Tech Not on ASR5500: Bad Voltage Alarm, Low Voltage Error

Introduction

This document describes how to resolve the issue of bad voltage alarm and low voltage error.

Problem: Bad Voltage and Low Voltage Error

Syslog showing zero voltage

 [drvctrl 39008 error] [6/0/13445 <hwctrl:0> hwctrl_msgs.c:501] [software internal system syslog] [DRIVER CONTROLLER] Low voltage error at card 5 with voltage (CPU0 VTT) current 0.000000 V low 0.839000 V high 1.277000 V.

Show card information reports "Bad ***ALARM***" in voltage section 

[local]# show card info 5
Card 5:
Slot Type               : MMIO
Card Type               : Management & 20x10Gb I/O Card
Daughter Cards          : DC1 DC2
Operational State       : Standby
Redundant With          : 6
Framing Mode            : Unspecified
Last State Change       : Wednesday April 13 02:03:56 PDT 2016
Administrative State    : Enabled
Card Lock               : Locked
Reboot Pending          : No
Upgrade In Progress     : No
Card Usable             : Yes
Single Point of Failure : n/a
Temperature             : Normal
Voltages                : Bad  ***ALARM*** 
Card LEDs               : Run/Fail: Green | Active: Off   | Redundant: Green
Card LEDs               : Master:   Off
CPU 0                   : Diags/Kernel Running, Tasks Running
Error ID Log          : None
Boot Progress Log     : None

Please verify the show card diag output it shows voltage looks good. 

 [local]# show card diag 5 
 
 Card 5:
   Counters:
     Successful Warm Boots : 4
       (last at Wednesday April 13 02:03:51 PDT 2016)
     Successful Cold Boots : 22
       (last at Wednesday March 16 02:23:29 PDT 2016)
     Total Boot Attempts   : 38
     In Service Date       : Sat Feb 09 19:53:01 2013 (Estimated)
   Status:
     IDEEPROM Magic Number : Good
     Boot Mode             : Normal
     Card Diagnostics      : Pass
     Current Failure       : None
     Last Failure          : None
     Card Usable           : Yes
     Last Reset Cause CPU 0: Hard Reset, ICH PLT Reset, ITP Reset
   Current Environment:
     Temp: DDR-C0D0        : 31.00 C (limit 95.00 C)
     Voltage: 12V-A        : 12.035 V
     Voltage: CPU0 VTT     : 1.106 V (min 0.839 V, max 1.277 V)
     Voltage: 12V-B        : 11.973 V
     Voltage: 12V-C        : 11.812 V
     Voltage: 1.8V         : 1.772 V (min 1.700 V, max 1.900 V)
     Voltage: 1.5V         : 1.483 V (min 1.430 V, max 1.580 V)
     Voltage: CPU0 VCC     : 0.918 V (min 0.712 V, max 1.417 V)
     Voltage: 1.2V         : 1.187 V (min 1.140 V, max 1.260 V)
     Voltage: 3.3V         : 3.283 V (min 3.140 V, max 3.470 V)
     Voltage: 5V           : 4.947 V (min 4.750 V, max 5.250 V)
     Voltage: 2.5V         : 2.497 V (min 2.380 V, max 2.630 V)
     Voltage: 1.53V DDR0   : 1.520 V (min 1.282 V, max 1.606 V)
     Voltage: 7.5V         : 7.460 V (min 7.130 V, max 7.880 V)
     Voltage: 1.1V IOH     : 1.096 V (min 1.045 V, max 1.155 V)
     Voltage: 3.3V Stdby   : 3.302 V (min 2.970 V, max 3.630 V)
     Voltage: 48V-A        : 51.925 V
     Voltage: 48V-B        : 52.734 V
     Current: 48V-A        : 5.44 A
     Current: 48V-B        : 3.56 A
     Airflow: Lower Left   : 361 FPM
     Airflow: Lower Middle : 369 FPM

Solution

This is NOT a hardware issue, please do NOT RMA the card - the alarm and syslog errors are misleading.  Please refer to bug CSCuz23239. The error/alarm showed due to voltage error has not been cleared correctly.  We see the issue when the card boots successfully after a previous card reset or shutdown resulted in a transient or race condition voltage reading.

Workaround: Reseating the card (put the card into standby before reseating as to avoid traffic impact)

Fix CSCuz23239 is included into 21.0 and later SW.