PCI units missing IRQ jumper

World         Australia | Europe | North America | Singapore
Europe        Austria | Birmingham | Budapest | France | Italy | London | TeamOS/2 (Germany)
North America Berkeley | Chattanooga | Minnesota | Nova Scotia | Ontario | Pennsylvania | TeamOS/2
Last update: 7th February, 1996

Symptoms

Installation of Warp hung.

First, because the hard disk was not visible.

Secondly, with a message about being unable to read file.
This occurred even after the first six disks' worth of system files had been loaded,
while trying to read Disk1.

The problems occurred regardless of whether installation took place from diskettes or from CD.

Under DOS, the same machine ran all right. (As well as you'd expect... :-} )

Hardware

Pentium 100 PCI motherboard with 256K cache (Mematex 586 SLB-P102 from Chaintech, Hamburg)
Award BIOS 4.50G, 01/12/95
16MB RAM (8+8)
Quantum 730S Lightning SCSI hard disk
Mitsumi FX400 4x IDE CD-player
Soundblaster 16 VE
  Cirrus Logic PCI GP-5434 video accelerator with 2MB
(Chaintech)
NCR SP-810 PCI SCSI controller (Chaintech)

Problem

The cause turned out to be two jumpers missing or set at incorrect positions.

The SCSI controller can be set for a variety of IRQ values, any of which would probably work. In fact it was set to none. The jumper was missing. This upsets Warp, which seems to query the equipment for such settings and, finding no acceptable reply, simply stops.
It won't take 'I don't know' for an answer. Therefore, the hard disk was not accessible.

The second error was the video adapter being set for "IRQ9 disabled"
by a jumper in the wrong position (shorting pins 2&3 instead of 1&2).
During the initial stages of installation this didn't matter,
but once OS/2 tried to stand on its own feet it stopped.

Procedure

Check all jumpers and/or switches on adapter boards.
Move or replace them to set correct values.

Notes

This may be a trivial problem, but it kept me from using my new PC for two months!

Of course, had I paid extra for a brand name PC from a more knowledgeable,
OS/2 consciuos dealer there might never have been a problem at all. <:-(
Though I hasten to add that they did try to help.

It is interesting that DOS stumbled along without even noticing anything amiss.
Well, DOS can always crash later anyway, can't it, so why bother to check anything now? ;-)


Ronald Nillson
[ The Warp Pharmacy | News | Hardware | Software | Symptoms | Procedures | References | Commerce | Search | More Help!!! ]