12
dc2518af
DEMO MANUAL DC2518A
Fault Sharing Setup in the GUI
Fault sharing provides a means of propagating a fault
detected by a power manager to other power managers
via FAULT pins. Use the Fault Sharing Setup Tool to con-
figure the fault sharing in the GUI. Select the LTC2975 in
the system tree. Go to Utilities
→
Fault Sharing Diagram.
For more details on this topic, please refer to the “Fault
Management” section in the data sheet.
The fault sharing dialog will appear as shown in Figure 9.
All Respond and all Propagate switches are closed by
default. In this configuration, a fault on a channel will shut
down not only the faulted channel but all other channels
since the Propagate switches are closed.
ADVANCED DEMO BOARD OPERATIONS
There are two types of actions to fault conditions: How a
channel responds to another channel’s fault and whether
a particular channel propagates its fault to other channels.
FAULT pins are bidirectional, meaning the device may drive
its fault pin low (output) or may respond to the fault pin
when another device drives it low (input). By default, the
LTC2975 is configured to shut down all channels if the
FAULT pin is low and to propagate its own fault to the
other channels by driving both FAULT pins low. On the
DC2518A, the two FAULT pins are tied together, which
effectively creates one fault line. You can think of the
“Respond” switches as “shut this channel down” when
another channel faults, and the “Propagate” switches as
“drive the fault pin” to broadcast to other channels that
this channel faulted.
Figure 9. Fault Sharing Utility in LTpowerPlay GUI
Fault Configuration Example
Suppose we do not want channel CH0 (+12V rail) to
propagate its fault to the other channels when it faults.
And suppose we do not want channel CH1 (–12V rail) to
shut down in response to another channel’s fault. We can
configure the switches as shown in Figure 10. Simply click
the switches to open/close. Click OK to close the dialog
box. Click the “PC
→
RAM” icon to write the changes to
the LTC2975.
You can now create a fault on CH0 by pressing the FAULT
pushbutton. This action does not directly short the +12V
supply to GND but instead pulls the V
SENSE
pin to GND. You
will notice that the channel shuts off but the other channels
remain powered up because its fault is not propagated to
the other channels. After the retry period, channel CH0
will power back up. You can now observe the effect of
changing the response setting on CH1. If you short the
test point VSP3 to ground on CH3 (–48V rail), notice that
all rails shut down except CH1. This is an example of a
keep-alive channel that remains powered up independent
of faults to other channels.
NOTE: It is not recommended to short the +48V or –48V
power supplies directly to GND on this demo board as
you may damage an output. Damage might occur if you
change the fault response to “ignore”.
Figure 10. Updated Fault Sharing Configuration
