Page 2 of 2
Re: CSPPC SCSI problems
Posted: Sat Feb 04, 2012 9:36 am
by Slayer
Can you post these two text files from your Sys:Kickstart directory please
diskboot.config
Kicklayout
Re: CSPPC SCSI problems
Posted: Sat Feb 04, 2012 4:33 pm
by Calgor
drbrain wrote:Terminator - CSPPC - UWSCSI HDD (with termpwr on) - Terminator
I've tried three different cables and four different terminators (all working fine with my main CSPPC) but none will work with OS4.1 on this card...
I've also tried four different firmwares on the CSPPC too... v44.57, v44.67, v44.69 and v44.71 (all the same, OS3.x works but not OS4.x)
That sounds perfect and simple. But is it an active terminator or LVD/SE terminator on each end? If so it should be fine. They need to at least operate in Single-Ended(SE) mode for the CSPPC chain and most but not all terminators are backward compatible to SE mode. But apparently it is even better to have matched terminators (don't know if this last bit is true).
Given you tried 4 terminators I doubt it is the problem, but can't hurt to check.
That's me out of suggestions as you have tried all conservative csppc boot menu memory and scsi settings, except to try as DarrenHD suggested a different hard disk (and also a completely new install on that).
Re: CSPPC SCSI problems
Posted: Mon Feb 13, 2012 5:18 pm
by drbrain
I've officially given up now...it just won't work with OS4.x..
Oh well... luckily my main card works fine
Re: CSPPC SCSI problems
Posted: Tue Feb 14, 2012 3:30 am
by Slayer
Don't make me come down there!
I asked for two lousy print outs
no worries if you've given up
Re: CSPPC SCSI problems
Posted: Tue Feb 14, 2012 11:30 pm
by tonyw
Why do you have a terminator at the CPU end? I seem to recall that the Cyberstorm SCSI cards are internally terminated. If that is true, then you have a doubly-terminated bus at that end, which would reduce signal levels.
Re: CSPPC SCSI problems
Posted: Wed Feb 15, 2012 3:21 am
by DarrenHD
tonyw wrote:Why do you have a terminator at the CPU end? I seem to recall that the Cyberstorm SCSI cards are internally terminated. If that is true, then you have a doubly-terminated bus at that end, which would reduce signal levels.
There is no terminator on the CS PPC. The easiest cable setup (with one drive) is active terminator on each end of the cable with a non-terminated drive somewhere in the middle. If you add a SCSI CDROM then you need to terminate the lower 8-bits on the drive - with most drives this can be accomplished by enabling the terminator jumer on the CDROM drive, as far as I can tell - at least this setup works for me.
In my experience problems can exist with the old SCSI cables themselves. People probably had these cables for a long time and over time they can become unreliable even on a 100% correctly cabled system.
Re: CSPPC SCSI problems
Posted: Sun May 13, 2012 8:58 am
by danbeaver
It seems like, for unknown reasons, that the kernel is not loading (or utilizing) the Cyberstorm SCSI device kmod. In which case it would boot from another HD device (IDE or sii3112ide), but this makes no sense as the other CSPPC SCSI is recognized; you can make sure the device is not commented out or reinstall a fresh OS 4.1; I've done it a dozen times or so trying to to work out issues. Another option is to trade or sell it to someone who is a non-OS 4.1 believer.
Slayer's printouts WOULD help out in this situation.
Dan
Re: CSPPC SCSI problems
Posted: Tue Jul 10, 2012 11:21 am
by danbeaver
Thinking it over, the 4.1 kernel relies heavily on the CSPPS's cybppc.device (to the extent that it is the ONLY DMA allowed in 4.1) that the NCR 53C770 scsi controller chip that a small variance in the chip may cause issues; it, I believe is a SMD but easily replaced by someone with right tools. The chip shows up Amibay and eBay every so often that replacing it might fix the issue.
Re: CSPPC SCSI problems
Posted: Mon Jul 23, 2012 11:02 pm
by drbrain
The non-working (in OS4.1) card was sold a couple of months ago... no need for a spare CSPPC
So no more help needed regarding this topic from m side...