On 8/29/05, Wolfe, Julian <ISC277 at clcillinois.edu> wrote: There are a couple issues, the first being interface, the second being BIOS, and the third being DOS. 1. I'm not familiar with the IDE interface you mentioned. If it's an 8-bit IDE interface you're likely to have problems with 16 bit drives. The drive geometry information will be garbled at best... SCSI should work OK provided the drives support whatever the controller provides (which I would assume would be SCSI-1 (single ended, 5 MHz)) You should check whether TRM_PWR is supplied by your controller and try an active terminator, rather than any autotermination. I've found that some drives don't autoterminate SCSI-1 properly. 2. Your XT IDE controller is likely to be limited to 504MiB drives without a BIOS upgrade. Your XT IDE controller is unlikely to be doing any translation of block addresses required for LBA. It's probably doing cylinder-head-sector addressing. If the BIOS was written for DOS 4.X or earlier it's probably calculating size as follows.... cylinders=true_cylinders mod 1024 heads=true_heads mod 16 sectors=true_sectors mod 64 Although a SCSI controller shouldn't have these constraints, you never know till you look. 3. Depending upon what DOS you use, you may have a limit of 128MiB (Dos 4.X), 32MiB (Dos 3.X) or 16MiB (Dos 2.X and earlier). Your best bet for debugging is to upgrade to Dos 5.X+ and get smaller drives and use the largest one that works. 2. Should I leave the V20 in? I've heard it has compatibility issues I've never had a compatibility issue using a V30 in an 8086 machines, except with very early copy protected games... Fortunately I could select speed and wait states from the keyboard and found that 4.77 MHz with 3 wait states generally came close enough to an XT to fool the game. Eric