LTC4227
13
422712fa
For more information www.linear.com/LTC4227
applicaTions inForMaTion
Active Current Loop Stability
The active current loop on the HGATE pin is compensated
by the parasitic gate capacitance of the external N-channel
MOSFET. No further compensation components are nor-
mally required. In the case when a MOSFET with CISS ≤
2nF is chosen, an RHG and CHG compensation network
connected at the HGATE pin may be required. The value
of CHG is selected based on the inrush current allowed for
the output load capacitance. The resistor, RHG, connected
in series with CHG accelerates the MOSFET gate recovery
for active current limiting after a fast gate pull-down due
to an output short. The value of CHG should be ≤100nF
and RHG should be between 10Ω and 100Ω for optimum
performance.
TMR Pin Functions
An external capacitor, CT
, connected from the TMR pin
to GND serves as fault filtering when the supply output is
in active current limit. When the voltage across the sense
resistor exceeds the circuit breaker trip threshold (50mV),
TMR pulls up with 100µA. Otherwise, it pulls down with 2µA.
The fault filter times out when the 1.235V TMR threshold
is exceeded, causing the corresponding FAULT pin to pull
low. The fault filter delay or circuit breaker time delay is:
tCB = CT • 12[ms/µF].
After the circuit breaker timeout, the TMR pin capacitor
pulls down with 2µA from the 1.235V TMR threshold
until it reaches 0.2V. Then, it completes 14 cooling cycles
consisting of the TMR pin capacitor charging to 1.235V
with a 100µA current and discharging to 0.2V with a 2µA
current. At that point, the HGATE pin voltage is allowed to
start up if the fault has been cleared as described in the
Resetting Faults section. When the latched fault is cleared
during the cool-off period, the corresponding FAULT pin
pulls high. The total cool-off time for the MOSFET after
an overcurrent fault is:
tCOOL = CT • 11[s/µF]
If the latched fault is not cleared after the cool-off period,
the cooling cycles continue until the fault is cleared.
After the cool-off period, the HGATE pin is only allowed
to pull up if the fault has been cleared for the latch-off
part. For the auto-retry part, the latched fault is cleared
automatically following the cool-off period and the HGATE
pin voltage is allowed to restart.
Resetting Faults (LTC4227-1/LTC4227-3)
For the latch-off part, an overcurrent fault is latched after
tripping the circuit breaker, and the FAULT pin is asserted
low. Only the Hot Swap MOSFET is turned off and the ideal
diode MOSFETs are not affected.
To reset a latched fault and restart the output, pull the ON
pin below 0.6V for more than 100µs and then high above
1.235V. The fault latches reset and the FAULT pin deas-
serts on the falling edge of the ON pin. When ON goes
high again, a debounce timing cycle is initiated before
the HGATE pin voltage restarts. Toggling the EN pin high
and then low again also resets a fault, but the FAULT pin
pulls high at the end of the debounce timing cycle before
the HGATE pin voltage starts up. Bringing all the supplies
below the INTVCC undervoltage lockout threshold (2.2V)
shuts off all the MOSFETs and resets all the fault latches.
A debounce timing cycle is initiated before a normal start-
up when any of the supplies is restored above the INTVCC
UVLO threshold.
Auto-Retry After a Fault (LTC4227-2/LTC4227-4)
For the auto-retry part, the latched fault is reset automati-
cally after a cool-off timing cycle as described in the TMR
Pin Functions section. At the end of the cool-off period,
the fault latch is cleared and FAULT pulls high. The HGATE
pin voltage is allowed to start up and turn on the Hot Swap
MOSFET. If the output short persists, the supply powers
up into a short with active current limiting until the cir-
cuit breaker times out and FAULT again pulls low. A new
cool-off cycle begins with TMR ramping down with a 2µA
current. The whole process repeats itself until the output
short is removed. Since tCB and tCOOL are a function of
TMR capacitance, CT,the auto-retry duty cycle is equal to
0.1%, irrespective of CT.
Figure 6 shows an auto-retry sequence after an overcur-
rent fault.