Hi Shogun, it really sounds like the EIII doesn't like that particular drive for some reason. The more I fix up/upgrade Emax's, EIII's, EII's and now an EIIIXS, the more I see really weird things come out of nowhere. Especially SCSI stuff.
I just got into upgrading an EIIIXS for the first serious time, and what a nightmare so far. I have never been successful in getting one of them in a stable working state.
To give you an idea of what its done to me so far (SCSI wise), and what I have tried, read on...
Unit started off with a failing PSU, dead HD and bad display.
Installed OLED display. Issue fixed.
Replaced PSU. Issue fixed.
Installed Acard SCSI/IDE bridge with CF adapter. Machine ID 5, SCSI ID 4, CDROM drive ID 3. It booted the system and worked fine for an entire afternoon of saving/recalling banks. I initially used a Kingston 1GB card as we do with the original EIII mod. Interestingly, this card was the only one out of the 4 or 5 I tested on the setup that showed up on the machines's drive menu as 'Cf Card 1GB'. Any others, regardless of size or brand showed up simply as 'D4' (SCSI ID is 4 on the Acard). Initially, all tested cards were formatted with the default interleave, which for the EIIIX is 3 if I recall.
Problems started showing up when running the backup utility in Disk Management/Backup. I had an Apple CD300 CDROM drive hooked up, set the backup to ALL/AUTOMATIC/EMPTY BANKS and let it go to it. So I was wanting to copy the entire contents of an EIIIX CDROM library to the internal CF drive.
I watched it read and write until it got to around the tenth or so bank transfer, and then it displayed a 'SCSI Hardware Error'. I restarted and ran it again. This time it failed with the same error after around 6 banks transferred.
At this point, I changed CF cards. I tried a different model of 1GB Kingston, the one with the lizard or whatever on the front. Formatted it, saved blank bank, installed SW etc etc, and ran the same backup again. It failed with the same error again but at a different point, bank 14 or something.
So after seeing that, I figured I had a hardware issue. I have spares lying around, so I changed the SCSI bridge first and ran the backup again with similar failure results. Next I changed the card reader. Same thing again.
Then I changed the internal SCSI ribbon cable. Same results. At this point, I tried a different CDROM library disk just incase and that made no difference. Taking it a step up, I tried a different CDROM drive and SCSI cable with no effect on the resulting hardware error. Getting really pissed off, I changed the SCSI controller IC. No change.
Just FYI, the CDROM drive supplied its own TERM power, so I tried putting a terminator block on it. No change again.
Every now and again, I got a 'SCSI MEDIA ERROR' instead, just to keep me coming back for more I reckon.
The Acard bridge was set to be terminated and supply its own TERM power. I tried turning off TERM power on the Acard and enabled internal TERM power on the EIIIX while keeping the Acard terminated, just not supplying its own TERM power. NO CHANGE.
Now I tried a whole bunch of format interleave settings on various cards and kept running the backups. Same string of random failures.
The thing that really gets me on this whole thing is that there is no commonality to when the backup process quits. If it quit at the same point every time, it wouldn't be as frustrating.
Is it entirely possible that the CF solution is just not 100% compatible with the 3X machines for some reason? I would say so. I don't think it's any kind of SCSI termination or ID conflict, and all of that different hardware can't all be defective and produce the same failures.
Sorry if this was off on too much of a tangent, but it might give you a few things to try or keep in mind.
Jay
Jan 10 update:
A generic 256mb card resolved all of the errors. Anything above would not work for this machine.
Author