6
dc2313af
DEMO MANUAL DC2313A
The LTC2937 monitors each channel for individually-
programmed overvoltage and undervoltage thresholds.
The voltage monitoring is active while the supplies are
up or sequencing-up.
The LTC2937 is configured to re-try after the fault; it will
attempt to sequence-up again. There is a rich suite of fault
response capabilities in the part, including turning off and
staying off, turning-off then re-sequencing, continuing
operation without turning off, or entering a debug mode.
More details are available later in this manual, and in the
LTC2937 data sheet.
The FAULT indicators are self-clearing upon a re-sequence
initiation. The only fault indications after a successful
re-sequence-up are the asserted ALERTB pin, and the
EEPROM record of the first fault condition after the most
recent power-up that produced the fault. The ALERTB
pin will remain asserted low until the ALERT condition is
cleared with an I2C bus operation.
Sequence Fault
A SEQUENCE fault is caused by supplies failing to meet
programmed voltage thresholds within programmed time
allowances during sequencing (for example, not ramping
fast enough).
Create this type of fault on the DC2313A board by pulling
and holding down one of the ENn turrets to GND while
sequencing-up the supplies. This will hold down the as-
sociated LDO and create a permanent UV condition. The
LTC2937 will recognize the unresponsive LDO and signal
a fault.
1) Start with the system sequenced-down. The LDOs are
off.
2) Short turret EN4 to GND and hold it there (press and
hold the FAULT pushbutton).
3) Press the “SEQUENCE UP/DOWN” button to initiate a
sequence-up operation.
EN1 - EN3 go high in sequence
CH1 - CH3 start in response to EN1-EN3
CH4 fails to start. LEDs EN4-6, CH4-6 remain off.
4) Observetherepeatedfault retryresponse(seethe table)
Quick start proceDure (without software)
Pins EN1 - EN3 pull low immediately. All illuminated
ENn LEDs turn off.
All of the LDO regulated supplies turn off. Illuminated
CHn LEDs turn off.
The LTC2937 is configured to automatically re-try
after a delay, so it will attempt to sequence-up the
supplies after detecting the fault. Since the LDO is
shorted, the fault persists, and re-sequencing will
repeatedly fail. The cycle will repeat until the fault is
removed (by releasing the FAULT pushbutton).
Pin FAULTB will assert low for the retry interval, until
the re-sequence begins. LED FAULT will illuminate
during the interval.
The FAULTB pin clears as soon as a re-sequence
begins. The FAULT LED illuminates briefly, then goes
off.
Pin RSTB will remain low. The RST LED will remain
illuminated because not all of the supplies ever come
up.
Pin ALERTB will assert low. The ALERT LED will
illuminate. The alert state will persist until an alert
response or a read from the CLEAR_ALERTB (0x28)
comes from the I2C bus. Only a bus operation can
release the ALERTB pin.
5) Remove the EN4 fault by releasing the FAULT pushbut-
ton.
6) Observe that the part completes the re-sequence au-
tonomously,and the fault responseclearsautomatically.
The RST LED goes off after all supplies are up.
Each channel has an independently-configurabletime limit
for starting-up and rising above its UV threshold voltage.
Each channel also has independently-programmable
sequence-down time parameters.
The LTC2937 is configured by default in the FAULT_RE-
SPONSE (0x23) register to automatically re-sequence-up
after all of the supplies are disabled and the voltages fall
below their discharge thresholds. While the EN4 pin is held
low, the fault persists, and the LTC2937 is programmed
to automatically try to re-sequence the supplies forever,
so the behavior will repeat indefinitely.
