At 11:05 -0500 6/29/13, bear wrote: Agreed with bear Almost agreed with bear. This might be me, and if so, I'm flaky, but was telling the truth as I knew it. However the MO drive failed again shortly thereafter, so I gave up in frustration and never reported what I'd done. Apologies if I left anyone hanging! I have only one MO drive, so my statistics aren't as good as his, but yeah, I don't know of any that are still working. Mine worked for years, then quit working (more retries, then infinite retries). I took it out, got it far enough open to get a q-tip full of isopropanol to the lens, and cleaned the dust off the lens and the slide rails for the carriage. This was a fiddly process, involving de-mating connectors, switching back and forth between top and bottom of the drive, and figuring out how to prop the drive up with some parts hinged open. However I would not call it "perfectly impossible" - I did it, which probably means Tony could do it with the drive running and one hand tucked safely into his hip pocket. The unit ran normally when I got it back together, but for not more than a few months. I have not taken a second pass at it. Dust was likely an issue with my machine, because it had been running with the fan in the normal orientation (air comes in the drive and out the back of the machine). Good for cooling, but if there's any dust at all in the environment, it ends up forming bunnies on the MO lens. The fan is mechanically symmetrical enough that it's easy to invert it to run "backward". Air comes in the fan, dust cakes on the rest of the machine, and nothing settles in the MO drive. Good for optics - but bad for cooling. On my 68040 cube, this led to very occasional crashes due to SCSI system faults. Since they happened with at least two different hard drives which were otherwise faultless, I deduce it was something to do with the motherboard, which was then getting the warmest air in the box blowing across it. (The crashes happened only during my morning "clean out old .nfs files" cron job, which did a "find / -name ....". I diagnosed them by distributed.net log file messages.) I put my fan back like it was, and the SCSI crashes stopped. Dust was clearly not the *only* issue with my machine. I surmise that it was exacerbating whatever slow degradation process affects all of them, and I bought myself a few extra months with the cleaning. More details available if anybody really needs them; I still have the drive and the rest of the system, and can repeat the disassembly I think, but not just immediately - I'm using the machine at the moment. FWIW, MO drive retry errors hang the machine - no monitor mode or good way to continue. The low-level drivers for the MO presumably operate at high priority, and block interrupts. SCSI problems are not so bad, which is yet another reason that if you have a NeXT running on a SCSI drive, switching to MO is a downgrade. They were a technology with an extremely high "cool" factor but a limited timespan of cost-effectiveness. Agreed with bear on that one, too. I have not looked into trying to find a replacement for the optical head. I suspect that would be pretty tough. At 18:23 -0500 6/30/13, <Dave R.> wrote: Checking/replacing all the caps would take a while. The drives have a lot of parts. However this is as good an explanation as any I've heard. FWIW, there was no smoke from mine, then or yet (it's still powered, in my cube). Sorry for slow response, hope this is at least some useful. -- - Mark 210-379-4635 ----------------------------------------------------------------------- Large Asteroids headed toward planets inhabited by beings that don't have technology adequate to stop them: Think of it as Evolution in Fast-Forward.