Datasheet

UCD9081
www.ti.com SLVS813B –JUNE 2008–REVISED DECEMBER 2010
RAIL SHUTDOWN
Rail (or GPO) shutdown is the act of setting the ENx (or GPOx) pin to a state which disables the associated
power supply output. A rail can shutdown for one of the following reasons:
• A fault condition on the rail itself
• A fault condition on a parent rail resulting in a shutdown
• An I
2
C shutdown command
Each rail and GPO can be independently configured to shutdown according to a user-specified time delay
between 0 - 4095 ms. This is referred to as the system shutdown configuration.
ALARM PROCESSING
Each rail can be independently configured to respond to an alarm or fault in a variety of ways. A fault can be an
UV condition, OV condition, or a rail that did not start (MTFR exceeded before UV threshold achieved). The
options for alarm processing are as follows:
• Ignore
• Log only
• Retry n times (n = 0,1,2,3,4)
• Retry continuously
• Sequence (immediately)
• Sequence After Shutdown
In addition to these options, a rail can be independently configured to log errors to FLASH to aid in failure
analysis. For more details, see the section on Error Logging .
IGNORE
The UCD9081 can be configured to ignore all alarms on the rail. This is the recommended option for all unused
power supply rails on the UCD9081.
LOG ONLY
The UCD9081 can be configured to log a fault and take no additional action. For more information, see the
section on Error Logging .
RETRY n TIMES
The UCD9081 can be configured to attempt to restart a rail up to n times (n = 0,1,2,3, or 4) in response to a
sustained fault condition. With this option, the user can also specify which rails and GPOs are dependent upon
the configured rail. When a sustained fault is detected, the faulty rail will be disabled and re-enabled the desired
number of times. The rail remains enabled for the specified MTFR before attempting another retry. If the rail does
not achieve regulation after the desired number of retries, all user-specified dependent rails and GPOs will be
shutdown according to the times specified in the system shutdown configuration. Note that if any of the
dependent rails have other rails or GPOs marked as dependents, those dependent rails or GPOs will also be
forced to shutdown regardless of their alarm processing configurations.
RETRY CONTINUOUSLY
The UCD9081 can be configured to continuously attempt to restart a faulty rail. When the UCD9081 detects a
sustained fault condition on the configured rail, the rail is disabled and then re-enabled. The rail remains enabled
for the specified MTFR. The retry process repeats for this rail until it properly achieves regulation.
SEQUENCE
The UCD9081 can be configured to sequence the entire system in response to a sustained fault condition. When
the UCD9081 detects a fault on the configured rail, all rails and GPOs are shutdown immediately and UCD9081
RESET occurs (see section on Device Reset ). Note that for this configuration, a shutdown according to the
delay times specified by the system shutdown configuration does not occur prior to UCD9081 RESET.
Copyright © 2008–2010, Texas Instruments Incorporated Submit Documentation Feedback 11
Product Folder Link(s) :UCD9081