[Geoman]

OK so now I'm just curious. What was the symptom that convinced you you had to actually replace your SCSI controllers? I mean that's a big leap isn't it?

It's awesome that you're able to solve it but I like to learn from people's projects and I'm having a hard time understanding what the red flag was that told you to change out your SCSI controllers versus say cabling or a bus/backplane termination (bad PCB termination component) issue or connector corrosion or the like in the entire system?

Was there a unintended, unsupported, hot-swap event on the system that you then suspected later damaged the controllers because of these bus errors?

Thank you!

Yes, the gold-standard would have been to check the IO3B-board with an MSO for possible failures. But I only have a DSO at hand, and to 'logic-analyze' the board I'm lacking the proper training.

So there were 3 things that led me to believe the SCSI-controller-chip(s) were damaged:

1. CB_HK had problems with one of his SCSI channels some time ago:  WTB: Spare Crimson IO3B (FOUND)
2. During boot Crimson gave me persistent SCSI Controller Reset errors (wd93-errors, see 1st post)[url=https://techpubs.jurassic.nl/manuals/0630/admin/IA_ConfigOps/sgi_html/apb.html][/url]
3. The Amiga 3000 uses a AM33C93A-16PC chip as well; in the Amiga-hobbyist-scene this is known to fail very often

---

Is there a IO3-board-test suite or another possibility to test the VME bus on the Crimson?

That would the standalone diagnostics, ide.IP15. It should be in your /stand directory and you boot it like you would boot fx.IP15

Of course, if the system doesn't boot from disk in the first place than you may need to net/tape/cdrom boot it.

An original IRIX 5.3 CD (812-0119-006, 11/94) gives me

Code:

Can't load dksc(1,6,7)/stand/ide.IP17

as well as IRIX 6.2 (812-0469-001 03/96) [what!?: "Crimson (except GTX)" so this could never work in the first place, anyway I need another 6.2 CD later on...]

The SGI Diagnostic 4.0.5J gave me at least:

Code:

653472

and stops right there. So at the moment I cannot get IDE to run.

Aside from that Crimson is still happy without wd93-errors. But I don't dare to insert that Vigra board nonetheless, until I know what's going on with the VME bus.