[GanjaTron]

Thanks for the pix, weblacky!

That board's actually denser than on the XM-3401. I couldn't bend the leads back as you did with some of the replacement caps as I had to keep them short so the caps wouldn't protrude into a cutout for the CD spindle motor. As in the floptical, there's not much headroom between the PCB and the mech, but the 4x7mm 10µF's were a pretty good fit. Dropped them in and the drive works again.

   

But... Achtung, Baby! There are two bipolar (!) SMD caps on the PCB. Lacking viable replacements (and space for two back-to-back polarised electrolytics), I left the originals. They still measure ok, but the capacitance is a bit on the low side (<9µF). I expect these aren't stressed as much as the polarised caps. Should replace them at some point, probably with SMD bipolars again.

Now if I could only get my old HDDs to work too... the ST43400 _seriously_ acted up again as I was testing the CDROM, then came to life about 1/2h after tons of failed startups and SCSI bus resets. I have 3 of these bastards, and they all developed the same issue. But that's another story...