AD8565/AD8566/AD8567
–8– REV. C
APPLICATIONS
Theory of Operation
The AD856x family is designed to drive large capacitive loads in
LCD applications. It has high output current drive, rail-to-rail
input/output operation, and is powered from a single 16 V supply.
It is also intended for other applications where low distortion and
high output current drive are needed.
Figure 1 illustrates a simplified equivalent circuit for the AD856x.
The rail-to-rail bipolar input stage is composed of two PNP
differential pairs, Q4 to Q5 and Q10 to Q11, operating in series
with diode protection networks, D1 to D2. Diode network
D1 to D2 serves as protection against large transients for
Q4 to Q5 to accommodate rail-to-rail input swing. D5 to D6
protect Q10 to Q11 against Zenering. In normal operation,
Q10 to Q11 are off and their input stage is buffered from the
operational amplifier inputs by Q6 to D3 and Q8 to D4. Opera-
tion of the input stage is best understood as a function of applied
common-mode voltage: when the inputs of the AD856x are
biased midway between the supplies, the differential signal
path gain is controlled by resistive loads (via R9, R10) Q4 to Q5.
As the input common-mode level is reduced toward the negative
supply (V
NEG
or GND), the input transistor current sources, I1
and I2, are forced into saturation, thereby forcing the Q6 to D3
and Q8 to D4 networks into cutoff. However, Q4 to Q5 remain
active, providing input stage gain. Inversely, when common-mode
input voltage is increased toward the positive supply, Q4 to Q5
are driven into cutoff, Q3 is driven into saturation, and Q4
becomes active, providing bias to the Q10 to Q11 differential
pair. The point at which Q10 to Q11 differential pair becomes
active is approximately equal to (V
POS
– 1 V).
R1
R3 R4
D1 D2
Q4
Q3 BIAS LINE
V–
D3 D4
Q5
Q4
R5 R6
Q10 Q11
C1
C2
D5
D6
Q8
Q6
R9 R10
FOLDED
CASCADE
V+
I1 I2
VNEG
VPOS
Figure 1. AD856x Equivalent Input Circuit
The benefit of this type of input stage is low bias current. The
input bias current is the sum of base currents of Q4 to Q5 and
Q6 to Q8 over the range from (V
NEG
+ 1 V) to (V
POS
– 1 V). Out-
side of this range, input bias current is dominated by the sum of
base currents of Q10 to Q11 for input signals close to V
NEG
and of
Q6 to Q8 (Q10 to Q11) for signals close to V
POS
. From this type
of design, the input bias current of AD856x not only exhibits
different amplitude but also exhibits different polarities. Figure 2
provides the characteristics of the input bias current versus the
common-mode voltage. It is important to keep in mind that the
source impedances driving the AD856x inputs are balanced for
optimum dc and ac performance.
INPUT COMMON-MODE VOLTAGE (V)
1,000
–1,000 0162
INPUT BIAS CURRENT (nA)
468101214
800
200
–200
–600
–800
600
400
0
–400
VS = 16V
TA = 25C
Figure 2. AD856x Input Bias Current vs. Common-Mode
Voltage
In order to achieve rail-to-rail output performance, the AD856x
design uses a complementary common-source (or gmRL) output.
This configuration allows output voltages to approach the power
supply rails, particularly if the output transistors are allowed to
enter the triode region on extremes of signal swing, which are
limited by V
GS
, the transistor sizes, and output load current.
Also, this type of output stage exhibits voltage gain in an open-loop
gain configuration. The amount of gain depends on the total
load resistance at the output of the AD856x.
Input Overvoltage Protection
As with any semiconductor device, whenever the input exceeds
either supply voltages, attention needs to be paid to the input
overvoltage characteristics. As an overvoltage occurs, the amplifier
could be damaged, depending on the voltage level and the magnitude
of the fault current. When the input voltage exceeds either supply
by more than 0.6 V, internal pn junctions allow current to flow
from the input to the supplies.
This input current is not inherently damaging to the device as
long as it is limited to 5 mA or less. If a condition exists using
the AD856x where the input exceeds the supply more than 0.6 V,
an external series resistor should be added. The size of the resis-
tor can be calculated by using the maximum overvoltage divided
by 5 mA. This resistance should be placed in series with either
input exposed to an overvoltage.