Browse Prior Art Database

Method to Determine First VRM Failure in a Multiple VRM System

IP.com Disclosure Number: IPCOM000126476D
Original Publication Date: 2005-Jul-20
Included in the Prior Art Database: 2005-Jul-20
Document File: 2 page(s) / 29K

Publishing Venue

IBM

Abstract

This method detects the first VRM (in multiple VRM systems) failure. The continuing trend of increasing system power, where circuit and system densities go up, has the effect that many VRM's (Voltage Regulation Module) must be used instead of a single power supply.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Method to Determine First VRM Failure in a Multiple VRM System

     This method is used to detect a marginally bad VRM in a multi-VRM system or subsystem. If one VRM fault occurs (glitch or total failure), all the VRMs may be affected and the system goes down. It is then difficult to determine which VRM caused the fault for replacement. The trend to use more VRMs continues thus requiring this power supply technique to be more pervasive in the future. To meet Printed Circuit Board (PCB) safety requirements, it is necessary to limit the amount of current that traverses a plane more so than the voltage. Presently, power supplies do not output low voltages because current requirements exceed safety limits.The continuing trend of increasing system power, where circuit and system densities go up, has the effect that many VRM's (Voltage Regulation Module) must be used instead of a single power supply.

     In the lab, a common method is to check which VRM does not power-up correctly. A typical and tedious method to determine VRM fault is to look at error capture registers and determine which card failed first to help narrow the search. However, each card has several VRMs. Many IBM server diagnostics programs are able to determine which subassembly failed, but not a particular VRM. The system will not power up, so probes must be used to determine which VRM is getting the enable signal but isn't providing power good before it can be replaced. Drawbacks for this method are special test equipment, high degree of system understanding, difficult access to system test points, waiting hours or days for a failure and partial system disassembly. Field service is more difficult and costly because more time and replacement parts must be used.

     System failures due to VRM glitches cause system faults but not VRM fault indicatio...