I somehow missed the very beginning of this discussion on Usenet, but the topic seemed clear and important enough nonetheless. Apparently several people have had instances in which they believe a SyQuest drive trashed their hard disk (so watch out for roaming SyQuest drives!). No one was able to confirm that the problem was absolutely linked to the drive and there is no reason to believe that the SyQuest mechanisms in general have problems. (For those wary of the SyQuest mechanisms, Iomega now has Bernoulli drives that are probably more reliable given their method of operation.) Some people suggested the possibility, and one with which I’ve fought recently, that there were problems with the SCSI chain which resulted in the damaged hard drives.
Essentially, the SCSI theory is that the first and last devices in the chain must be terminated. Everything else should not be terminated. I think the maximum length of a SCSI chain is 16′, but don’t hold me to that. The problem I ran into involved a Mac Plus, a LaserWriter SC (no termination), a Dataframe XP60 (internally terminated) and a PLI SyQuest drive (also internally terminated). The only way this setup would work was if the XP60 was hooked to either the LaserWriter SC or the PLI Infinity, but not both at once. The LaserWriter only worked when hooked to the XP60 if it wasn’t externally terminated, in defiance of the rules, and the PLI Infinity only worked if it was connected to the XP60 by a short cable, was the last device in the chain, and was turned on first. I tried invoking a few daemons, but they were no more effective than applying the rules.
In any event, if you are using a SyQuest drive and are having troubles, try to isolate the drive as much as possible and test it. Some problems are related to the fact that the mechanism is hot and the media doesn’t like the heat (for this reason BMUG (Berkeley Macintosh Users’ Group) recommends the DPI SyQuest drive). Other problems are undoubtedly related to the termination problems I mentioned above, so getting a SyQuest drive with external termination (APS SyQuest drives have been recommended on the net and have external termination) will make your SCSI trial-and-error session more productive. If you already have a drive with internal termination, the company you purchased it from can tell you how to remove the termination resistors, at which point you can buy a normal external terminator and try that. A final possibility is that internal termination takes its power from the drive, whereas external termination gets power from the Mac on some models, at least. Thus, if the SyQuest drive is turned off, the termination might not be working as expected. Of course, all bets are off if you are using a Mac IIfx.
Now that SCSI-2 has been ratified (ratified: to be thrown before the rats, or whatever standards committees do – actually I’m not sure of that except for the fact that the new NeXT machines have SCSI-2) Apple will of course implement it fully in the Mac IV and there will be no more problems. Or at least very few. In the meantime, I’m trying to remember those incantations.