted.
The user should now reserve the CO, reset the flag TS and
delete CANxIDX so that other interrupts can also be
reported. Should he wish to send further data, he can now
enter this.
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 149
21.3.2.4. Receive Telegram
CM = Receive
With a receive telegram, data is received. If the EXF flag and
the unmasked bits of the identifier of a received telegram are
the same as those of a receive CO, the telegram will be cop-
ied to the CO. ID, DLC and data bytes are overwritten by the
received ID, DLC and data. Only as many data bytes as the
received DLC specify will be overwritten (max. 8). The DLC
actually received will be entered. A permitted receive CO is
only used when TS has been deleted or OW has been set.
Once a telegram has been received and copied to a CO, the
flag TS is set. An interrupt will also be initiated if receive
interrupts are permitted and CANxIDX contains the value
minus one (255).
If the user detects the reception of a telegram (TS set), he
must reserve the CO. Then he can read the data and, before
releasing the CO again, delete TS.
21.3.2.5. Receive All Telegrams
CM = Rx-All
If, while searching for an RX-CO, the BI comes across a free
Rx-All-CO, the received telegram will be entered here with-
out regard to ID and EXF.
Rx-All-COs should be applied at the end of the CA.
21.3.2.6. Fetch Telegram
CM = Fetch
A fetch CO is used to request data from another node. This
is done by sending a telegram with the identifier of the
desired data. The remote transmission request flag is set in
this Tg. No data is therefore sent with it. If another node has
the desired data available, this is transmitted with the same
ID as soon as bus traffic allows.
In this mode, only the reception of the data telegram can trig-
ger an interrupt.
The sequence of a fetch cycle is represented for the user in
pseudo-code.
if (TS == FALSE && SR == FALSE) /* CO is empty */
{
LCK = TRUE; /* claim CO */
/* wait until BI released this CO */
while (ACC == FALSE) {/* do anything else */}
SR = TRUE; /* send this Tg */
TS = FALSE;
LCK = FALSE; /* release CO */
}
The BI now transmits the telegram with the RTR flag set. The
other node receives the Tg, provides the data and returns
the telegram with RTR flag deleted. After the reply telegram
has been received, the BI sets the flag TS. The user waits for
the data.
/* wait for answer */
while (TS == FALSE) {/* do anything else */}
LCK = TRUE; /* claim CO */
/* wait until BI released this CO */
while (ACC == FALSE) {/* do anything else */}
/* copy data */
TS = FALSE;
LCK = FALSE; /* release CO */
Instead of waiting for the answer, it is also possible for notifi-
cation to be given by a receive interrupt.
21.3.2.7. Provide Telegram
CM = Provide
A provide CO is used to prepare data for fetching. It is the
counterpart of a fetch CO. In a provide CO the RSR flag is
cleared. It will be set and deleted by the BI. The data can be
prepared in two ways:
In the first case, the user does not become active until a
remote frame has been received (Rx interrupt or polling from
RSR). After the CO has then been reserved, the data is writ-
ten, the SR flag is set and the CO is released. The BI then
ensures that the data is transferred back.
In the second case, the data has already been entered, SR
has been set and TS deleted before the request. When the
remote frame is received, the user does not need to become
active. Also, no Rx interrupt will be initiated. The data is sim-
ply fetched. In this case the requesting RTR telegram must
contain the correct DLC because, with an RTR telegram too,
a received DLC overwrites the local DLC.
In both cases a Tx interrupt can occur after the data telegram
has been transmitted.
21.3.2.8. Data Length Code
The data length code is 4 bits long. It can therefore contain
values between 0 and 15. In principle, no more than 8 bytes
can be transmitted. Empty data telegrams (DLC = 0) are also
possible.
If a telegram with a DLC greater than 8 is received, this value
will be written into the DLC of the CO, but exactly 8 bytes of
data will be copied.
If the DLC of a Tx-CO contains a value greater than 8, this
DLC will be transmitted, but only 8 bytes of data.
21.3.2.9. Overwrite Mode
The BI normally processes a CO only when the transfer sta-
tus TS has been deleted; i.e. the user has processed the CO
since the last transmission. In the case of COs with which
telegrams are received, the TS flag can be by-passed. If
overwrite (OW) is permitted, the BI may overwrite a previ-
ously received telegram. When accessing data therefore, the
user always receives the most up-to-date data.
21.3.3. Interrupts
All interrupts are enabled or disabled by the global interrupt
enable flags, GTIE for Tx interrupts, GRIE for Rx interrupts
and EIE for error interrupts in the CANxCTR register. Each
error interrupt can also be masked individually in the Error
Status Mask register. A Tx interrupt can be enabled in the
corresponding CO with the Tx interrupt enable flag TIE. An
Rx interrupt can be enabled in the corresponding CO with
the Rx interrupt enable flag RIE.
An interrupt can only be initiated when the interrupt index
CANxIDX is empty (minus one). To initiate an interrupt, the
BI enters the number (0...253) of the appropriate CO in the
CDC16xxF-E ADVANCE INFORMATION
150 March 31, 2003; 6251-606-2AI Micronas
CANxIDX. When an error interrupt is involved, the number
254 is entered.
The BI attempts to initiate an interrupt immediately after suc-
cessful transfer. If this does not work (CANxIDX not empty),
the interrupt is pending (also error interrupt).
The BI permanently scans the CA. If, while doing so, it finds
a CO whose interrupt condition is satisfied (e.g. TIE and TS
are set), it generates an interrupt. This means that interrupts
not yet reported will not be reported in the sequence of their
occurrence, but in the sequence in which they are discov-
ered later.
The interrupt service routine of the user must read the CANx-
IDX. The interrupt source is stored here. If CANxIDX points
to a CO (0...253), the user must reserve this. After this, he
must first delete TS so that this CO does not initiate an inter-
rupt again. Only then he may release CANxIDX (CANxIDX =
255) so that the BI can enter further interrupts.
21.3.4. Rescan
The normal transmit strategy searches for the next transmit
CO in the CA. If all the transmit COs are ready to send, they
are processed one after the other. This is a democratic strat-
egy.
If higher-priority TxTgs are reported in the meantime, these
are not processed until the complete list has been finished.
With rescan, the search for Tx telegrams is started again at
the beginning of the CA. By this means the user can force
the normal strategy to be interrupted and a search to be
made first of all for higher-priority TxTgs. A transmit CO
already reported will of course be transmitted first.
The rescan requirement can be achieved dynamically, when
a transmit CO is reported, by setting the global rescan flag
GRSC.
It is also possible to configure a rescan strategy statically.
Each Tx-CO has the rescan flag RSC. If it is set, the system
starts from the beginning with the transmit search after this
CO has been processed. It is possible, for instance, to set
RSC in the low-priority Tx-COs. Each time a low-priority Tx-
CO has been handled, the search continues for higher-prior-
ity objects.
The user must ensure that each Tx-CO is processed.
21.3.5. Time Stamp
The time stamp of a CO shows the user how much time has
elapsed since the transmission of the object. For this pur-
pose, he compares the time stamp with the capture timer
CANxCTIM. Because the time stamp contains the value of
the CANxCTIM at the time of the start of transmission, the
difference is proportional to the time which has elapsed.
The time stamp mechanism also enables network-wide syn-
chronization. A master transmits a Tg. All nodes note the
transmission time (local time). Then the master transmits its
own (global) transmission time. The difference between local
and global time shows by how much one’s own clock (timer)
is wrong.
21.3.6. Errors
In the error status register (CANxESTR) error messages and
status data are collected which can generate an error inter-
rupt. As long as a flag is set in the CANxESTR and not
masked in the CANxESM, the flag ERS is also set in the sta-
tus register. This means that the value 254 is written in
CANxIDX and an interrupt is generated when EIE has been
set.
An error interrupt is deleted by first deleting CANxESTR and
then releasing CANxIDX.
The 5 flags BIT, STF, CRC, FRM and ACK originate from the
protocol manager. The flag GDM (Good Morning) is not an
error flag. GDM is set when the BI is aroused from the sleep
mode by a dominant bus level.
The flag ECNT (error counter level) indicates that an error
counter has exceeded a limit value. It is set when the trans-
mit error counter exceeds the values 95, 127 and 255 or the
receive error counter exceeds the values 95 and 127.
When the BI is in the Bus-Off mode, it no longer actively par-
ticipates in the bus traffic. Nor does it receive telegrams, but
continues to observe the bus. As soon as the BI has
detected 128 x 11 successive recessive bits, it either reverts
to the error-active mode if flag BOST is zero, or it sets the
flag HLT and enters the HALT mode if flag BOST is set. At
the same time the error counters are cleared.
A Bus-Off sequence triggers two interrupts, if the error inter-
rupt is enabled. The first interrupt (ECNT=TRUE) indicates
that the transmit error counter has exceeded the value 255.
This means that the module is in the Bus-Off mode now
(BOFF=TRUE). The receive error counter is used to count
the reception of 128 x 11 successive recessive bits in the
Bus-Off mode. This is the reason for the second interrupt
(ECNT=TRUE), which indicates that the receive error
counter has exceeded the value 95 (warning level). The sec-
ond interrupt can be ignored in Bus-Off mode. The error
interrupt can be disabled during Bus-Off mode to avoid this
second interrupt.
21.3.7. Layout of the CA
The CA contains all COs beginning with the lowest identifier.
The three MSBs must be set in the byte after the last CO
(End of CA).
If the BI has received an identifier complete, it starts at the
beginning of the CA with the search for an appropriate Rx-
CO. If a rescan is initiated, the BI also starts from the begin-
ning with the transmit search.
21.3.7.1. Buffers
Several successive receive COs may be allocated with the
same identifier. The BI stores a received Tg in the first free
Rx-CO. Using this mechanism it is possible to construct a
receive buffer. If RIE is set in the last CO, the CPU is not
informed until the buffer is full.
21.3.7.2. Basic/Full CAN
For a Basic CAN application, a single Tx-CO will be used. All
outgoing telegrams will be transmitted with this. The user
must receive all Rx-Tgs and must himself decide whether he
needs it (acceptance filtering). For this case it is possible to
use an Rx-All-CO. But it is necessary to ensure that this can
be processed before the next Tg arrives.
For this reason, it is a good idea to employ 2 or 3 Rx-All-COs
as buffers after the Tx-CO.
In the case of a FullCAN application, one uses the built-in
acceptance filtering and sets up a CO specifically for each
desired Rx-Tg and Tx-Tg.
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 151
If the CAN-RAM is not big enough, mixed strategies are also
possible. The acceptance filtering, of course, burdens the
CPU with communication tasks.
Fig. 21–7: Example: CA of a BasicCAN with 2 Rx-buffers
Fig. 21–8: Example: CA of a FullCAN with 2 Rx-objects, 2
Tx-objects, and 2 Rx-buffers
Fig. 21–9: Example: CA of a BasicCAN with 4 Rx-buffers
21.3.7.3. Bus Monitor
With some Rx-All-COs it is possible to construct a user-
friendly bus monitor. The CPU has merely to observe
whether anything has been received. The contents of the CO
must be stored. The transmission time can be calculated
from the time stamp.
21.3.7.4. Maximum number of COs
The maximum number of COs depends on the size of the
CAN-RAM, the baud rate, the system clock, the BI and the
CPU accesses to the CAN-RAM.
– The BI can handle a maximum of 254 objects. The limiting
factor is the 8-bit register CANxIDX in the GCS. CANxIDX
can contain 256 different values. The values 255 (empty)
and 254 (error) are reserved. The remaining values
0...253 can indicate 254 objects.
– The maximum number of COs is, of course, limited to a
greater extent by the size of the CAN-RAM. The BI can
only access the CAN-RAM. Therefore the CA can only be
applied there.
16 bytes are reserved for each CO. One extra byte for
coding EoCA after the last CO must not be forgotten. The
CAN-RAM area after the EoCA is freely available to the
user. No EoCA is necessary if the CAN-RAM is filled com-
pletely with COs.
TD:CM = 7
Tx-Obj
TD:
CM=Send
Rx-Obj
TD:
CM=
Rec. All
Rx-Obj
TD:
CM=
Rec. All
End of Com. Area
TD:
CM=Send
TD:CM = 7
TD:
CM=
Rec. All
TD:
CM=
Rec. All
TD:
CM=
Receive
TD:
CM=Send
TD:
CM=
Receive
Tx-Obj
End of Com. Area
Rx-Obj
Rx-Obj
Rx-Obj
Tx-Obj
Rx-Obj
TD:
CM=Send
TD:
CM=
Rec. All
TD:
CM=
Rec. All
TD:CM = 7
TD:
CM=
Rec. All
TD:
CM=
Rec. All
Tx-Obj
Rx-Obj
Rx-Obj
End of Com. Area
Rx-Obj
Rx-Obj
CDC16xxF-E ADVANCE INFORMATION
152 March 31, 2003; 6251-606-2AI Micronas
There is a maximum number of 16 COs possible in a
CAN-RAM of 256 bytes.
– The next limiting factor can be calculated from the baud
rate and system clock. After the BI has received an identi-
fier, it must be possible for it to scan the entire CA before
the telegram comes to an end.
tCA Scan is the time from having received an ID to the end
of a minimum telegram (11 bit ID, no data), which is at the
BI’s disposal to scan the CA.
tCO Scan is the worst case time needed by the BI to pro-
cess an object (A value of 6 I/O cycles is a more realistic
size than 9).
With an input frequency of 8 MHz and a baud rate (1/tbit)
of 1 MBd, the BI could handle 24 COs. Naturally, this
value needs to be rounded off.
– The value thus calculated is further limited, however, by
the CPU accesses to the CAN-RAM. Each I/O cycle
required by the CPU to write or read data in the CAN-
RAM is missing from the BI. The BI is halted by CPU
accesses. This reduces the time which the BI has to scan
the CA. Where there is a reduced CPU clock, in particular,
the user should have only limited access to the CAN-
RAM.
ZBI is the number of BI cycles in the total cycles (ZG), over a
relatively long period (mean value). KBI therefore represents
a correction factor.
Example for an 8-bit CPU:
The Load and Store-Accu commands require 4 cycles.
OP code Adr.L Adr.H DB
Of the 4 cycles, only the last occupies the CAN-RAM. If a
block move without loop is programmed,
LDA 600;
STA 680;
LDA 601;
STA 681;
etc.
then only 3 of 4 cycles remain for the BI, i.e. 75%. KBI would
then be 0.75. The maximum number of COs is then 18. This
applies only when source and destination lie in the CAN-
RAM. If one of the two lies outside, then KBI is 0.875.
If, of course, a loop is programmed,
LDA 600,X ZBI = 3 of 4
STA 680,X ZBI = 3 of 4
DEX ZBI = 2 of 2
BNE NXT ZBI = 2 of 2
10 of 12 cycles are available to the BI. This gives a KBI
of 0.833. The BI can then handle 20 COs. If source or desti-
nation are not in the CAN-RAM, there are as many as 22.
Example for an 16-bit CPU:
The Load and Store-Accu commands require 5 cycles.
OP code Adr.L Adr.H DBL DBH
Of the 5 cycles, the last two occupy the CAN-RAM. In the
worst case, 3 of 5 cycles remain for the BI, i.e. 60%. KBI
would then be 0.6. The maximum number of COs is then 14.
21.4. Bit Timing Logic
In the bit timing logic the transmission speed (baud rate) and
the sample point within one bit will be configured. By shifting
the sample point it is possible to take account of the signal
propagation delay in different buses. Furthermore, the nature
of the sampling and the bit synchronization can also be
defined.
21.4.1. Baud Rate Pre-scaler
The baud rate pre-scaler is a 6-bit counter. It divides the sys-
tem clock down by the factor 1...64. The output is the clock
for the bit timing logic. This clock TQCLK defines the time
quantum (tQ). The time quantum is the smallest time unit into
which a bit is subdivided.
21.4.2. Bit Timing
A bit duration consists of a programmable number of TQCLK
cycles. The cycles are split up into the segments SYNCSEG,
TSEG1 and TSEG2.
21.4.2.1. Bit Timing Definition
Sync.Seg.
It is expected that a bit will begin in the synchronization seg-
ment. If the bit level changes, the resynchronization ensures
that the edge lies inside this segment. The sync.seg is
always one time quantum long.
Prop.Seg.
This part of a bit is necessary to compensate for delay times
Max. Number CO CAN RAM Size
16
-----------------------------------------=
tBit ( 3 + TSEG1 + TSEG2 ) tQ
=
tCA SCAN 28 tBit
=
Max. Number CO tCA SCAN
tCO SCAN
-----------------------=
tQBPR 1+() tXTAL
=
tCO SCAN 9 tXTAL
=
Max. Number CO KBI tCA SCAN
tCO SCAN
----------------------------------=
KBI
ZBI
ZG
--------=
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 153
of the network. It is twice the sum of the signal propagation
delay on the bus plus input comparator delay plus output
driver delay.
Phase Seg.
Phase segments 1 and 2 are necessary to compensate
phase differences. They can be lengthened or shortened by
resynchronization.
Sample Point
The bus level is read at this point and interpreted as a
received bit.
TSEG1
The CAN implementation combines propagation delay seg-
ment and phase segment 1 to form time segment TSEG1.
TSEG2
TSEG2 corresponds to phase segment 2.
SJW
The synchronization jump width gives the maximum number
of time quanta by which a bit may be lengthened or short-
ened by resynchronization.
Fig. 21–10: Bit Timing Definition
The baud rate is then calculated as follows:
21.4.2.2. Bit Timing Configuration
Certain boundary conditions need to be observed when pro-
gramming the bit timing registers. The correct location of the
sample point is especially important with maximum bus
length and at high baud rate.
The information processing time is the internal processing
time. After reception of a bit (sample point) this time is
needed to calculate the next bit for transmission.
With a baud rate of 1 MBd a bit should be at least 8 tQ long.
In case of a triple sample mode (MSAM = 1), the following
boundary condition must also be observed:
The triple sample mode offers better immunity to interference
signals. In the single sample mode a higher transmission
speed is possible.
For high baud rates and maximum bus length, neither SYN
nor MSAM may be switched on. Bosch advises against both
adjustment facilities. When an input filter matched to the
baud rate or a bus driver is used, the triple sample mode is
not necessary. If SYN is set, synchronization will also be
made with the soft edge (dominant to recessive) and this will
mean higher demands being imposed on the clock toler-
ances.
21.4.2.3. Influence of ERM on CAN Timing
When 29 bits are transferred without intermediate resynchro-
nization and with a synchronization jump width of 4 time
tBit tSYNCSEG tTSEG1tTSEG2
++=
tTSEG1TSEG11+() tQ
=
tSYNCSEG 1 tQ
=
tQtXTAL BPR 1+()=
tTSEG2TSEG21+() tQ
=
tSJW SJW tQ
=
Sync Seg
tSYNCSEG
Prop Seg Phase Seg1 Phase Seg2
tTSEG1 tTSEG2
tBit
1 Time quant
def. CAN-SPEC
impl. CAN
Sample Point
BR 1
tBit
--------=
tBit tXTAL( BPR + 1 ) ( 3 + TSEG1 + TSEG2 )=
BR fXTAL
( BPR + 1 ) ( 3 + TSEG1 + TSEG2 )
-----------------------------------------------------------------------------------------=
tTSEG13 tQ
≥
tTSEG1tPROP tSJW
+≥
tTSEG1tTSEG2
≥
tTSEG22 tQ = Information Processing Time≥
tTSEG2tSJW
≥
tTSEG1tPROP tSJW 2tQ
++≥
CDC16xxF-E ADVANCE INFORMATION
154 March 31, 2003; 6251-606-2AI Micronas
quants and prescaler value of 0, the CAN module can handle
a maximum baudrate offset of
With fXTAL = 8MHz and fBAUD = 1MBd the result is ±0.86%.
The ERM introduces a limited uncertainty in the position of
the actual sample time point which may vary from clock to
clock by 0 to 0.121 fXTAL.
Due to this phase modulation, the above maximum baud rate
offset is reduced by a small amount:
With fXTAL = 8MHz and fBAUD = 1MBd the result is ±0.026%,
and the maximum baud rate offset is reduced to ±0.83%.
Furthermore, due to this modulation, the propagation delay
that the CAN node can produce increases by 0.121 fXTAL.
During system design, this increased delay has to be taken
into consideration.
21.4.2.4. Synchronization
The BTL carries out synchronization at an edge (change of
the bus level) in order to compensate for phase shifts
between the oscillators of the different CAN nodes.
21.4.2.5. Hard Synchronization
Hard synchronization is carried out at the start of a telegram.
The BTL ensures that the first negative edge is in the sync.
seg.
21.4.2.6. Resynchronization
Resynchronization takes place during the transmission of a
telegram. If the BTL detects an edge outside the sync. seg., it
can lengthen or shorten the bit. If it detects the edge during
TSEG1, tTSEG1 is lengthened. If it detects the edge during
TSEG2, tTSEG2 is shortened. In this way, it ensures that the
edges lie in the sync. seg. TSJW is the maximum time a bit
can be lengthened or shortened.
Two forms of resynchronization are possible. In normal oper-
ation, synchronization is carried out only with the negative
edge (recessive to dominant). At low transmission speeds,
synchronization can also be carried out with the rising edge
(SYN = 1).
21.5. Bus Coupling
The bus coupling describes the connection of the internal
signals rx (receive line) and tx (transmit line) to the pins to
the CAN bus.
The output pins are push/pull drivers for TLL levels. The
input pins are also designed for TTL levels.
Integrated transceivers (Siliconix Si9200, Philips 82C250
etc.) are available for physical coupling in the high-speed
range in compliance with ISO/DIS 11898.
For a laboratory system a “minimum bus” can be constructed
by means of a wire-Or circuit.
To utilize the advantages of differential signal transmission,
an analogue comparator is necessary.
Fig. 21–11: Bus Coupling
4f
BAUD
⋅
229f
XTAL
⋅⋅
--------------------------------
±
0.121 fBAUD
⋅
229f
XTAL
⋅⋅
---------------------------------
±
Prop.Seg. 2 fXTAL ext. delay tdtxmax tsrxmin 0,121
fXTAL
--------------+–+
⋅⋅≥
TxD
1
0
1
ITX
tx
0
1REF0
+5V
0
1REF1
RxD
OR rx
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 155
Fig. 21–12: Minimum Bus
Table 21–2: Logical Level Transmitting
ITX tx TxD Bus Level Remarks
0 0 0 Dominant direct
011Recessive
101Recessiveinverted
110Dominant
Table 21–3: Logical Level Receiving
REF1
REF0
RxD rx Bus Level Remarks
0 0 x 1 Does not work
0 1 0 1 Recessive inverted
0110Dominant
1 0 0 0 Dominant direct
1 0 1 1 Recessive
1 1 x 0 Does not work
TxD
1
0
1
ITX
tx
0
1REF0
+5V
0
1REF1
RxD
OR rx
+5V
Bus
CAN
CDC16xxF-E ADVANCE INFORMATION
156 March 31, 2003; 6251-606-2AI Micronas
22. DIGITbus System Description
22.1. Bus Signal and Protocol
The DIGITbus is a single-line serial master-slave-bus that
allows clock recovery from the sign stream. Data on the bus
is represented by a pulse width modulated signal. There are
three different signs:
“0”: 25% High Time
“1”: 50% High Time
“T”: 75% High Time
A permanently high bus (100% High Time) means that the
bus is passive high. The bus is active if there are consecu-
tive T-Signs, ones or zeros.
A permanently low bus (0% High Time) is interpreted as bus
reset or failure indicator. Reasons may be shorts, or opens,
or even a low level forced by a bus node to indicate an inter-
nal failure or reset condition.
The sign “T” is used to provide a system-wide clock for the
bus nodes and to separate the address and data fields and
consecutive telegrams.
A telegram normally consists of an address and a data field
separated by one “T”. These fields may be as long as neces-
sary. Thus the length of an address or data field may carry
information. The end is marked by a “T”. The end of a tele-
gram is marked by two T-Signs.
One system implementation may be confined to certain
address and/or data field lengths, thus reducing the hard-
ware or software requirements.
The transmitter of an address has to guarantee that the
address is preceded by four T-Signs at least.
An isolated data field is not possible. Each non-“T”
sequence, which is preceded by two or more consecutive T-
Signs, must be interpreted as an address. An address field is
valid after the reception of the following “T”. The minimum
address length is one bit. The minimum data field length is
zero bit.
Telegrams with more than one data field are also permitted.
For instance TTTTAAATDDDDTDDDDDTT is a valid tele-
gram format on the DIGITbus.
A telegram consisting of one address only is possible, too. In
this case, the length of the data field is zero.
A data field is preceded by an address field and separated
from this by a single “T”. It is followed by one T-Sign. After
reception of two T-Signs the telegram is finished and valid.
In the idle phase (no information exchange) of the bus traffic,
only the bus clock is transmitted.
After the reception of two consecutive T-Signs, all bus nodes
have to be prepared to receive a new telegram starting with
an address field. They are ready to send an address after the
reception of four consecutive T-Signs.
The modification of a T-Sign to a zero or one is done by pull-
ing the bus line to low (dominant state) at the right time. This
is done by a master sending an address or a data bit or by a
slave sending a data bit.
When reading data from a slave, the master first sends the
address. After receiving the address , the slave waits one T-
Sign and then modifies the following T-Signs to zeros and
ones which the master can recognize.
Slaves do not have the possibility to become active on the
bus if they want to communicate a local event or if they need
data from a master. It is a polling bus. Only a master is able
to send an address. The master has to scan the slaves for
their data. But it is possible to transfer data from one slave
directly to another slave. The master has to transmit an
address for which one slave is the source and the second
slave is the destination. Telegrams on the bus are broadcast.
Every bus node may receive them.
01T
bit time
Address DataTT T T T T
AddressTT T TTTTT
TT T T T T TTTTTT
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 157
22.2. Other Features
There are two possibilities for a slave to signal a local event
to the master. They are called wake-up and bus reset.
22.2.1. Wake-up
If the DIGITbus is passive high, (permanent high level for
more than one bit period), a slave can pull the bus line down
to low level. This will awake the master who has to store this
event in a flag, to start the bus clock and to scan the bus for
the source of this event. The minimum low time of the reset
pulse is 1/16 of the nominal bit time (1/Baudrate).
22.2.2. Bus Reset
The rising edge of a bit or bus clock is only controlled by the
bus node which generates the bus clock (clock master). No
other bus node may hold down the bus line at that moment.
When the clock master releases the bus line at the end of a
bit, he must watch the bus line. If the bus level does not rise
after at least 1/2 bit time, this must be interpreted as a proto-
col violation. Delay of 1/2 of a bit time is the latest moment
for a master. He can indicate this protocol violation if the ris-
ing edge is delayed 1/8 bit time. Slaves may use this mecha-
nism to signal an exception to the master. They must pull
down the bus for at least 2 bit times. After such an event,
normal communication may be impossible until the PLL of
bus nodes have synchronized again.
22.2.3. Phase Correction
On a physical bus the signal edges may be delayed by the
bus load. An extra delay may be added by different trigger
edges. The bus nodes see the edges at different times. This
causes them to pull the bus line delayed. To compensate this
effect, the phase correction mechanism allows the bus
nodes to adjust their internal counters.
The master sends a special address, to which the slave
answers with a single zero. The master measures the time
between the rising and the falling edge. With this value, he
can calculate a phase correction value and transmit it to the
slave. The slave may use it to adjust his internal counter.
The Phase Correction has to be done separately for each
bus node.
22.2.4. Abort Transmission
The Abort Transmission feature is an option that allows the
implementation of some kind of rip cord with the DIGITbus.
In the event of an alarm, the SW of the sending master bus
node may break the current telegram and send another tele-
gram instead. The reception of an address/data field cannot
be stopped. The transmission of the alarm telegram is
delayed until after the end of the reception in this case. Only
the currently sending bus node can abort the transmission.
22.3. Standard Functions
The following standard functions have to be included in
every DIGITbus implementation.
22.3.1. Send Bus Clock
The Bus Clock is the sequence of T-Signs on the DIGITbus.
The rising edges of the bus signal are of constant distance.
Only one bus node may generate this Bus Clock even in a
multimaster system. All bus nodes use this stream of
T-Signs to generate telegrams. The bus clock generator
knows two states. “Active Bus” means the transmission of
the Bus Clock. “Passive Bus” means permanent high bus
level. “Passive Bus” may be a low power mode.
22.3.2. Receive Bus Clock
Bus nodes which do not generate the bus clock need an
internal clock for their operation. They may use a separate
clock source or derive their clock from the bus clock by a
PLL. Bus nodes which use own clock sources nevertheless
have to synchronize on the bus clock if they want to transmit
or receive data.
22.3.3. Send Address
The Address is the first bit field in a telegram. Only a master
may send this field. The sender must guarantee that at least
two consecutive T-Signs have been visible on the DIGITbus
before sending this field. Therefore he has to send four T-
Signs. If one of those four transmitted T-signs is disturbed,
only one of the separated telegrams is corrupted for a
receiver. Sending of an address requires synchronization on
the bus clock and, in the case of a multimaster system, colli-
sion detection and arbitration capability.
22.3.4. Receive Address
Every slave and all multimaster-capable bus nodes must be
able to receive an address. For a receiver, a valid address
field must be preceded by two consecutive T-Signs. To verify
a received address it is not sufficient to compare the value.
The length of the address must be correct too, because of
the arbitrary length of the address field.
22.3.5. Send Data
Every master must be able to send a data field, and some
slaves are also able to. A data field is preceded by an
address or data field and one T-Sign.
22.3.6. Receive Data
Every master must be able to receive a data field, and some
slaves are also able to. A data field is preceded by an
address or data field and one T-Sign. It is a good idea to ver-
ify the length of a received data field, if possible. But data
fields of variable lengths are possible too.
22.3.7. Collision Detection
Collision detection together with arbitration is necessary in
multimaster systems. It is necessary to avoid the disturbance
of telegrams if two masters try to send a telegram at the
CDC16xxF-E ADVANCE INFORMATION
158 March 31, 2003; 6251-606-2AI Micronas
same time. As long as both transmit the same sign (one or
zero) at the same time, they don’t detect a collision. If one
master is sending a one and the other is sending a zero, a
zero will be seen at the bus. In this case the master whose
one was modified to the zero, immediately stops sending
and should receive this telegram.
The sender has to arbitrate his part of the telegram.
Write telegram: TTTTTAAAATDDDDTTTTT
Read telegram: TTTTTAAAATDDDDTTTTT
The separator (T-Sign) after an address or data field is object
of arbitration too.
In a single master system arbitration loss has to be managed
as a bus error.
22.4. Optional Functions
The following optional functions may be designed into a cer-
tain DIGITbus implementation.
22.4.1. Abort Transmission
A master controlling the transmission of a telegram can abort
the sending of the address and data field. After four T-Signs
after the last bit he can send another, more urgent telegram.
If he is receiving a data field from a slave, he must wait until
the slave has finished the data field. Then he can insert a
new telegram.
22.4.2. Measure Pulse Width
The capability to measure the pulse width of a high pulse at
the DIGITbus may be used for a phase correction by some
bus nodes. The bus node generating the bus clock sends a
data read telegram to another bus node. The other bus node
answers with a data field which consists of a single zero. The
pulse width of this zero is measured by the master. With this
value he can calculate a phase correction value and transmit
it to this bus member, which may adjust its time slots to the
system dependencies.
22.4.3. Correct Phase
Bus nodes which do not generate the bus clock may use the
procedure described above to adjust their phase. They have
to answer to a special address with sending back a zero.
Afterwards they will receive a correction value with another
special address. With this value they can adjust the point
where they pull the bus line to modify a “T” to a one or a
zero.
22.4.4. Generate Wake-up
If the DIGITbus is passive high (no bus clock, always high
level), the clock master may become wake-up by pulling the
bus level to low (dominant state) for 1/16 bit time at least. All
nodes without the clock master may be able to do that.
22.4.5. Receive Wake-up
If there is a low pulse of at least 1/64 bit time on a passive
high DIGITbus, the clock master must start to transmit the
bus clock by sending T-Signs. All Masters with a bus clock
generation unit must be able to do so in a system which uses
this feature.
22.4.6. Generate Reset
During active DIGITbus a slave may be allowed to pull down
the bus line longer than up to the end of the actual bit time
(2 bit times at least). The rising edge at the end of the bit will
be delayed in this case. This will disturb the bus clock for all
bus nodes.
22.4.7. Receive Reset
The clock master generates the rising edge at the end of a
bit time. He will detect the reset condition described above
and set a flag if the rising edge is delayed for at least 1/8 of
the bit time.
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 159
23. DIGITbus Master Module
23.1. Introduction
The DIGITbus is a single-line serial master-slave-bus that
allows clock recovery from the sign stream. The address and
data field are of arbitrary length.
The DIGITbus Master module is a HW-Module for connect-
ing a single-chip controller to the DIGITbus. It generates the
bus clock and manages short telegrams autonomously.
Transmission and reception of long telegrams are supported
by a FIFO each. The DIGITbus Master may be used in a sin-
gle or in a multimaster bus system.
Features
– Single master in a singlemaster system.
– Clock master in a multimaster system.
– Passive master in a multimaster system.
– Bus clock generation.
– Receive and transmit a telegram with address and data
field.
– Transmit FIFO and receive FIFO.
– Collision detection and arbitration.
– Abort transmission.
– Sleep mode.
– Bus monitor mode.
– Measurement of pulse width for phase correction.
– Phase correction.
– Reception of wake-up and bus reset signal.
– Register interface to the CPU.
23.2. Context
Apart from reset and clock line, the interface to the CPU con-
sists of registers connected to the internal address and data
bus. An output signal may be connected to the interrupt con-
troller.
A modified universal port builds the output logic which is con-
nected with its special input and output to the DIGITbus Mas-
ter. This provides an easy way for the SW to hold the bus line
permanent low or high, or investigate bus level directly, with-
out support of DIGITbus Master HW.
An open drain output instead of a push/pull output is neces-
sary for the universal port to build a single-line wired-and
bus.
Fig. 23–1: Context Diagram
DIGITbus
Master
ADB
DB
R/W
Reset
from clock
Interrupt
tx
rx
Universal Port with
Open Drain Output
Port Pin DIGITbus
Other
Transmitter
+U
divider
SO
SI
CDC16xxF-E ADVANCE INFORMATION
160 March 31, 2003; 6251-606-2AI Micronas
23.3. Functionality
23.3.1. 3-bit-Prescaler
The programmable 3-bit-Prescaler supplies the module with
clock signals. It scales down the clock divider clock by factor
1, 2, 3 to 8 (see Table 23–2 on page 162). The output is 64
times the bus clock. The desired input frequency from the
clock divider is hardware programmable.
23.3.2. Internal Clocks
In low-power mode, the clock supply of the whole module
with exception of the receive bit logic can be stopped. The
receive bit logic needs a clock in low-power mode too, as it
must filter and watch the bus line for a wake-up signal.
23.3.3. Transmit T
The transmit T logic sends a continuous stream of T-signs, if
active. It outputs a permanent high if it is inactive.
23.3.4. Transmit Bit
Depending on the input signals, the transmit bit logic modi-
fies the T-signs to ones or zeros.
A phase correction can be done by adjusting the start time of
a transmit bit sequence.
Other bus behavior than sending zeros, ones or T-signs may
be enforced by the SW using the universal port in normal
mode directly. The bus line may be released or pulled low.
23.3.5. Receive Bit
The receive bit logic samples the bus level at a frequency of
64 times of the bus clock. It filters the input signal and
decodes the input stream to supply the receive telegram
logic with the logical bus signals (0, 1 and T) and the receive
clock. In addition, it measures the pulse width of each non T-
sign. It creates a bus reset signal if the active bus is held
down beyond the end of a bit time. It creates a wake-up sig-
nal if there is a low level on the passive high bus.
23.3.6. Send Telegram
The send telegram logic will be enabled by the transmit FIFO
and the receive telegram logic when four consecutive T-
signs have been received. It supports the transmit bit logic
with the transmit bit sequence. If it recognizes the beginning
of a new field, it waits one bit time (separator T-sign).
23.3.7. Receive Telegram
The receive telegram logic traces the bus and indicates the
state to the status register and other related modules. The
received bit field is written to the receive FIFO. The receive
telegram logic is active all the time. Even if the module is
transmitting a telegram, all bits must also be received in a
multimaster system, because arbitration may be lost. Recep-
tion of own telegrams can be disabled (in a singlemaster sys-
tem).
23.3.8. Collision Detection
The collision detection logic compares each incoming bit with
the currently outgoing bit. A difference is signalled to the
send telegram logic. If the module is transmitting, the send
telegram logic is stopped immediately, and the transmit FIFO
and shift register are flushed.
23.3.9. Transmit FIFO
The transmit FIFO has five entry addresses. One for the field
length of address or data field, one for a address byte, one
for a data byte, one for more address bytes and one for more
data bytes. The field length has to be written once before the
corresponding field is entered into the FIFO unless the field
length is not a multiple of 8.
An entry into the address register is inserted into the bus
clock after the reception of 4 consecutive T-signs. An entry
into the data register is inserted into the bus clock after the
reception of a non-T-sign and one T-sign. Thus it is possible
to append a second data field (maybe acknowledge) after
the reception of a telegram.
The transmit FIFO may be flushed to abort a transmission. It
is also flushed if the transmit telegram logic is active and a
collision is detected.
23.3.10. Receive FIFO
The receive FIFO will be filled from the receive shift register.
It has two exit addresses. One for the field length and field
type and one for the bit field. The field length has to be read
before the corresponding field is taken from the FIFO. The
receive FIFO will be frozen if it is full. The receive shift regis-
ter will be overwritten.
23.3.11. Interrupt
Several flags of the status registers are connected with the
interrupt source signal by a logical-OR. The interrupt output
can be masked by a flag in the control register.
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 161
Fig. 23–2: Block Diagram
DIGITbus
Master
ADB
DB
R/W
Reset
from clock
DIGITbus
rx
tx
3-Bit-Prescaler
Control/Status
Pulse Width
Transmit
Bit
Receive
Bit
Receive
Telegram
T
dat
Address
Decoder
64 x bus clk
Transmit
Telegram
wake-up/bus reset
Collision
Detection
generate bus clock
arbitration lost
RxSR
TxFIFO
RxFIFO
divider
Transmit
64 x bus clk
rxclk
full
empty
Tx More Data Field
Rx Field Length
flush
64 x bus clk
data lost
Tx More Addr. Field
Tx Field Length
Rx Field
T
dat
T
&
T-Seq.
0-Seq.
txclk
rx external only
rise
Phase
run
Tx Data Field
Tx Addr. Field
1-Seq.
TxSR
Interrupt
Source
CDC16xxF-E ADVANCE INFORMATION
162 March 31, 2003; 6251-606-2AI Micronas
23.4. Registers
The register mnemonic prefix “DG” stands for DIGITbus.
An “x” in a writable bit location means that this flag is
reserved. The user has to write a zero to this location for fur-
ther compatibility. An “x” in a readable bit location means that
this flag is reserved. A read from this location results in an
undefined value.
RUN Run
r/w1: Module clock is active.
r/w0: Module is not clocked.
The module is absolutely inactive if RUN is zero. Other flags
are not functional then.
GBC Generate Bus Clock
r/w1: Module generates bus clock
r/w0: No bus clock
ACT Activate
r/w1: Module is active (reception and transmission).
r/w0: Module is sleeping (low power mode).
Only the receive bit logic is active in the low-power mode.
RXO Receive External Only
r/w1: Don’t receive own telegrams.
r/w0: Receive all.
PSC Prescaler
r/w: Scaling value
INTE Enable Interrupt
r/w1: Enable interrupt
r/w0: Disable interrupt
ENEM Enable Not Empty Interrupt
r/w1: Enable
r/w0: Disable
ENOF Enable Not Full Interrupt
r/w1: Enable
r/w0: Disable
PHASE Phase Correction Field
r/w: Transmit phase.
The start of the transmit frame can be selected in increments
of 1/64 of a total bit time related to the rising edge. Values
between 0 and 15 are possible, but only the interval from
0 to 9 results in correct behavior.
Table 23–1: Register Mapping
Addr.
Offs. Mnem. readable writable
0 DGC0 Control 0
1 DGC1 Control 1
2 DGS0 Status 0
3 DGRTMD Rx Length Tx More Data
4 DGTL Tx Length
5 DGS1TA Status 1 Tx Addr.
6 DGTD reserved Tx Data
7 DGRTMA Rx Field Tx More Addr.
DGC0 Control Register 0
76543210
r/w RUN GBC ACT RXO X PSC 2 to 0
0000x000Res
Table 23–2: Clock Prescaler
PSC
hex Divide
by Bus Clock in kHz
@ 6 MHz @ 8 MHz @ 10 MHz
0 1 93.75 125.0 156.25
1 2 46.9 62.5 78.1
2 3 31.25 41.7 52.1
3 4 23.4 31.25 39.1
4 5 18.75 25.0 31.25
5 6 15.6 20.8 26.0
6 7 13.4 17.9 22.3
7 8 11.7 15.6 19.5
DGC1 Control Register 1
76543210
r/w INTE ENEM ENOF x PHASE
000x0000Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 163
Fig. 23–3: Phase Correction
RDL Receive Data Lost
r1: Data lost
r0: No data lost
This flag is set if the receive FIFO is full and the shift register
tries to store its contents to the FIFO, because a new bit
arrives. In this case the FIFO is frozen but the shift register is
overwritten. It must be interpreted and cleared by the user. It
is cleared by reading an entry from the FIFO.
NEM Rx FIFO is Not Empty
r1: There is at least one entry to read.
r0: Empty.
(see Fig. 23–4 on page 163)
NOF Tx FIFO is Not Full
r1: There is at least one entry free.
r0: Full.
It generates an interrupt only at the precise moment when
the limit is passed. It doesn’t generate interrupts when the
FIFO is empty (see Fig. 23–4 on page 163).
TGV Telegram Valid
r1: Telegram valid
r0: Telegram not valid
w0: Clear flag
This flag will be set if two consecutive T-signs have been
received. It is reset by the HW if a non-T-sign is received. It
can be cleared by the user if the related telegram is evalu-
ated.
PV Protocol Violation
r1: Wake-up if bus is passive high.
Bus reset if bus is active.
r0: No trouble
w0: Clear flag
It must be interpreted and cleared by the user. It is set when
the receive bit logic enters or leaves state passive high or
when it enters the state passive low.
Fig. 23–4: Rx- and TxFIFO Timing
ERR Error
r1: Fatal error.
r0: No error
w0: Clear flag
The HW sets this flag either if a dominant level is transmitted
and a recessive level is detected (collision error), or if there
was a wrong edge within a received bit. If a collision error is
detected during transmission, the flag ARB will be set too
and transmission stops immediately. This flag has to be
cleared by the user.
ARB Arbitration Lost
r1: Arbitration lost.
r0: No arbitration loss.
w0: Clear flag
This flag will be set if a collision is detected during transmis-
sion. It must be cleared by the user. The transmit buffer is
flushed if ARB is true. It is impossible to write to the transmit
FIFO as long as ARB is true. Wait until flag TGV is true
before reloading TxFIFO. This is automatically done if ARB
is evaluated within the TGV interrupt subroutine only.
The Flags RDL, NEM, NOF, TGV, and PV trigger the inter-
rupt source signal (see Section 23.5.7. on page 167).
The first byte of an address field must be written to DGS1TA.
DGS0 Status Register 0
76543210
0 163248 0
0 163248 0
416 32 48 0
Phase delay
PHASE = Start value of transmit counter.
Bit time
Transmitted
Received
Corrected
wxxxTGVPVERRxARB
rRDL NEM NOF
x01000x0Res
DGS1TA Status 1 & Tx Address Register
76543210
NEM
RxFIFO
Interrupt
NOF
TxFIFO
Interrupt
EMPTY
wTransmit Address
rSTATE PW5 to 0
01000000Res
CDC16xxF-E ADVANCE INFORMATION
164 March 31, 2003; 6251-606-2AI Micronas
STATE Bus State
r: State of receive bit logic.
PW Pulse Width
r: Pulse width
The pulse width of the most recent non-T-sign is stored in
this register. It is measured in increments of 1/64 of the bus
clock period.
More bytes of a data field must be written to DGRTMD.
The read part of register DGRTMD is associated with the
front entry in the receive FIFO (the receive field DGRTMA). It
has to be read and interpreted before the corresponding
FIFO entry.
RDL Receive Data Lost
r1: Data lost
r0: No data lost
The flag RDL from the status register DGS0 is mirrored here.
It is cleared by a read access to register DGRTMA.
NEM Receive FIFO is Not Empty
r1: There is at least one entry.
r0: Empty
The flag NEM from the status register DGS0 is mirrored
here. FTYP, EOF, LEN and register DGRTMA are not valid if
NEM is false.
FTYP Field Type
r1: Address field
r0: Data field
EOF End of Field
r1: Last byte of a field
r0: Not last byte of a field
If EOF is set, the corresponding FIFO entry is the last part of
the actual field. The next entry, if there is one, belongs to a
new field.
LEN Length of Field
r: Length of valid data bit
The three-bit length does not limit the overall length of the
corresponding field. The field length defines how many bits
of the front entry of the receive FIFO carry valid bits. They
are right-aligned (Table 23–4). The real length of the field is
unlimited. The user must count the bytes fetched from the
FIFO to calculate the real field length.
The examples in Table 23–5 illustrate the interpretation of
register DGRTMD. They are valid for an address field (FTYP
= 1) or a data field (FTYP = 0).
More bytes of an address field must be written to DGRTMA.
The bytes of a received field must be read from register
DGRTMA. The meaning of this field (address or data) is
defined by the flag FTYP.
Received bytes of a bit field are right-aligned. The last byte
of a long bit field (with the LSB) may be filled partially. To get
the whole bit field right-aligned it is necessary to shift all pre-
ceding bytes to the right.
A read access to this register takes the top entry of the
receive FIFO. Both registers DGRTMA and DGRTMD are
overwritten by the next FIFO entry as result of a read access.
Table 23–3: Receiver States
STATE Bus
0 0 Passive low
0 1 Passive high
1 0 Active low
1 1 Active high
DGRTMD Rx Length & Tx More Data Register
76543210
wTransmit More Data
rRDL NEM FTYP EOF x LEN2 to 0
00xxxxxxRes
Table 23–4: LEN usage, Receive and Transmit Length
LEN
2 1 0 Valid Bit Numbers
7 6 5 4 3 2 1 0
1 0 0 1 _ _ _ _ _ _ _ x
2 0 1 0 _ _ _ _ _ _ x x
3 0 1 1 _ _ _ _ _ x x x
4 1 0 0 _ _ _ _ x x x x
5 1 0 1 _ _ _ x x x x x
6 1 1 0 _ _ x x x x x x
7 1 1 1 _ x x x x x x x
0 0 0 0 x x x x x x x x
Table 23–5: DGRTMD Interpretation Examples
LEN EOF
6 1 Last byte of a field. The six rightmost
bits belong to the field.
0 0 A byte of a field. All bits belong to the
field. At least one byte follows.
0 1 Last byte of a field. Eight bits belong
to the field.
≠0 0 Impossible.
DGRTMA Rx Field & Tx More Address Register
76543210
wTransmit More Address
rReceive Field
xxxxxxxxRes
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 165
The Transmit Length Register is associated with the whole
field (address or data) which will be written into the transmit
FIFO. It has to be written before the first entry of the field.
FLUSH Flush Tx FIFO
w1: Empty Tx FIFO and abort transmission.
w0: No action.
This flag will be reset by the HW autonomously. After FLUSH
wait at least one bit time before rewriting TxFIFO.
EMPTY Tx FIFO is Empty
r1: No transmit telegram in FIFO.
r0: Transmit telegram in FIFO.
LEN Length of Field
w: Length of address or data field.
These three bits correspond to the first byte of a bit field.
They define how many bits of this byte carry valid information
and should be transmitted (see Table 23–4 on page 164).
DGTL must be written before the first byte of the actual bit
field is written to the FIFO. It only has to be written once for
each bit field. The overall length of the bit field is not limited.
The first byte of a data field must be written to DGTD.
The first byte of a bit field (with the MSB) which is entered
into DGS1TA or DGTD, may be partially filled. In the follow-
ing bytes all bits must contain valid data.
23.5. Principle of Operation
23.5.1. Reset
The module reset signal resets all registers and internal HW.
The same module reset signal does a standby bit in a
standby register.
Setting flag RUN in register DGC0 resets all internal HW and
registers, with exception of registers DGC0, DGC1, DGS0
and DGS1TA. These registers are accessible all the time,
they are not reset by any setting of the DIGITbus Master
flags.
Internal HW is reset to an inactive state (not transmitting, not
receiving). Internal counters are reset to zero. FIFOs and
shift registers are empty. Internal representations of the bus
line are reset to passive bus level (high).
Fig. 23–5: Reset Structure
23.5.2. Initialization
The corresponding port must be configured as special out
open drain.
After reset, and after setting flag DGB in standby register
SR2, the DIGITbus master is inactive. The global enable flag
RUN must be set together with the appropriate prescaler
entry PSC, to activate the module.
23.5.2.1. Clock Master
The flag GBC (generate bus clock) must be set, if the DIGIT-
bus master should generate the bus clock. The module now
acts as clock master of the connected DIGITbus system. It
outputs a stream of T-signs.
23.5.2.2. Receiver/Transmitter
Setting the flag ACT activates the receive and transmit logic.
From now on, all telegrams are received in the receive FIFO.
Writing to the transmit FIFO initiates transmission of a tele-
gram.
The bus clock (T-signs) must be activated some time before
the first telegram is transmitted. This is necessary, as other
modules may use a PLL for generating the internal clock
from the bus clock. No telegram shall be transmitted before
all modules have locked on the bus clock.
23.5.2.3. Singlemaster System
In a singlemaster system (no collision possible), you can
suppress reception of transmitted telegrams by setting flag
RXO (receive external only). This unburdens the CPU from
clearing the receive FIFO of those telegrams.
DGTL Transmit Length Register
76543210
wxFLUSHx x x LEN2 to 0
x0xxx000Res
rxEMPTYxxxxxx
x1xxxxxxRes
DGTD Transmit Data Register
76543210
wTransmit Data
xxxxxxxxRes
QR
C0.RUN
Registers
C0
C1
Internal
HW
reset
reset S1.TSTn
and
remaining
registers
QR
SR2.DGB
reset
CDC16xxF-E ADVANCE INFORMATION
166 March 31, 2003; 6251-606-2AI Micronas
23.5.2.4. Multimaster System
In a multimaster system it is necessary that each transmitted
telegram is received too, because arbitration may be lost and
then the transmitter becomes a receiver. If arbitration has not
been lost, the receive FIFO must be read to empty it. The
flag RXO has to be cleared in a multimaster system.
23.5.3. Transmission
Transmission is initiated by writing a telegram into the trans-
mit FIFO.
If the field length is not a multiple of 8 bit, the total field length
module 8 has to be written to register DGTL. This must be
done once for each field, and before any entry in registers
DGS1TA, DGTD, DGRTMA or DGRTMD. If the total field
length is a multiple of 8, it is not necessary to write the field
length into register DGTL.
The first entry of a field (address or data) has to be written
right-aligned to register DGS1TA (address) or DGTD (data).
Further entries of the same field, if it is longer than 8 bit, have
to be written to DGRTMA (more address) or DGRTMD (more
data). A telegram is transmitted MSB first, hence fields have
to be written to transmit FIFO MSB first.
A new address field is transmitted if at least four consecutive
T-signs were on the bus. A new data field is transmitted if
there was exactly one T-sign. If the last bit of a field was
transmitted and there are no more entries in the transmit
FIFO, the transmitter stops sending. After reception of two
consecutive T-signs the telegram valid flag TGV is set. This
is the signal for the SW to evaluate whether transmission
was correct or whether an arbitration loss or an error have
cancelled transmission (flags ARB, PV and ERR). In the lat-
ter case SW must initiate retransmission.
A telegram has been transmitted correctly if ARB and ERR
are false and EMPTY is true.
23.5.3.1. Transmit FIFO
SW must ascertain that there is an empty entry in the trans-
mit FIFO, before writing to it. Flag NOF (not full) indicates
that there is at least one entry free. Flag EMPTY indicates
complete emptiness of transmit FIFO. After reset, FLUSH or
ARB wait until flag TGV is true before rewriting TxFIFO.
Short telegrams can be completely buffered in the FIFO.
Managing long telegrams is a SW job. The SW must buffer
long telegrams and write the parts in time. The transmit FIFO
is intended to unburden the CPU from immediate reaction to
a NOF interrupt. If an entry becomes free, the SW has time
to write, as long as it needs to transmit two FIFO entries and
the contents of the transmit shift register. This time must not
necessarily be the duration of sending 24 bits. Possibly, only
one bit of each remaining FIFO entry has to be sent.
The transmit FIFO is not intended for telegram tracking. Only
one transmit telegram at a time must be entered.
23.5.4. Reception
Every non-T-sign is shifted into the receive shift register. If it
is full, or if a T-sign was received, the shift register is stored
into the receive FIFO. This is done until the receive FIFO is
full. In this case, the FIFO is frozen, but the shift register con-
tinues operation. The flag RDL indicates the latter case.
If the shift register is stored to the receive FIFO because a T-
sign was received, the corresponding flag EOF is set, indi-
cating that this is the last entry of a field.
The corresponding flag FTYP is modified at the same time. If
two or more consecutive T-signs were received in front of the
actual field, it is set, indicating that this field has to be inter-
preted as an address field. If only one T-sign has been
received in front of the actual field, it is cleared, indicating
that it has to be interpreted as a data field.
The flag TGV is set if two consecutive T-signs were received.
This is the moment to read status flags and Receive FIFO.
The flags PV and ERR have to be interpreted. Even if an
error has occurred, the Receive FIFO must be emptied by
reading it because every telegram or fragment is stored
there. Otherwise reception of the next telegram may overflow
the receive FIFO, which is indicated by flag RDL.
Every time you want to read DGRTMA, it is ingenious to read
DGRTMD first, because DGRTMD and DGRTMA are over-
written with a read access to DGRTMA.
23.5.4.1. Receive FIFO
The receive FIFO contains entries as long as flag NEM is
true.
Short telegrams can be buffered completely in the receive
FIFO. SW must buffer long telegrams and read parts of it in
time.
23.5.5. Sleep Mode
Only the receive bit logic is active in sleep mode. Neither
transmission nor reception of telegrams is possible.
A wake-up (passive high to low edge) is signalled by flag PV.
The DIGITbus master is not automatically activated by a
wake-up. This has to be done by SW. The flag PV can be
used to trigger an interrupt.
Switching to Sleep Mode while a telegram is being transmit-
ted can cause problems. Hence, please make sure that bus
clock generation is switched off only if bus is idle (T-signs).
Table 23–6: Operating modes
RUN GBC ACT RXO Remarks
0 x x x Standby mode
1 0 x x Passive master. Exter-
nal bus clock genera-
tion is necessary.
1 1 x x Clock master
1 0 0 x Sleep mode
1x1xActive mode
1 x 1 0 Receive all. (Recom-
mended in multimaster
system)
1 x 1 1 Receive external only.
(Recommended in
singlemaster system)
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 167
23.5.6. Abort Transmission
Writing a one to flag FLUSH aborts the transmission of a
telegram after completion of the actual transmitted bit, if the
DIGITbus master is the transmitter. The transmit FIFO is
emptied and another, more urgent telegram can be transmit-
ted. Transmission of the new telegram starts as soon as 4
consecutive T-signs have been received after the aborted
telegram.
It is not possible to abort a telegram or a field which is trans-
mitted by another bus node.
23.5.7. Interrupt
Five flags (RDL, NEM, NOF, TGV, PV) are connected to the
interrupt source output by an or operation. This output can
be enabled globally by flag INTE. The interrupt generation of
two flags (NEM, NOF) can be enabled locally by flags ENEM
and ENOF. A rising edge of a flag triggers the interrupt
source output.
Fig. 23–6: Interrupt Sources
23.5.8. Measure Pulse Width
The pulse width (high time) of every non-T sign is stored with
the falling edge of the bus signal in status register DGS1TA
in the field PW. T-signs don’t affect PW. It must be read
before the falling edge of the next non-T sign.
23.5.9. Correct Phase
The rising edge of the bus signal can be delayed by inner
(sampling and filter) or outer (bus load) influences. This
delayed rising edge resets a 6-bit transmit counter in the
transmit bit logic. The transmit counter pushes the bus line
low when it reaches 15 (transmitting 0) or 31 (transmitting 1).
It releases the bus line when it reaches 55.
The transmit counter is reset to a value which contains two
zeros at the most significant position and the four PHASE
bits of the control register DGC1 at the least significant posi-
tion. This allows an adjustment of the transmitted non T
signs between 0 and 1/15 of the whole bit length.
23.5.10. Error
The setting of flag ERR may have one of the following
causes:
– Wrong baud rate of DIGITbus Master or other bus nodes.
– Wrong port configuration of DIGITbus Master.
– Disturbances on bus line.
– DIGITbus Master HW damaged.
23.5.11. Precautions
Don’t use indirect addressing when you write to a DIGITbus
register. An unwanted read access to the same address can
be the result and a read access to the RxFIFO output regis-
ter modifies the content of this FIFO. Received data are lost
in this case.
If a telegram is aborted by FLUSH, normally there is a TGV
interrupt with the reception of the second T sign after the last
bit of the aborted telegram. The next TGV interrupt signals
the transmission of the alarm telegram. If a transmit telegram
is aborted by FLUSH before transmission has actually
started, then the first TGV interrupt of the aborted telegram
doesn’t occur. In this case the TGV interrupt signals the
transmission of the alarm telegram.
Don’t access DIGITbus registers in CPU SLOW mode. This
can cause interrupts. Operation of the module in CPU SLOW
mode is allowed.
RDL
NEM
NOF
TGV
PV
INTE
DIGITbus
Interrupt
Source
OR &
&
&
ENEM
ENOF
CDC16xxF-E ADVANCE INFORMATION
168 March 31, 2003; 6251-606-2AI Micronas
23.6. Timings
Fig. 23–7: Tx Timing
Fig. 23–8: Rx Timing
D T T T T T A A T D D T T T T
D T T T T T A A T D D T T T T
Tx stream
Rx stream
txa
TGV
Bus Clock
NEM
ARB
collision
D T T T T T T T T T T T T T T
D T T T T T A A T D D T T T T
Tx stream
Rx stream
txa
TGV
Bus Clock
NEM
ARB
collision
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 169
24. Audio Module (AM)
The Audio Module AM provides a gong output signal that
may be used to drive a speaker circuit.
The output signal is a square wave signal with selectable
gong frequency.
The gong signal amplitude is defined by the pulse-width of a
PWM signal. An internal accumulator is selectable to auto-
matically decrease this pulse-width, and thus the gong ampli-
tude following an exponential function.
Features
– Programmable gong frequency
– Programmable gong duration
– Programmable initial amplitude
– Gong can be stopped and retriggered
– Generation of an exponentially decreasing gong
amplitude function without CPU interaction
Fig. 24–1: Block diagram of the audio module
&8 Bit - PWM
COMP Amplitude=0 ?
R
SQ
FCL
FPWM
1/ 2FGONG
Amplitude - Latch (13 Bit)
+-
1/32
MSB
7-Bit Counter
13
8
13
8
Set Frequency-Register
Data Bus
Write AMF
Data Bus
AMAS
1313
LSB’s
’1FH’
CLK
Clear Counter
Write AMAS
AMA
7
CLR
5
U 5.3
AM-OUT
5-Bit Counter
Set Decrement-Register
Data Bus
Write AMDEC
CLK
Clear Counter
3
FDecrement
(Start/Stop gong sound)
Adder
813
13
1/(2n)
Read AMAS
= 15.625 kHz
= 4 MHz
&
&
Data Bus
(Bit 7)
AMA
U 5.2
AM-PWM
VDD
AMMCA
0
HW Options
1/n
AM Trigger
AM Clock
CDC16xxF-E ADVANCE INFORMATION
170 March 31, 2003; 6251-606-2AI Micronas
24.1. Functional Description
The gong sound frequency is adjusted with the Audio Mod-
ule Frequency Register (AMF). FGONG is FPWM divided by
twice the [AMF value + 1]. The length of AMF is 7 bit, so val-
ues from 0 to 127 can be written.
The initial gong sound amplitude is set by writing the Audio
Module Amplitude & Status Register (AMAS), this write also
starts the gong sound. An active audio module is indicated
by the read only Audio Module Active Bit (AMA) in the
AMAS.
Every 1st..32nd cycle of the gong sound frequency (depend-
ing on the Audio Module Decrement Register (AMDEC)), a
new amplitude value is calculated (FDecrement). The falling
edge of the amplitude decrement frequency FDecrement
latches the output of the adder into the amplitude latch (13 -
bit), and simultaneously the 8 MSB’s into the PWM.
During the first low cycle of FGONG following the active FDec-
rement edge, the PWM ialready runs with the newly calculated
amplitude, but takes effect at the output not until the next
high cycle of FGONG. FGONG is modulating the PWM-output
to generate the gong sound frequency, while the decreasing
PWM-value generates an exponentially decreasing ampli-
tude.
As soon as the 8 MSBs of the amplitude latch reach zero,
the AMA will be reset, which deactivates the audio module.
The audio module is only operable in the CPU FAST mode.
24.1.1. Hardware Settings
The AM clock frequency FCL is set by HW option FFB7h. The
AM trigger frequency FPWM (PWM reload frequency) is set
by HW option FFC3h. Select 4MHz for FCL and FCL/256 for
FPWM to achieve the standard Audio Module functionality.
24.1.2. Initialization
To connect the audio module output with the corresponding
output pin (U5.3), in the register U5M32 the flag PMODE has
to be set to switch into the Port Mode. Additionally the port
U5.3 has to be switched into the Special Out Mode by setting
the flag TRI1 to ’0’ and the flag N/S1 to ’1’ in the register
U5SEG32.
There is one register to set the gong sound frequency (AMF)
and another one to set the gong sound duration (AMDEC)
before the gong sound can be started. Both their reset val-
ues are zero.
24.1.3. Start Gong
The gong sound is started by writing the initial amplitude
value into the Audio Module Amplitude & Status Register
(AMAS). Simultaneously with the write to AMAS the Flag
Audio Module Active (AMA) is set, which enables the FPWM-
and FCL-inputs. The setting of AMA to ’1’ also enables the
FGONG- and FDecrement - counter.
24.1.4. Restart Gong
It’s possible to restart the gong sound simply by writing a
new initial amplitude value to the AMAS (independently from
the former initial value or the current value of the register.
Note: The current amplitude value can’t be read out). The
new gong sound will start immediately with a low cycle of
FGONG.
24.1.5. Stop Gong
The gong sound will stop automatically, as soon as the
amplitude value in the AMAS reaches zero. This will reset
the AMA, which indicates the inactive audio module.
To stop the gong sound, just write 00H into the AMAS. The
gong sound then will stop immediately with the writing of
00H. (also indicated by AMA).
A continuous tone will never stop automatically. It can also
be stopped by writing 00H into AMAS.
24.1.6. Decay of Sound
The decay characteristic used for this gong sound is
described by the following exponential function (see Fig. 24–
3 on page 172):
An = A0 (1 - 1/32)n
with A0= initial amplitude
An= amplitude after n FDecrement cycles
n= int (t * FDecrement )
Each FDecrement cycle the amplitude is decreased by 1/32.
FDecrement is determined by the value of GDF in the register
AMDEC and by the value of the Audio Module Frequency
Register (AMF):
FDecrement = FGong / 2 GDF
for GDF settings of 0 .. 5
With GDF settings of 6 and 7 the gong sound amplitude
update frequency FDecrement is zero (continuous tone).
The time constant τ of the above exponential function is
defined as the time interval within which the amplitude A is
decreasing to 36.8%.
Given
the number nτ of FDecrement cycles needed to reduce the ini-
tial amplitude to 36.8% is
nτ ≅ 32
This means that τ is correlating with FDecrement. The higher
FDecrement, the shorter is τ.
With an initial amplitude of FFH the total time t255->0 needed
to reach zero amplitude in the 8 Bit - AMAS is n = 193 FDecre-
ment cycles, which is approximately 6τ.
With an initial amplitude lower than FFH the gong sound
duration is shorter.
To sum up, it can be said that the total duration of the gong
sound depends on FGong, set with AMF, the setting of the
Gong sound Duration Factor GDF and the setting of the ini-
0 368,11
32
------–
nτ
=
t255 0→193 1
FDecrement
------------------------- 193 2GDF
FGONG
-----------------
==
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 171
tial amplitude (see Table 24–1 on page 171). Please note
that senseless combinations of register values are also pos-
sible.
Fig. 24–2: Example sections of the audio module output signal
Table 24–1: Total gong sound running time from A0 = FFH to AFinal = 00H for selected gong sound frequencies
(approx. 6τ; FPWM=15.625 kHz)
FGong (AMF) GDF=0 GDF=1 GDF=2 GDF=3 GDF=4 GDF= 5
61.0 Hz (min.) 3.20 s 6.39 s 12.8 s 25.6 s 51.1 s 86.6 s
100 Hz 1.95 s 3.9 s 7.8 s 15.6 s 31.2 s 62.4 s
601 Hz 0.324 s 0.649 s 1.30 s 2.60 s 5.19 s 10.4 s
2.60 kHz 75.0 ms 150 ms 300 ms 600 ms 1.2 s 2.4 s
7.81kHz (max.) 25.0 ms 49.9 ms 99.9 ms 200 ms 399 ms 799 ms
123012
13 25 26
FPWM
FGONG
50 %
PWM
100 %
15 %
Conditions: (FPWM = 15.625 kHz; FGONG = 601 Hz; AMDEC value = 2, FDecrement = 150.25Hz)
after 0 s (initial)
after 0.398 s (1.87τ)
after 0.146 s (0.68τ)
Pulse
Duty
Factor
A
B
C
14 15
time
F
Decrement
Gong
Output
Pin
4 x B4 x A 4 x C
0 s 0.146 s (0.68τ) 0.398 s (1.87τ)
start gong new amplitude (n.a.) n.a. n.a. n.a. n.a.
FGONG
zoomed
gong
output
signal
CDC16xxF-E ADVANCE INFORMATION
172 March 31, 2003; 6251-606-2AI Micronas
Fig. 24–3: Decay of Sound - Function
20
40
60
80
100
120
140
160
180
200
220
240
260
AMAS
(MSB of amplitude latch)
time
1
τ
81624
32 40 48 56 64 72 80
2
τ
no. F
Decrement
-cycles
36.8%
13.5%
5.0%
1/F
Decrement
96
3
τ
128
4
τ
160
5
τ
192
6
τ
88 104 112 120 152136 144 184168 176 200
1.8%
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 173
24.2. Registers
The Audio Module control registers are located in the I/O
area.
Initial Amplitude
A write access to this register starts or stops the gong sound,
while the value written is the initial gong sound amplitude.
Writing the value 00H into this register during an active gong
sound deactivates the gong sound immediately, while writing
a value > 00H is restarting the gong sound immediately with
the new Initial Amplitude.
wxx: (Re-)Start gong sound with Initial Amplitude.
w00: Stop gong sound.
AMA Audio Module Active Flag
This flag indicates an active Audio Module generating a gong
sound.
r1: Audio Module is active.
r0: Audio Module is not active.
With this register the gong sound frequency is programmed.
The PWM frequency is divided by twice the register value
increased by one.
The value which has to be written, resp. the resulting gong
sound frequency is calculated with:
With the 7bit value ranging from 0 to 127, the programmable
gong sound frequency range is 61 Hz .. 7.81 kHz (see
Table 24–3 on page 173).
It’s possible to write a new gong sound frequency during an
active audio module (AMA = ’1’).
AMMCA Audio Module Maximal Constant
Amplitude Flag
w1: Activate the AMMCA mode.
w0: Deactivate the AMMCA mode.
With the flag AMMCA the Audio Module Maximal Constant
Amplitude (AMMCA) mode is selected. If this Flag is set, the
gong sound with the maximum, not decreasing amplitude is
available at the audio module output pin. The only difference
between this tone and a ’normal’ gong sound is the constant,
not decreasing amplitude. The handling of this tone (i.e.
start, stop, frequency, duration) is the same. The tone is
started by writing an initial value to AMAS, but this value will
only influence the duration of the tone, not its amplitude.
GDF Gong sound Duration Factor
This register sets the gong sound duration in dependence of
FGONG. With GDF=0 the amplitude will be decreased every
FGONG - cycle, values 1 to 5 will result in a amplitude update
frequency of FGONG / 2 to FGONG / 32 according this equa-
tion:
A value of 6 or 7 means no decrease of the amplitude, so a
continuous tone with the initial amplitude will be generated
(FDecrement = 0). To stop the continuous tone write a ’00H’ to
AMAS or change the gong sound duration factor to let the
Table 24–2: Audio Module Control Registers
Mnemonic Name
AMAS Audio Module Amplitude/Status
AMF Audio Module Frequency
AMDEC Audio Module Decrement
AMAS Audio Module Amplitude and Status
Register
76543210
Note
AMF Audio Module Frequency Register
76543210
Note
wInitial Amplitude
rAMAxxxxxxx
0xxxxxxxRes
wx Sound Frequency
-0000000Res
value FPWM
2FGONG
-------------------- 1–=
FGONG
FPWM
2value 1+()
--------------------------------=
Table 24–3: Examples for AMF-values
AMF - value resulting FGong
127 (max.) 61.0 Hz (min.)
::
77 100 Hz
::
12 601 Hz
::
22.60 kHz
::
0 (min.) 7.81kHz (max.)
AMDEC Audio Module Decrement Register
76543210
Note
wAMMCA x x x x GDF
0----000Res
FDecrement
FGONG
2GDF
-----------------= GDF 0…5=
CDC16xxF-E ADVANCE INFORMATION
174 March 31, 2003; 6251-606-2AI Micronas
tone decay. It’s possible to change GDF during an active
gong sound (AMA = ’1’).
Table 24–4: Definition of GDF
GDF gong sound duration factor
0H 1
1H 2
2H 4
3H 8
4H 16
5H 32
6H continuous tone
7H
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 175
25. Hardware Options
25.1. Functional Description
Hardware Options are available in several areas to adapt the
IC function to the host system requirements:
– clock signal selection for most of the peripheral modules
from fosc to fosc/217 plus some internal signals (see
Table 25–2 on page 179)
– interrupt source selection for interrupt inputs 0, 1, 5, 6, 7,
10, 13, 14 and 15
– Special Out signal selection for some U- and H-ports
– Rx/Tx polarity selection for SPI and UART modules
– U-port Port Slow Mode selection
The setting of the Hardware Option takes place in two steps:
1. selection is done by programming dedicated address loca-
tions with the desired options’ code
2. activation is done by a read access to these dedicated
address locations at least once after each reset.
Address locations 00FFB8h through 00FFBFh do not allow
random setting. Their respective Hardware Options are hard-
wired and can only be altered by changing a production
mask for this IC. By default all U-Ports have the Port Slow
Option set with the exception of U1.0 to U1.3 (Port Fast
Option set). The Watchdog and Clock Monitor are SW acti-
vated by default.
Future mask ROM derivatives of this IC will not require (but
will tolerate) activation of option settings by read accesses
because ROM as well as options will be hard-wired. Instead,
the manufacturer will automatically process the setting of the
dedicated address locations, as given in the ROM code file,
to set the required mask changes.
To ensure compatible option settings in this IC and mask
ROM derivatives when run with the same ROM code, it is
recommended to always read locations 00FFA0h through
00FFC3h directly after reset. Be aware that the non-pro-
grammable locations 00FFB8h through 00FFBFh may not be
compatible among this IC and the mask ROM derivative.
25.2. Listing of Dedicated Addresses and Corresponding Hardware Options
Table 25–1: Hardware-Option-Dedicated Addresses
7 6 5 4 3 2 1 0
00FFA0 Timer 0 Clock Options
x x x Clock options f1 to f31
00FFA1 PWM0, 3 Clock Options
x x x Clock options f0 to f31 (all)
00FFA2 PWM0, 3 Trigger Options
x x x Clock options f0 to f31 (all)
The high pulse width of the trigger period must be greater than the high pulse width of the clock the PWM is
provided with.
00FFA3 PWM1, 4 Clock Options
x x x Clock options f0 to f31 (all)
00FFA4 PWM1, 4 Trigger Options
x x x Clock options f0 to f31 (all)
The high pulse width of the trigger period must be greater than the high pulse width of the clock the PWM is
provided with.
CDC16xxF-E ADVANCE INFORMATION
176 March 31, 2003; 6251-606-2AI Micronas
00FFA5 PWM2 Clock Options
x x x Clock options f0 to f31 (all)
00FFA6 PWM2 Trigger Options
x x x Clock options f0 to f31 (all)
The high pulse-width of the trigger period must be greater than the high pulse-width of the clock the PWM is
provided with.
00FFA7 Timer 1 Option
x x x Clock options f0 to f31 (all)
00FFA8 Timer 2 Option
x x x Clock options f0 to f31 (all)
00FFA9 CAPCOM Counter Clock Option
x x x Clock options f0 to f31 (all)
00FFAA DIGITbus Clock Option
x x x Clock options f0 to f31 (all)
00FFAB Clock Out 0: Mux0 Prescaler and Clock Option
x x0: Mux out direct
01: Mux out / 1.5
11: Mux out / 2.5
Clock options f0 to f31 (all)
00FFAC Clock Out 1: Prescaler and Clock Option
x x0: direct
01: 1/ 1.5
11: 1/ 2.5
Clock options f0 to f31 (all)
00FFAD LCD Module Prescaler and Clock Option
x x0: direct
01: 1/ 1.5
11: 1/ 2.5
Clock options f0 to f31 (all)
00FFAE SMM, SPI0, SPI1 Clock Prescaler and SMM Clock Option
x x0: direct
01: 1/ 1.5
11: 1/ 2.5
Clock options f0 to f31 (all)
Table 25–1: Hardware-Option-Dedicated Addresses
7 6 5 4 3 2 1 0
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 177
00FFAF SPI0 Input, Output and F0SPI Clock Option 1)
SPI0-D-OUT
0: direct
1: inverted
SPI0-D-IN
0: direct
1: inverted
x Clock options f0 to f31 (all)
1) FFAE, Bits 6 & 5 define also the SPI0 and SPI1 prescaler setting.
00FFB0 SPI1 Input, Output and F1SPI Clock Option 1)
SPI1-D-OUT
0: direct
1: inverted
SPI1-D-IN
0: direct
1: inverted
x Clock options f0 to f31 (all)
1) FFAE, Bits 6 & 5 define also the SPI0 and SPI1 prescaler setting.
00FFB1 F2SPI Clock Options
x x x Clock options f0 to f31 (all)
00FFB2 Clock Out 0: Mux1 Clock Option
x x x Clock options f0 to f31 (all)
00FFB3 Clock Out 0: Mux2 Clock Option
x x x Clock options f0 to f31 (all)
00FFB4 UART0, 2 Input and Output
UART0 Tx
0: direct
1: inverted
UART0 Rx
0: direct
1: inverted
UART2 Tx
0: direct
1: inverted
UART2 Rx
0: direct
1: inverted
xxxx
00FFB5 UART1 Options
UART1 Tx
0: direct
1: inverted
UART1 Rx
0: direct
1: inverted
xxxxxx
00FFB6 Clock Out 0: Mux3 Clock Option
x x x Clock options f0 to f31 (all)
00FFB7 AM Clock Option
x x x Clock options f0 to f31 (all)
00FFB8 Clock Monitor Options
x Wdog & Clk
Monitor:
0: deact. by
Software
1: always
active
xxxxxx
Table 25–1: Hardware-Option-Dedicated Addresses
7 6 5 4 3 2 1 0
CDC16xxF-E ADVANCE INFORMATION
178 March 31, 2003; 6251-606-2AI Micronas
00FFB9 Universal Port 1 Slow/Fast Options
0: Fast mode pin only
1: Slow or fast mode pin possible
00FFBA Universal Port 2 Slow/Fast Options
0: Fast mode pin only
1: Slow or fast mode pin possible
00FFBB Universal Port 3 Slow/Fast Options
0: Fast mode pin only
1: Slow or fast mode pin possible
00FFBC Universal Port 4 Slow/Fast Options
0: Fast mode pin only
1: Slow or fast mode pin possible
00FFBD Universal Port 5 Slow/Fast Options
0: Fast mode pin only
1: Slow or fast mode pin possible
00FFBE Universal Port 6 Slow/Fast Options
0: Fast mode pin only
1: Slow or fast mode pin possible
00FFBF Universal Port 7 Slow/Fast Options
x x x x 0: Fast mode pin only
1: Slow or fast mode pin possible
00FFC0 Interrupt Sources Multiplexer 1 to 4
Mux4:
00 CAN 2
01 SPI 0
10 DMA
11 PINT3-IN
Mux3:
00 PINT3-IN
01 SPI 1
10 UART 1
11 CC1 COMP
Mux2:
00 UART 2
01 P06 COMP
10 SPI 0
11 Timer 1
Mux1:
00 CC0 COMP
01 Timer 2
10 CAN 2
11 Timer 1
00FFC1 Interrupt Sources Multiplexer 5 to 8
Mux8:
00 CC1OR
01 PINT2-IN
10 IR-RTC
11 IR-WAPI
Mux7:
00 CC0OR
01 UART 1
10 IR-RTC
11 IR-WAPI
Mux6:
00 Timer 2
01 DIGITbus
10 UART 2
11 PINT2-IN
Mux5:
00 Timer 2
01 UART 1
10 SPI 1
11 DMA
Table 25–1: Hardware-Option-Dedicated Addresses
7 6 5 4 3 2 1 0
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 179
00FFC2 Interrupt Sources Multiplexer 9 and Port Multiplexer
H-Port 1.1
0: SME
1: PWM2
x H-Port 1.0
0: SME
1: PWM0
PINT3-IN
0: at U5.6
1: at U5.7
xxMux9:
00 CAN 1
01 UART 1
10 IR-RTC
11 IR-WAPI
00FFC3 AM Trigger Option
x x x Clock options f0 to f31 (all)
Table 25–1: Hardware-Option-Dedicated Addresses
7 6 5 4 3 2 1 0
Table 25–2: Clock Option Selection Code
Clock Option
Number Clock Signal Selection Code
f0 fOSC/20 xxx0.0000
f1 fOSC/21 xxx0.0001
f2 fOSC/22 xxx0.0010
f3 fOSC/23 xxx0.0011
f4 fOSC/24 xxx0.0100
f5 fOSC/25 xxx0.0101
f6 fOSC/26 xxx0.0110
f7 fOSC/27 xxx0.0111
f8 fOSC/28 xxx0.1000
f9 fOSC/29 xxx0.1001
f10 fOSC/210 xxx0.1010
f11 fOSC/211 xxx0.1011
f12 fOSC/212 xxx0.1100
f13 fOSC/213 xxx0.1101
f14 fOSC/214 xxx0.1110
f15 fOSC/215 xxx0.1111
f16 fOSC/216 xxx1.0000
f17 fOSC/217 xxx1.0001
If the leading “x” in the Clock sampling table are not used
for the purpose of coding other options, they must be
replaced by zeros.
1) Clock option f22 is only available if the Stepper Motor
Module has been enabled by the standby bit.
f18 VSS xxx1.0010
f19 Timer 0 xxx1.0011
f20 VSS xxx1.0100
f21 fSM xxx1.0101
f22 1)fSM/2
8 xxx1.0110
f23 fCC0IN xxx1.0111
f24 fCC1IN xxx1.1000
f25, 26, 27 VSS xxx1.1001 ...
f28 fOSC/22 xxx1.1100
f29, 30 VSS xxx1.1101 ...
f31 fOSC/210 xxx1.1111
Table 25–2: Clock Option Selection Code
Clock Option
Number Clock Signal Selection Code
If the leading “x” in the Clock sampling table are not used
for the purpose of coding other options, they must be
replaced by zeros.
1) Clock option f22 is only available if the Stepper Motor
Module has been enabled by the standby bit.
CDC16xxF-E ADVANCE INFORMATION
180 March 31, 2003; 6251-606-2AI Micronas
26. Register Cross Reference Table V2.1
26.1. CAN Registers, memory page 1C
Address
(hex) Mnemonic Block
1C00 CAN0CTR CAN0
1C01 CAN0STR
1C02 CAN0ESTR
1C03 CAN0IDX
1C04 CAN0IDM
1C05
1C06
1C07
1C08 CAN0BT1
1C09 CAN0BT2
1C0A CAN0BT3
1C0B CAN0ICR
1C0C CAN0OCR
1C0D CAN0TEC
1C0E CAN0REC
1C0F CAN0ESM
1C10 CAN0CTIM
1C11
1C40 CAN1CTR CAN1
1C41 CAN1STR
1C42 CAN1ESTR
1C43 CAN1IDX
1C44 CAN1IDM
1C45
1C46
1C47
1C48 CAN1BT1
1C49 CAN1BT2
1C4A CAN1BT3
1C4B CAN1ICR
1C4C CAN1OCR
1C4D CAN1TEC
1C4E CAN1REC
1C4F CAN1ESM
1C50 CAN1CTIM
1C51
Address
(hex) Mnemonic Block
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 181
26.2. I/O Register 1, memory page 1E
1C80 CAN2CTR CAN2
1C81 CAN2STR
1C82 CAN2ESTR
1C83 CAN2IDX
1C84 CAN2IDM
1C85
1C86
1C87
1C88 CAN2BT1
1C89 CAN2BT2
1C8A CAN2BT3
1C8B CAN2ICR
1C8C CAN2OCR
1C8D CAN2TEC
1C8E CAN2REC
1C8F CAN2ESM
1C90 CAN2CTIM
1C91
Address
(hex) Mnemonic Block
Address
(hex) Mnemonic Block
1E64 PAR0 Patch Module
1E65 PAR1
1E66 PAR2
1E67 PDR
1E68 PER0
1E69 PER1
CDC16xxF-E ADVANCE INFORMATION
182 March 31, 2003; 6251-606-2AI Micronas
1E70 WUS Power Saving Module
1E71
1E74 SSR
1E75
1E76
1E78 SSC
1E79
1E7A
1E7C RTC
1E7D
1E7E
1E80 WPM0
1E81 WPM2
1E82 WPM4
1E83 WPM6
1E84 WPM8
1E88 WSC
1E90 OSC
1E94 RTCC
1E98 POL
1E99
1E9C SMX
1EA0 MULCAND Multiplier
1EA1 MULPLIER
1EA2 MULPROD
1EA3
Address
(hex) Mnemonic Block
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 183
26.3. I/O Register 0, memory page 1F
Address
(hex) Mnemonic Block
1F00 CSW0 Core Logic
1F01 CR
1F02 ERMC ERM
1F08 SR0 Core Logic
1F09 SR1
1F0A SR2
1F0B SR3
1F0C DBG Debug Register
1F0F ABR Memory Banking
1F10 SPI0D SPI0
1F11 SPI0M
1F12 SPI1D SPI1
1F13 SPI1M
1F14 CO0SEL Core Logic
1F15 CO1SEL
1F18 UA1D UART1
1F19 UA1C
1F1A UA1BR0
1F1B UA1BR1
1F1C UA1IM
1F1D UA1CA
1F1E UA1IF
1F1F IRE Interrupt Controller
1F20 IRC
1F21 IRRET
1F22 IRPRI10
1F23 IRPRI32
1F24 IRPRI54
1F25 IRPRI76
1F26 IRPRI98
1F27 IRPRIBA
1F28 IRPRIDC
1F29 IRPRIFE
1F2A IRP
1F2B IRPM0
1F2C IRPP
1F2D AMAS Audio Module
1F2E AMF
1F2F AMDEC
1F30 U2D Universal Port 2
1F32 U2SEG10
1F33 U2M10
1F34 U2SEG32
1F35 U2M32
1F36 U2SEG54
1F37 U2M54
1F38 U2SEG76
1F39 U2M76
1F4E TIM0 Timer 0
1F4F
1F50 PWM0 PWM
1F51 PWM1
1F52 PWM2
Address
(hex) Mnemonic Block
CDC16xxF-E ADVANCE INFORMATION
184 March 31, 2003; 6251-606-2AI Micronas
1F54 TIM1 Timer 1, 2
1F55 TIM2
1F5A SMVC Stepper Motor Module
1F5B SMVSIN
1F5C SMVCOS
1F5D SMVCMP
1F5E PWM3 PWM
1F5F PWM4
1F60 CSW1 Core Logic
1F61 CSW2
1F64 UA2D UART2
1F65 UA2C
1F66 UA2BR0
1F67 UA2BR1
1F68 UA2IM
1F69 UA2CA
1F6A UA2IF
1F6C CC0M Capture Compare Module
1F6D CC0I
1F6E CC0
1F6F
1F70 CC1M
1F71 CC1I
1F72 CC1
1F73
1F74 CC2M
1F75 CC2I
1F76 CC2
1F77
1F7C CCC
1F7D
1F7E P0D Analog Input Port 0
Address
(hex) Mnemonic Block
1F80 H0NS High Current Port 0
1F81 H0TRI
1F82 H0D
1F84 H1NS High Current Port 1
1F85 H1TRI
1F86 H1D
1F88 H2NS High Current Port 2
1F89 H2TRI
1F8A H2D
1F90 H3NS High Current Port 3
1F91 H3TRI
1F92 H3D
1F98 U1D Universal Port 1
1F99 U1SEG10
1F9A U1SEG32
1F9B U1M30
1F9C U1SEG54
1F9D U1M54
1F9E U1SEG76
1F9F U1M76
1FA0 UA0D UART0
1FA1 UA0C
1FA2 UA0BR0
1FA3 UA0BR1
1FA4 UA0IM
1FA5 UA0CA
1FA6 UA0IF
1FA8 AD0 AD Converter
1FA9 AD1
Address
(hex) Mnemonic Block
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 185
1FAC U3D Universal Port 3
1FAE U3SEG10
1FAF U3M10
1FB0 U3SEG32
1FB1 U3M32
1FB2 U3SEG54
1FB3 U3M54
1FB4 U3SEG76
1FB5 U3M76
1FB8 U4D Universal Port 4
1FBA U4SEG10
1FBB U4M10
1FBC U4SEG32
1FBD U4M32
1FBE U4SEG54
1FBF U4M54
1FC0 U4SEG76
1FC1 U4M76
1FC4 U5D Universal Port 5
1FC6 U5SEG10
1FC7 U5M10
1FC8 U5SEG32
1FC9 U5M32
1FCA U5SEG54
1FCB U5M54
1FCC U5SEG76
1FCD U5M76
Address
(hex) Mnemonic Block
1FD0 U6D Universal Port 6
1FD2 U6SEG10
1FD3 U6M10
1FD4 U6SEG32
1FD5 U6M32
1FD6 U6SEG54
1FD7 U6M54
1FD8 U6SEG76
1FD9 U6M76
1FDC U7D Universal Port 7
1FDE U7SEG10
1FDF U7M10
1FE0 U7SEG32
1FE1 U7M32
1FE8 DCS DMA
1FE9 DIC
1FEA DSA
1FEB
1FEC
1FED DEA
1FEE
1FEF
1FF0 DGC0 DIGITbus
1FF1 DGC1
1FF2 DGS0
1FF3 DGRTMD
1FF4 DGTL
1FF5 DGS1TA
1FF6 DGTD
1FF7 DGRTMA
1FFD TST3 TST
1FFE TST1
1FFF TST2
Address
(hex) Mnemonic Block
CDC16xxF-E ADVANCE INFORMATION
186 March 31, 2003; 6251-606-2AI Micronas
27. Register Quick Reference
Table 27–1: A/D Converter
Mnemonic Register Name Addr.
(hex) Register Configuration Section
AD0 ADC Register 0 1FA8 12.2.
AD1 ADC Register 1 1FA9
76543210
rEOC CMPO x x x x AN1 AN0
wTSAMP CHANNEL
00xx0000Res
rAN9 AN8 AN7 AN6 AN5 AN4 AN3 AN2
Table 27–2: Audio Module
Mnemonic Register Name Addr.
(hex) Register Configuration Section
AMAS Audio Module Ampli-
tude & Status Regis-
ter
1F2D 24.2.
AMF Audio Module Fre-
quency Register 1F2E
AMDEC Audio Module Decre-
ment Register 1F2F
76543210
wInitial Amplitude
rAMAxxxxxxx
0xxxxxxxRes
wx Sound Frequency
-0000000Res
wAMMCA x x x x GDF
0----000Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 187
Table 27–3: Capture-Compare-Unit
Mnemonic Register Name Addr.
(hex) Register Configuration Section
CC0M CAPCOM 0 Mode
Register 1F6C 15.2.
CC0I CAPCOM 0 Interrupt
Register 1F6D
CC0 CAPCOM 0 Capture/
Compare Register
low byte
1F6E
CAPCOM 0 Capture/
Compare Register
high byte
1F6F
CC1M CAPCOM 1 Mode
Register 1F70
CC1I CAPCOM 1 Interrupt
Register 1F71
CC1 CAPCOM 1 Capture/
Compare Register
low byte
1F72
CAPCOM 1 Capture/
Compare Register
high byte
1F73
CC2M CAPCOM 2 Mode
Register 1F74
CC2I CAPCOM 2 Interrupt
Register 1F75
76543210
r/w MSK MSK MSK FOL OAM IAM
00000000Res
r/w CAP CMP OFL LAC RCR x x x
00000000Res
rRead low byte of capture register and lock it.
wWrite low byte of compare register and lock it.
11111111Res
rRead high byte of capture register and unlock it.
wWrite high byte of compare register and unlock it.
11111111Res
r/w MSK MSK MSK FOL OAM IAM
00000000Res
r/w CAP CMP OFL LAC RCR x x x
00000000Res
rRead low byte of capture register and lock it.
wWrite low byte of compare register and lock it.
11111111Res
rRead high byte of capture register and unlock it.
wWrite high byte of compare register and unlock it.
11111111Res
r/w MSK MSK MSK FOL OAM IAM
00000000Res
r/w CAP CMP OFL LAC RCR x x x
00000000Res
CDC16xxF-E ADVANCE INFORMATION
188 March 31, 2003; 6251-606-2AI Micronas
CC2 CAPCOM 2 Capture/
Compare Register
low byte
1F76 15.2.
CAPCOM 2 Capture/
Compare Register
high byte
1F77
CCC CAPCOM Counter
low byte 1F7C
CAPCOM Counter
high byte 1F7D
Table 27–3: Capture-Compare-Unit
Mnemonic Register Name Addr.
(hex) Register Configuration Section
76543210
rRead low byte of capture register and lock it.
wWrite low byte of compare register and lock it.
11111111Res
rRead high byte of capture register and unlock it.
wWrite high byte of compare register and unlock it.
11111111Res
rRead low byte and lock CCC
00000000Res
rRead high byte and unlock CCC
00000000Res
Table 27–4: Controller Area Network 0
Mnemonic Register Name Addr.
(hex) Register Configuration Section
CAN0CTR Control Register 1C00 21.2.
CAN0STR Status Register 1C01
CAN0ESTR Error Status Register 1C02
CAN0IDX Interrupt Index Reg-
ister 1C03
CAN0IDM Identifier Mask Reg-
ister 1C04
1C05
1C06
1C07
76543210
r/w HLT SLP GRSC EIE GRIE GTIE BOST rsvd
1000000xRes
rHACK BOFF EPAS ERS rsvd rsvd rsvd rsvd
1000xxxxRes
r/w GDM CTOV ECNT BIT STF CRC FRM ACK
00000000Res
r/w Interrupt Index
11111111Res
r/w Identifier Mask Bits 29 to 21 low
r/w Identifier Mask Bits 20 to 13
r/w Identifier Mask Bits 12 to 5
r/w Identifier Mask Bits 4 to 0 x x x high
00000000Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 189
CAN0BT1 Bit Timing Register 1 1C08 21.2.
CAN0BT2 Bit Timing Register 2 1C09
CAN0BT3 Bit Timing Register 3 1C0A
CAN0ICR Input Control Regis-
ter 1C0B
CAN0OCR Output Control Reg-
ister 1C0C
CAN0TEC Transmit Error
Counter 1C0D
CAN0REC Receive Error
Counter 1C0E
CAN0ESM Error Status Mask
Register 1C0F
CAN0CTIM Capture Timer 1C10
1C11
Table 27–4: Controller Area Network 0
Mnemonic Register Name Addr.
(hex) Register Configuration Section
76543210
r/w MSAM SYN BPR BPR BPR BPR BPR BPR
00000000Res
r/w rsvd TSEG2 TSEG2 TSEG2 TSEG1 TSEG1 TSEG1 TSEG1
00000000Res
r/w rsvd rsvd rsvd rsvd rsvd SJW SJW SJW
xxxxx000Res
r/w rsvd rsvd rsvd rsvd rsvd XREF REF1 REF0
xxxxx000Res
r/w rsvd rsvd rsvd rsvd rsvd rsvd rsvd ITX
xxxxxxx0Res
rCounter Bit 7 to 0
00000000Res
rx Counter Bit 6 to 0
x0000000Res
r/w EGDM ECTV EECT EBIT ESTF ECRC EFRM EACK
11111111Res
rTimer Bit 7 to 0 low
rTimer Bit 15 to 8 high
00000000Res
Table 27–5: Controller Area Network 1
Mnemonic Register Name Addr.
(hex) Register Configuration Section
CAN1CTR Control Register 1C40 21.2.
CAN1STR Status Register 1C41
76543210
r/w HLT SLP GRSC EIE GRIE GTIE BOST rsvd
1000000xRes
rHACK BOFF EPAS ERS rsvd rsvd rsvd rsvd
1000xxxxRes
CDC16xxF-E ADVANCE INFORMATION
190 March 31, 2003; 6251-606-2AI Micronas
CAN1ESTR Error Status Register 1C42 21.2.
CAN1IDX Interrupt Index Reg-
ister 1C43
CAN1IDM Identifier Mask Reg-
ister 1C44
1C45
1C46
1C47
CAN1BT1 Bit Timing Register 1 1C48
CAN1BT2 Bit Timing Register 2 1C49
CAN1BT3 Bit Timing Register 3 1C4A
CAN1ICR Input Control Regis-
ter 1C4B
CAN1OCR Output Control Reg-
ister 1C4C
CAN1TEC Transmit Error
Counter 1C4D
CAN1REC Receive Error
Counter 1C4E
CAN1ESM Error Status Mask
Register 1C4F
CAN1CTIM Capture Timer 1C50
1C51
Table 27–5: Controller Area Network 1
Mnemonic Register Name Addr.
(hex) Register Configuration Section
76543210
r/w GDM CTOV ECNT BIT STF CRC FRM ACK
00000000Res
r/w Interrupt Index
11111111Res
r/w Identifier Mask Bits 29 to 21 low
r/w Identifier Mask Bits 20 to 13
r/w Identifier Mask Bits 12 to 5
r/w Identifier Mask Bits 4 to 0 x x x high
00000000Res
r/w MSAM SYN BPR BPR BPR BPR BPR BPR
00000000Res
r/w rsvd TSEG2 TSEG2 TSEG2 TSEG1 TSEG1 TSEG1 TSEG1
00000000Res
r/w rsvd rsvd rsvd rsvd rsvd SJW SJW SJW
xxxxx000Res
r/w rsvd rsvd rsvd rsvd rsvd XREF REF1 REF0
xxxxx000Res
r/w rsvd rsvd rsvd rsvd rsvd rsvd rsvd ITX
xxxxxxx0Res
rCounter Bit 7 to 0
00000000Res
rx Counter Bit 6 to 0
x0000000Res
r/w EGDM ECTV EECT EBIT ESTF ECRC EFRM EACK
11111111Res
rTimer Bit 7 to 0 low
rTimer Bit 15 to 8 high
00000000Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 191
Table 27–6: Controller Area Network 2
Mnemonic Register Name Addr.
(hex) Register Configuration Section
CAN2CTR Control Register 1C80 21.2.
CAN2STR Status Register 1C81
CAN2ESTR Error Status Register 1C82
CAN2IDX Interrupt Index Reg-
ister 1C83
CAN2IDM Identifier Mask Reg-
ister 1C84
1C85
1C86
1C87
CAN2BT1 Bit Timing Register 1 1C88
CAN2BT2 Bit Timing Register 2 1C89
CAN2BT3 Bit Timing Register 3 1C8A
CAN2ICR Input Control Regis-
ter 1C8B
CAN2OCR Output Control Reg-
ister 1C8C
CAN2TEC Transmit Error
Counter 1C8D
CAN2REC Receive Error
Counter 1C8E
76543210
r/w HLT SLP GRSC EIE GRIE GTIE BOST rsvd
1000000xRes
rHACK BOFF EPAS ERS rsvd rsvd rsvd rsvd
1000xxxxRes
r/w GDM CTOV ECNT BIT STF CRC FRM ACK
00000000Res
r/w Interrupt Index
11111111Res
r/w Identifier Mask Bits 29 to 21 low
r/w Identifier Mask Bits 20 to 13
r/w Identifier Mask Bits 12 to 5
r/w Identifier Mask Bits 4 to 0 x x x high
00000000Res
r/w MSAM SYN BPR BPR BPR BPR BPR BPR
00000000Res
r/w rsvd TSEG2 TSEG2 TSEG2 TSEG1 TSEG1 TSEG1 TSEG1
00000000Res
r/w rsvd rsvd rsvd rsvd rsvd SJW SJW SJW
xxxxx000Res
r/w rsvd rsvd rsvd rsvd rsvd XREF REF1 REF0
xxxxx000Res
r/w rsvd rsvd rsvd rsvd rsvd rsvd rsvd ITX
xxxxxxx0Res
rCounter Bit 7 to 0
00000000Res
rx Counter Bit 6 to 0
x0000000Res
CDC16xxF-E ADVANCE INFORMATION
192 March 31, 2003; 6251-606-2AI Micronas
CAN2ESM Error Status Mask
Register 1C8F
CAN2CTIM Capture Timer 1C90 21.2.
1C91
Table 27–6: Controller Area Network 2
Mnemonic Register Name Addr.
(hex) Register Configuration Section
76543210
r/w EGDM ECTV EECT EBIT ESTF ECRC EFRM EACK
11111111Res
rTimer Bit 7 to 0 low
rTimer Bit 15 to 8 high
00000000Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 193
Table 27–7: Core Logic
Mnemonic Register Name Addr.
(hex) Register Configuration Section
CSW0 Clock, Supply and
Watchdog Register 0 1F00 6.
CR Control Register 1F01
SR0 Standby Register 0 1F08
SR1 Standby Register 1 1F09
SR2 Standby Register 2 1F0A
SR3 Standby Register 3 1F0B
CO0SEL Clock Out 0 Selec-
tion 1F14
CO1SEL Clock Out 1Selection 1F15
CSW1 Clock, Supply and
Watchdog Register 1 1F60
CSW2 Clock, Supply and
Watchdog Register 2 1F61
76543210
wxxxxxxxCMA
xxxxxxx1Res
r/w RESLNG TSTTOG x MFM TSTROM IROM IRAM ICPU ROM
r/w RESLNG TSTTOG EBTRI MFM FLASH IROM IRAM ICPU Emu
Value of 00FFF3h Res
r/w SM PWM1 PWM0 UART2 SPI1 CAN0 CCC SPI0
00000000Res
r/w LCD CPUFST PSLW UART0 ADC P0DIN TIM1 ERM
01000000Res
r/w TIM2 PWM3 PWM2 UART1 PWM4 DGB EXTIR ABM
00000000Res
r/w x x x XTAL WAID FCLO CAN2 CAN1
xxx10
*) 000Res
wxxxxxxCO01CO00
xxxxxx00Res
wxxxxxxxCO10
xxxxxxx0Res
rxxxxxxxWDRES
wWatchdog Time and Trigger Value
11111111Res
rTST xWKID FHR CLM PIN POR x0
wx x 0 FHR 000x0
-x------Res
CDC16xxF-E ADVANCE INFORMATION
194 March 31, 2003; 6251-606-2AI Micronas
Table 27–8: Debug Register
Mnemonic Register Name Addr.
(hex) Register Configuration Section
DBG Debug Register 1F0C 8.3.5.
Table 27–9: DIGITbus
Mnemonic Register Name Addr.
(hex) Register Configuration Section
DGC0 Control Register 0 1FF0 23.4.
DGC1 Control Register 1 1FF1
DGS0 Status Register 0 1FF2
DGRTMD Rx Length & Tx More
Data Register 1FF3
DGTL Tx Length Register 1FF4
DGS1TA Status 1 & Tx
Address Register 1FF5
DGTD Tx Data Register 1FF6
DGRTMA Rx Field & Tx More
Address Register 1FF7
76543210
r/w xxxxxxxDCS0
xxxxxxx0POR
76543210
r/w RUN GBC ACT RXO X PSC 2 to 0
0000x000Res
r/w INTE ENEM ENOF x PHASE
000x0000Res
wxxxTGVPVERRxARB
rRDL NEM NOF
x01000x0Res
wTransmit More Data
rRDL NEM FTYP EOF x LEN2 to 0
00xxxxxxRes
wxFLUSHx x x LEN2 to 0
x0xxx000Res
rxEMPTYxxxxxx
x1xxxxxxRes
wTransmit Address
rSTATE PW5 to 0
01000000Res
wTransmit Data
xxxxxxxxRes
wTransmit More Address
rReceive Field
xxxxxxxxRes
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 195
Table 27–10: DMA
Mnemonic Register Name Addr.
(hex) Register Configuration Section
DCS DMA Control and
Status Register 1FE8 18.2.
DIC DMA Initial Configu-
ration Register 1FE9
DSA DMA Start Address 1FEA
1FEB
1FEC
DEA DMA End Address 1FED
1FEE
1FEF
76543210
r/w x x x DTA DSI DCC STP BSY
00000000Res
wx WS2 WS1 WS0 x x x WSA
00000000Res
wBit 7 to 0
wBit 15 to 8
wBit 23 to 16
00000000Res
wBit 7 to 0
wBit 15 to 8
wBit 23 to 16
00000000Res
Table 27–11: EMI Reduction Module
Mnemonic Register Name Addr.
(hex) Register Configuration Section
ERMC EMI Reduction Mod-
ule Control Register 1F02 4.4.8.
76543210
rxxxxxxxCLKSEL
wxx00000CLKSEL
xx001000Res
CDC16xxF-E ADVANCE INFORMATION
196 March 31, 2003; 6251-606-2AI Micronas
Table 27–12: High Current Port 0
Mnemonic Register Name Addr.
(hex) Register Configuration Section
H0NS High Current Port 0
Normal/Special Reg-
ister
1F80 11.5.
H0TRI High Current Port 0
Tristate Register 1F81
H0D High Current Port 0
Data Register 1F82
76543210
wx x N/S5 N/S4 N/S3 N/S2 N/S1 N/S0
00000000Res
wx x TRI5 TRI4 TRI3 TRI2 TRI1 TRI0
00000000Res
r/w x x D5 D4 D3 D2 D1 D0
00000000Res
Table 27–13: High Current Port 1
Mnemonic Register Name Addr.
(hex) Register Configuration Section
H1NS High Current Port 1
Normal/Special Reg-
ister
1F84 11.5.
H1TRI High Current Port 1
Tristate Register 1F85
H1D High Current Port 1
Data Register 1F86
76543210
wx x N/S5 N/S4 N/S3 N/S2 N/S1 N/S0
00000000Res
wx x TRI5 TRI4 TRI3 TRI2 TRI1 TRI0
00000000Res
r/w x x D5 D4 D3 D2 D1 D0
00000000Res
Table 27–14: High Current Port 2
Mnemonic Register Name Addr.
(hex) Register Configuration Section
H2NS High Current Port 2
Normal/Special Reg-
ister
1F88 11.5.
H2TRI High Current Port 2
Tristate Register 1F89
H2D High Current Port 2
Data Register 1F8A
76543210
wx x N/S5 N/S4 N/S3 N/S2 N/S1 N/S0
00000000Res
wx x TRI5 TRI4 TRI3 TRI2 TRI1 TRI0
00000000Res
r/w x x D5 D4 D3 D2 D1 D0
00000000Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 197
Table 27–15: High Current Port 3
Mnemonic Register Name Addr.
(hex) Register Configuration Section
H3NS High Current Port 3
Normal/Special Reg-
ister
1F90 11.5.
H3TRI High Current Port 3
Tristate Register 1F91
H3D High Current Port 3
Data Register 1F92
76543210
wx x N/S5 N/S4 N/S3 N/S2 N/S1 N/S0
00000000Res
wx x TRI5 TRI4 TRI3 TRI2 TRI1 TRI0
00000000Res
r/w x x D5 D4 D3 D2 D1 D0
00000000Res
Table 27–16: Interrupt Controller
Mnemonic Register Name Addr.
(hex) Register Configuration Section
IRE Interrupt Enable
Register
IRC Interrupt Control
Register 1F20 10.2.
IRRET Interrupt Pending
and Return Register 1F21
IRPRI10 Interrupt Priority,
Inputs 0 and 1 1F22
IRPRI32 Interrupt Priority,
Inputs 2 and 3 1F23
IRPRI54 Interrupt Priority,
Inputs 4 and 5 1F24
IRPRI76 Interrupt Priority,
Inputs 6 and 7 1F25
76543210
wA write access enables interrupts according to priority setting (same
effect as setting IRC.DINT)
00000000Res
rx x x x DAINT DINT x x
wx x x RESET DAINT DINT A1INT CLEAR
x11xxRes
rIPF7 IPF6 IPF5 IPF4 IPF3 IPF2 IPF1 IPF0
wA write access signals to the Interrupt Controller that the current
request has been served
00000000Res
r/w PRIO1 PRIO0
00000000Res
r/w PRIO3 PRIO2
00000000Res
r/w PRIO5 PRIO4
00000000Res
r/w PRIO7 PRIO6
00000000Res
CDC16xxF-E ADVANCE INFORMATION
198 March 31, 2003; 6251-606-2AI Micronas
IRPRI98 Interrupt Priority,
Inputs 8 and 9 1F26 10.2.
IRPRIBA Interrupt Priority,
Inputs 10 and 11 1F27
IRPRIDC Interrupt Priority,
Inputs 12 and 13 1F28
IRPRIFE Interrupt Priority,
Inputs 14 and 15 1F29
IRP Interrupt Pending
Register 1F2A
IRPM0 Interrupt Port Mode
Register 0 1F2B
IRPP Interrupt Port Pres-
caler Register 1F2C
Table 27–16: Interrupt Controller
Mnemonic Register Name Addr.
(hex) Register Configuration Section
76543210
r/w PRIO9 PRIO8
00000000Res
r/w PRIO11 PRIO10
00000000Res
r/w PRIO13 PRIO12
00000000Res
r/w PRIO15 PRIO14
00000000Res
rIPF15 IPF14 IPF13 IPF12 IPF11 IPF10 IPF9 IPF8
00000000Res
wPit3 Pit2 PIT1 PIT0
00000000Res
wxxxxxxP1INT32P0INT4
00Res
Table 27–17: Memory Banking
Mnemonic Register Name Addr.
(hex) Register Configuration Section
ABR Alternative Banking
Register 1F0F 5.2.2.
76543210
r/w Alternative Bank Address
00000001Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 199
Table 27–18: Multiplier
Mnemonic Register Name Addr.
(hex) Register Configuration Section
MULCAND Multiplicand 1EA0
MULPLIER Multiplier 1EA1
MULPROD Multiplication
Product 1EA3
1EA2
Table 27–19: Patch Module
Mnemonic Register Name Addr.
(hex) Register Configuration Section
PAR0 Patch Address Reg-
ister 0 1E64 9.2.
PAR1 Patch Address Reg-
ister 1 1E65
PAR2 Patch Address Reg-
ister 2 1E66
PDR Patch Data Register 1E67
PER0 Patch Enable Regis-
ter 0 1E68
PER1 Patch Enable Regis-
ter 1 1E69
76543210
r/w multiplicand
xxxxxxxxRes
r/w multiplier
xxxxxxxxRes
rProduct high byte 1
rProduct low byte 0
xRes
76543210
wPA7 PA6 PA5 PA4 PA3 PA2 PA1 PA0
11111111Res
wPA15 PA14 PA13 PA12 PA11 PA10 PA9 PA8
11111111Res
wPA23 PA22 PA21 PA20 PA19 PA18 PA17 PA16
11111111Res
wPD7 PD6 PD5 PD4 PD3 PD2 PD1 PD0
00000000Res
wPSEL6 PSEL5 PSEL4 PSEL3 PSEL2 PSEL1 PSEL0 PMEN
00000000Res
wx x x x x PSEL9 PSEL8 PSEL7
xxxxx000Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 201
Table 27–21: Power Saving Module
Mnemonic Register Name Addr.
(hex) Register Configuration Section
WUS Wake-Up Source
Register 1E71 8.2.
1E70
SSR Sub Second Reload
Register 1E77
1E76
1E75
1E74
SSC Sub Second Counter 1E7B
1E7A
1E79
1E78
RTC Real Time Counter 1E7F
1E7E
1E7D
1E7C
WPM0 Wake Port Mode
Register 1E80
WPM2 1E81
WPM4 1E82
WPM6 1E83
WPM8 1E84
WSC Wake Source Con-
trol 1E88
OSC Oscillator Source
Register 1E90
RTCC RTC Control Regis-
ter 1E94
76543210
r/w RTCxxxxxWP9WP81
r/w WP7 WP6 WP5 WP4 WP3 WP2 WP1 WP0 0
No HW reset Res
r/w xxxxxxxx3
r/w xxxx Bit 19 to 16 2
r/w Bit 15 to 8 1
r/w Bit 7 to 0 0
No HW reset Res
rxxxxxxxx3
rxxxx Bit 19 to 16 2
rBit 15 to 8 1
rBit 7 to 0 0
No HW reset Res
r/w xxxxxxxx3
r/w xxx HR 2
r/w xx MIN 1
r/w x x SEC 0
No HW reset Res
r/w x MOD1 x MOD0 0
No HW reset Res
r/w xxxxxASTRTCP0
0x00 after UVDD power on Res
r/w RC XK XM x LD PRE SRC 0
1 1 1 No HW reset Res
r/w x x x SEL 0
No HW reset Res
CDC16xxF-E ADVANCE INFORMATION
202 March 31, 2003; 6251-606-2AI Micronas
POL Polling Register 1E99
1E98
SMX Signal Multiplexer
Register 1E9C
Table 27–22: Pulse Width Modulator
Mnemonic Register Name Addr.
(hex) Register Configuration Section
PWM0 PWM 0 Register 1F50 14.2.
PWM1 PWM 1 Register 1F51
PWM2 PWM 2 Register 1F52
PWM3 PWM 3 Register 1F5E
PWM4 PWM 4 Register 1F5F
Table 27–23: Serial Synchronous Peripheral Interface 1
Mnemonic Register Name Addr.
(hex) Register Configuration Section
SPI1D SPI 1 Data Register 1F12 19.2.
SPI1M SPI 1 Mode Register 1F13
Table 27–21: Power Saving Module
Mnemonic Register Name Addr.
(hex) Register Configuration Section
76543210
r/w xCLKx PER 1
r/w ENA OE x DEL 0
0x00 Res
r/w BYP x x x x MUX 0
0x00 Res
76543210
wPulse width value
00000000Res
76543210
r/w Bit 7 to 0 of Rx/Tx Data
00000000Res
r/w BIT8 LEN9 RXSEL INTERN NEDGE x CSF
00000000Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 203
Table 27–24: Stepper Motor Module
Mnemonic Register Name Addr.
(hex) Register Configuration Section
SMVC SMM Control Regis-
ter 1F5A 16.2.
SMVSIN SMM Sine Register 1F5B
SMVCOS SMM Cosine Regis-
ter 1F5C
SMVCMP SMM Comparator
Register 1F5D
Table 27–25: Test Registers
Mnemonic Register Name Addr.
(hex) Register Configuration Section
TST3 Test Register 3 1FFD 6.4.
TST1 Test Register 1 1FFE
TST2 Test Register 2 1FFF
76543210
wx x SEL x QUAD
xx000x00Res
rxxxxxxxBUSY
w8bit Sine Value
00000000Res
w8bit Cosine Value
00000000Res
r/w x x ACRD ACRB x ACRE ACRC ACRA
xx00x000Res
76543210
wFor testing purposes only
00000000Res
wFor testing purposes only
00000000Res
wFor testing purposes only
00000000Res
CDC16xxF-E ADVANCE INFORMATION
204 March 31, 2003; 6251-606-2AI Micronas
Table 27–26: Timer
Mnemonic Register Name Addr.
(hex) Register Configuration Section
TIM0L Timer 0 low byte 1F4E 13.
TIM0H Timer 0 high byte 1F4F
TIM1 Timer 1 Register 1F54
TIM2 Timer 2 Register 1F55
76543210
rRead low byte of down-counter and latch high byte
wWrite low byte of reload value and reload down-counter
11111111Res
rLatched high byte of down-counter
wHigh byte of reload value
11111111Res
wReload value
00000000Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 205
Table 27–27: Universal Asynchronous Receiver Transmitter 0
Mnemonic Register Name Addr.
(hex) Register Configuration Section
UA0D UART 0 Data Regis-
ter 1FA0 20.3.
UA0C UART 0 Control and
Status Register 1FA1
UA0BR0 UART 0 Baudrate
Register low byte 1FA2
UA0BR1 UART 0 Baudrate
Register high byte 1FA3
UA0IM UART 0 Interrupt
Mask Register 1FA4
UA0CA UART 0 Compare
Address Register 1FA5
UA0IF UART 0 Interrupt
Flag Register 1FA6
76543210
r Receive register
wTransmit register
xxxxxxxxRes
rRBUSY BRKD FRER OVRR PAER EMPTY FULL TBUSY
0xx0x100Res
wxxxxSTPBODDPARLEN
xxxx0000Res
wBit 7 to 0 of Baud Rate
00000000Res
wx x x Bit 12 to 8 of Baud Rate
- - -00000Res
wx x x x x ADR BRK RCVD
-----000Res
wBit 7 to 0 of address
00000000Res
rTest Test Test Test Test ADR BRK RCVD
-----x00Res
CDC16xxF-E ADVANCE INFORMATION
206 March 31, 2003; 6251-606-2AI Micronas
Table 27–28: Universal Asynchronous Receiver Transmitter 1
Mnemonic Register Name Addr.
(hex) Register Configuration Section
UA1D UART 1 Data Regis-
ter 1F18 20.3.
UA1C UART 1 Control and
Status Register 1F19
UA1BR0 UART 1 Baudrate
Register low byte 1F1A
UA1BR1 UART 1 Baudrate
Register high byte 1F1B
UA1IM UART 1 Interrupt
Mask Register 1F1C
UA1CA UART 1 Compare
Address Register 1F1D
UA1IF UART 1 Interrupt
Flag Register 1F1E
76543210
r Receive register
wTransmit register
xxxxxxxxRes
rRBUSY BRKD FRER OVRR PAER EMPTY FULL TBUSY
0xx0x100Res
wxxxxSTPBODDPARLEN
xxxx0000Res
wBit 7 to 0 of Baud Rate
00000000Res
wx x x Bit 12 to 8 of Baud Rate
- - -00000Res
wx x x x x ADR BRK RCVD
-----000Res
wBit 7 to 0 of address
00000000Res
rTest Test Test Test Test ADR BRK RCVD
-----x00Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 207
Table 27–29: Universal Asynchronous Receiver Transmitter 2
Mnemonic Register Name Addr.
(hex) Register Configuration Section
UA2D UART 2 Data Regis-
ter 1F64 20.3.
UA2C UART 2 Control and
Status Register 1F65
UA2BR0 UART 2 Baud rate
Register low byte 1F66
UA2BR1 UART 2 Baud rate
Register high byte 1F67
UA2IM UART 2 Interrupt
Mask Register 1F68
UA2CA UART 2 Compare
Address Register 1F69
UA2IF UART 2 Interrupt
Flag Register 1F6A
76543210
r Receive register
wTransmit register
xxxxxxxxRes
rRBUSY BRKD FRER OVRR PAER EMPTY FULL TBUSY
0xx0x100Res
wxxxxSTPBODDPARLEN
xxxx0000Res
wBit 7 to 0 of Baud Rate
00000000Res
wx x x Bit 12 to 8 of Baud Rate
- - -00000Res
wx x x x x ADR BRK RCVD
-----000Res
wBit 7 to 0 of address
00000000Res
rTest Test Test Test Test ADR BRK RCVD
-----x00Res
CDC16xxF-E ADVANCE INFORMATION
208 March 31, 2003; 6251-606-2AI Micronas
Table 27–30: Universal Port 1
Mnemonic Register Name Addr.
(hex) Register Configuration Section
U1D Universal Port 1 Data
Register 1F98 11.3.
U1SEG10 Universal Port 1 Seg-
ments of U1.0, U1.1 1F99
U1SEG32 Universal Port 1 Seg-
ments of U1.2, U1.3 1F9A
U1M30 Universal Port 1
Mode of U1.0 to U1.3 1F9B
U1SEG54 Universal Port 1 Seg-
ments of U1.4, U1.5 1F9C
U1M54 Universal Port 1
Mode of U1.4, U1.5 1F9D
U1SEG76 Universal Port 1 Seg-
ments of U1.6, U1.7 1F9E
U1M76 Universal Port 1
Mode of U1.6, U1.7 1F9F
76543210
r/w D7 D6 D5 D4 D3 D2 D1 D0
00000000Res
wxN/S1TRI1x xN/S0TRI0xPort
wLCDSLV LCD
00100010Res
wxN/S3TRI3x xN/S2TRI2xPort
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S5TRI5x xN/S4TRI4xPort
wSEG5.3 SEG5.2 SEG5.1 SEG5.0 SEG4.3 SEG4.2 SEG4.1 SEG4.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S7TRI7x xN/S6TRI6xPort
wSEG7.3 SEG7.2 SEG7.1 SEG7.0 SEG6.3 SEG6.2 SEG6.1 SEG6.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 209
Table 27–31: Universal Port 2
Mnemonic Register Name Addr.
(hex) Register Configuration Section
U2D Universal Port 2 Data
Register 1F30 11.3.
U2SEG10 Universal Port 2 Seg-
ments of U2.0, U2.1 1F32
U2M10 Universal Port 2
Mode of U2.0, U2.1 1F33
U2SEG32 Universal Port 2 Seg-
ments of U2.2, U2.3 1F34
U2M32 Universal Port 2
Mode of U2.2, U2.3 1F35
U2SEG54 Universal Port 2 Seg-
ments of U2.4, U2.5 1F36
U2M54 Universal Port 2
Mode of U2.4, U2.5 1F37
U2SEG76 Universal Port 2 Seg-
ments of U2.6, U2.7 1F38
U2M76 Universal Port 2
Mode of U2.6, U2.7 1F39
76543210
r/w D7 D6 D5 D4 D3 D2 D1 D0
00000000Res
wxN/S1TRI1x xN/S0TRI0xPort
wSEG1.3 SEG1.2 SEG1.1 SEG1.0 SEG0.3 SEG0.2 SEG0.1 SEG0.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S3TRI3x xN/S2TRI2xPort
wSEG3.3 SEG3.2 SEG3.1 SEG3.0 SEG2.3 SEG2.2 SEG2.1 SEG2.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S5TRI5x xN/S4TRI4xPort
wSEG5.3 SEG5.2 SEG5.1 SEG5.0 SEG4.3 SEG4.2 SEG4.1 SEG4.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S7TRI7x xN/S6TRI6xPort
wSEG7.3 SEG7.2 SEG7.1 SEG7.0 SEG6.3 SEG6.2 SEG6.1 SEG6.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
CDC16xxF-E ADVANCE INFORMATION
210 March 31, 2003; 6251-606-2AI Micronas
Table 27–32: Universal Port 3
Mnemonic Register Name Addr.
(hex) Register Configuration Section
U3D Universal Port 3 Data
Register 1FAC 11.3.
U3SEG10 Universal Port 3 Seg-
ments of U3.0, U3.1 1FAE
U3M10 Universal Port 3
Mode of U3.0, U3.1 1FAF
U3SEG32 Universal Port 3 Seg-
ments of U3.2, U3.3 1FB0
U3M32 Universal Port 3
Mode of U3.2, U3.3 1FB1
U3SEG54 Universal Port 3 Seg-
ments of U3.4, U3.5 1FB2
U3M54 Universal Port 3
Mode of U3.4, U3.5 1FB3
U3SEG76 Universal Port 3 Seg-
ments of U3.6, U3.7 1FB4
U3M76 Universal Port 3
Mode of U3.6, U3.7 1FB5
76543210
r/w D7 D6 D5 D4 D3 D2 D1 D0
00000000Res
wxN/S1TRI1x xN/S0TRI0xPort
wSEG1.3 SEG1.2 SEG1.1 SEG1.0 SEG0.3 SEG0.2 SEG0.1 SEG0.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wx N/S3 TRI3 x DPM2 N/S2 TRI2 x Port
wSEG3.3 SEG3.2 SEG3.1 SEG3.0 SEG2.3 SEG2.2 SEG2.1 SEG2.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S5TRI5x xN/S4TRI4xPort
wSEG5.3 SEG5.2 SEG5.1 SEG5.0 SEG4.3 SEG4.2 SEG4.1 SEG4.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S7TRI7x xN/S6TRI6xPort
wSEG7.3 SEG7.2 SEG7.1 SEG7.0 SEG6.3 SEG6.2 SEG6.1 SEG6.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 211
Table 27–33: Universal Port 4
Mnemonic Register Name Addr.
(hex) Register Configuration Section
U4D Universal Port 4 Data
Register 1FB8 11.3.
U4SEG10 Universal Port 4 Seg-
ments of U4.0, U4.1 1FBA
U4M10 Universal Port 4
Mode of U4.0, U4.1 1FBB
U4SEG32 Universal Port 4 Seg-
ments of U4.2, U4.3 1FBC
U4M32 Universal Port 4
Mode of U4.2, U4.3 1FBD
U4SEG54 Universal Port 4 Seg-
ments of U4.4, U4.5 1FBE
U4M54 Universal Port 4
Mode of U4.4, U4.5 1FBF
U4SEG76 Universal Port 4 Seg-
ments of U4.6, U4.7 1FC0
U4M76 Universal Port 4
Mode of U4.6, U4.7 1FC1
76543210
r/w D7 D6 D5 D4 D3 D2 D1 D0
00000000Res
wxN/S1TRI1x xN/S0TRI0xPort
wSEG1.3 SEG1.2 SEG1.1 SEG1.0 SEG0.3 SEG0.2 SEG0.1 SEG0.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S3TRI3x xN/S2TRI2xPort
wSEG3.3 SEG3.2 SEG3.1 SEG3.0 SEG2.3 SEG2.2 SEG2.1 SEG2.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S5TRI5x xN/S4TRI4xPort
wSEG5.3 SEG5.2 SEG5.1 SEG5.0 SEG4.3 SEG4.2 SEG4.1 SEG4.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S7TRI7x xN/S6TRI6xPort
wSEG7.3 SEG7.2 SEG7.1 SEG7.0 SEG6.3 SEG6.2 SEG6.1 SEG6.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
CDC16xxF-E ADVANCE INFORMATION
212 March 31, 2003; 6251-606-2AI Micronas
Table 27–34: Universal Port 5
Mnemonic Register Name Addr.
(hex) Register Configuration Section
U5D Universal Port 5 Data
Register 1FC4 11.3.
U5SEG10 Universal Port 5 Seg-
ments of U5.0, U5.1 1FC6
U5M10 Universal Port 5
Mode of U5.0, U5.1 1FC7
U5SEG32 Universal Port 5 Seg-
ments of U5.2, U5.3 1FC8
U5M32 Universal Port 5
Mode of U5.2, U5.3 1FC9
U5SEG54 Universal Port 5 Seg-
ments of U5.4, U5.5 1FCA
U5M54 Universal Port 5
Mode of U5.4, U5.5 1FCB
U5SEG76 Universal Port 5 Seg-
ments of U5.6, U5.7 1FCC
U5M76 Universal Port 5
Mode of U5.6, U5.7 1FCD
76543210
r/w D7 D6 D5 D4 D3 D2 D1 D0
00000000Res
wxN/S1TRI1x xN/S0TRI0xPort
wSEG1.3 SEG1.2 SEG1.1 SEG1.0 SEG0.3 SEG0.2 SEG0.1 SEG0.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S3TRI3x xN/S2TRI2xPort
wSEG3.3 SEG3.2 SEG3.1 SEG3.0 SEG2.3 SEG2.2 SEG2.1 SEG2.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S5TRI5x xN/S4TRI4xPort
wSEG5.3 SEG5.2 SEG5.1 SEG5.0 SEG4.3 SEG4.2 SEG4.1 SEG4.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S7TRI7x xN/S6TRI6xPort
wSEG7.3 SEG7.2 SEG7.1 SEG7.0 SEG6.3 SEG6.2 SEG6.1 SEG6.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 213
Table 27–35: Universal Port 6
Mnemonic Register Name Addr.
(hex) Register Configuration Section
U6D Universal Port 6 Data
Register 1FD0 11.3.
U6SEG10 Universal Port 6 Seg-
ments of U6.0, U6.1 1FD2
U6M10 Universal Port 6
Mode of U6.0, U6.1 1FD3
U6SEG32 Universal Port 6 Seg-
ments of U6.2, U6.3 1FD4
U6M32 Universal Port 6
Mode of U6.2, U6.3 1FD5
U6SEG54 Universal Port 6 Seg-
ments of U6.4, U6.5 1FD6
U6M54 Universal Port 6
Mode of U6.4, U6.5 1FD7
U6SEG76 Universal Port 6 Seg-
ments of U6.6, U6.7 1FD8
U6M76 Universal Port 6
Mode of U6.6, U6.7 1FD9
76543210
r/w D7 D6 D5 D4 D3 D2 D1 D0
00000000Res
wxN/S1TRI1x xN/S0TRI0xPort
wSEG1.3 SEG1.2 SEG1.1 SEG1.0 SEG0.3 SEG0.2 SEG0.1 SEG0.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S3TRI3x xN/S2TRI2xPort
wSEG3.3 SEG3.2 SEG3.1 SEG3.0 SEG2.3 SEG2.2 SEG2.1 SEG2.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S5TRI5x xN/S4TRI4xPort
wSEG5.3 SEG5.2 SEG5.1 SEG5.0 SEG4.3 SEG4.2 SEG4.1 SEG4.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S7TRI7x xN/S6TRI6xPort
wSEG7.3 SEG7.2 SEG7.1 SEG7.0 SEG6.3 SEG6.2 SEG6.1 SEG6.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
CDC16xxF-E ADVANCE INFORMATION
214 March 31, 2003; 6251-606-2AI Micronas
Table 27–36: Universal Port 7
Mnemonic Register Name Addr.
(hex) Register Configuration Section
U7D Universal Port 7 Data
Register 1FDC 11.3.
U7SEG10 Universal Port 7 Seg-
ments of U7.0, U7.1 1FDE
U7M10 Universal Port 7
Mode of U7.0, U7.1 1FDF
U7SEG32 Universal Port 7 Seg-
ments of U7.2, U7.3 1FE0
U7M32 Universal Port 7
Mode of U7.2, U7.3 1FE1
76543210
r/w D7 D6 D5 D4 D3 D2 D1 D0
00000000Res
wxN/S1TRI1x xN/S0TRI0xPort
wSEG1.3 SEG1.2 SEG1.1 SEG1.0 SEG0.3 SEG0.2 SEG0.1 SEG0.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
wxN/S3TRI3x xN/S2TRI2xPort
wSEG3.3 SEG3.2 SEG3.1 SEG3.0 SEG2.3 SEG2.2 SEG2.1 SEG2.0 LCD
00100010Res
wxxxxxxxPMODE
00000001Res
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 215
28. Differences
This chapter describes differences of this document to pre-
decessor document “CDC16xxF-E Automotive Controller
Family User Manual”, Feb. 17, 2003, 6251-606-1AI.
#Section Description
1 Features Table 1–1: "CDC16xxF Family Feature List" on page 5:
Name and features of “Example E-Family” changed into “CDC1631F-E”,
Multiplier, 8 by 8 bit added.
Fig. 1–1: “Block diagram of CDC1605F-E/CDC1607F-E” on page 10:
Multiplier, 8 by 8 bit added
2 External Components New section with Fig. 2–4: “Recommended external supply and quartz connection for low electromag-
netic interference (EMI)” out of section “Pin Function Description”. Value of C at RESETQ changed
from 47 µ to 47 n, value of C at VREF changed from 10 µ to 10 n and text added.
3 Pin Circuits New chapter
4 Electrical Data Characteristics:
mistake in writing corrected:
Outputs, Vol, Port Low Output Voltage, H-ports, ... “Io = 40 mA@TCASE = 40 °C” changed to “Io =
40 mA@TCASE = −40 °C”
5 CPU and Clock System Table 4–1: "Major Differences between Processors and Modes" on page 32:
Item “flags after reset” for 65C816 Emulation changed from “D not modified” to “D=0”.
6 Boot System Principle of operation:
Boot Loader check if started by wake-up from Power Saving Mode added.
The Boot Loader:
Fig. 5–7: “Boot Loader flow-chart” on page 46:
Boot Loader check if started by wake-up from Power Saving Mode (CSW2.WKID = ’1’) added.
Used RAM:
Page 2 demand added
7 Multiplier New Chapter
8 Core Logic Table 6–1: "Control byte source" on page 49:
Updated / minimized
Fig. 6–1: “UVDD Section” on page 51:
Updated
9 Power-Saving Module
(PSM) Timing:
Fig. 8–3: “Power-on Reset” on page 67: updated
10 Interrupt Controller (IR) Registers:
New: Interrupts Enable Register (IRE)
Interrupt Assignment:
INT-MUX 1 assignment changed
11 CAN Manual Updated from version LCAN0009 to version LCAN000F:
Mainly bit “Bus-Off Stop Select” in control register and
“Error Status Mask” register added
12 Hardware Options Listing of Dedicated Addresses and Corresponding Hardware Options:
INT-MUX 1 assignment changed
13 Register Cross Reference
Table V2.1 New registers added
ADVANCE INFORMATION CDC16xxF-E
Micronas March 31, 2003; 6251-606-2AI 217
All information and data contained in this data sheet are without any
commitment, are not to be considered as an offer for conclusion of a
contract, nor shall they be construed as to create any liability. Any new
issue of this data sheet invalidates previous issues. Product availability
and delivery are exclusively subject to our respective order confirmation
form; the same applies to orders based on development samples deliv-
ered. By this publication, Micronas GmbH does not assume responsibil-
ity for patent infringements or other rights of third parties which may
result from its use.
Further, Micronas GmbH reserves the right to revise this publication
and to make changes to its content, at any time, without obligation to
notify any person or entity of such revisions or changes.
No part of this publication may be reproduced, photocopied, stored on a
retrieval system, or transmitted without the express written consent of
Micronas GmbH.
CDC16xxF-E ADVANCE INFORMATION
218 March 31, 2003; 6251-606-2AI Micronas
Micronas GmbH
Hans-Bunte-Strasse 19
D-79108 Freiburg (Germany)
P.O. Box 840
D-79008 Freiburg (Germany)
Tel. +49-761-517-0
Fax +49-761-517-2174
E-mail: docservice@micronas.com
Internet: www.micronas.com
Printed in Germany
Order No. 6251-606-2AI
29. Data Sheet History
1. Advance Information: “CDC16xxF-E Automotive Control-
ler Family User Manual”, Feb. 17, 2003, 6251-606-1AI. First
release of the advance information. Originally created for the
HW version CDC16xxF-E1.
2. Advance Information: “CDC16xxF-E Automotive Control-
ler Family User Manual”, March 31, 2003, 6251-606-2AI.
Second release of the advance information. Originally cre-
ated for the HW version CDC16xxF-E2.
