VOGONS


First post, by Kahenraz

User metadata
Rank l33t
Rank
l33t

I have two Compaq LTE laptops with floppy drives, both of which failed to work. Disassembly revealed that they contain a single drive belt that had gone soft and was beginning to melt.

My measurements are 1x0.5mmx23mm-ish, 11.5mm folded. The length is only an estimate, as the belt was no longer well formed.

Finding an adequate replacement will be difficult, I think, since it is much thinner than the standard 1.25mm kits found online, and especially because it isn't a square belt.

I believe that most belts of this vintage are well past their expiration date, and I would highly recommend that you replace any of yours, regardless of whether or not they still work. The consequence is a tar-like slime inside of a mechanical assembly that can be difficult or even impossible to clean.

I'm starting to see this probablem more and more. And I expect that it will only get worse over time.

The attachment 20221126_195154_resize_89.jpg is no longer available
The attachment 20221126_195553_resize_68.jpg is no longer available
The attachment 20221126_203446_resize_99.jpg is no longer available
Last edited by Kahenraz on 2022-11-27, 02:04. Edited 4 times in total.

Reply 1 of 14, by pentiumspeed

User metadata
Rank l33t
Rank
l33t

Infamous W1D indeed.

Yes, there a seller selling new belts for these on ebay.

Cheers,

Great Northern aka Canada.

Reply 3 of 14, by pentiumspeed

User metadata
Rank l33t
Rank
l33t

Yes that one. I got two from this while ago.

Cheers,

Great Northern aka Canada.

Reply 4 of 14, by Kahenraz

User metadata
Rank l33t
Rank
l33t

I've ordered two as well. I hope that I can get at least one working drive.

Reply 5 of 14, by Kahenraz

User metadata
Rank l33t
Rank
l33t

This drive appears to be very similar to the NEC FD1238T. I wonder how close they really are.

The attachment Screenshot_20221126-222020_eBay.jpg is no longer available

Reply 6 of 14, by PD2JK

User metadata
Rank Oldbie
Rank
Oldbie

I'm expecting a new belt for a Matsushita EME279TD drive in a few days. The drive internals from the photo looks alike.

i386 16 ⇒ i486 DX4 100 ⇒ Pentium MMX 200 ⇒ Athlon Orion 700 | TB 1000 ⇒ AthlonXP 1700+ ⇒ Opteron 165 ⇒ Dual Opteron 856

Reply 7 of 14, by Kahenraz

User metadata
Rank l33t
Rank
l33t

It took a while, but I finally got my belts. This fixed the disk spinning, but neither one of my drives seem to work. One of them I know never worked (it came with a used laptop years ago), but I wasn't sure if a new belt would help. The other one only works partially; it will read a few sectors but then will always fail. My guess is that I somehow knocked the heads out of alignment.

I looked at the drive carefully and found test points on the bottom, but I couldn't find any way to reliably adjust any of the alignment characteristics, even if I could monitor it on a scope.

Both drives are otherwise in excellent physical condition. I'm not sure what to do with them now. And getting a replacement might not be worth the trouble.

The attachment 20221218_093444.jpg is no longer available

Reply 8 of 14, by Kahenraz

User metadata
Rank l33t
Rank
l33t

I had nothing to lose by tinkering with this further, so I took stock of the current status of the drives and started experimenting to see if I could find a way to get a single working drive between the two.

One of the drives is completely dead no matter how much I touch it. The heads simply refuse to read track 0. The stepper motor will my always move the back to the correct position, but it can never read the track. I even tried a head swap, and this did not improve things, which suggests that it is the heads.

The other unit, which worked fine years ago, is no longer able to read HD (1.44MB) disks fully, always failing somewhere between tracks 58-60, near the end of the disk. The drive is able to cope with DD (740K) disks, with their wider tracks, so it is at least partially usable.

I did my more volatile tinkering on the dead unit, until I found out how to realign the track 0 sensor and the top head. The bottom head appears to be fixed. The heads on the dead drive appeared to be perfectly aligned, and the heads on the partially working unit were not; the top head was noticably askew. I carefully realigned it by hand, but this did not improve or worsen its ability to read HD and DD disks.

Does anyone know what physical characteristics would cause the drive to fail to read the inner tracks while being able to read the outter tracks just fine?

Also of note, the partially working heads do not work at all in the other body. I don't know why this is, but the two units, both Citizen W1Ds, are slightly different, so maybe there is some subtle incompatibility between the two revisions.

The attachment 20221220_100733.jpg is no longer available
The attachment 20221220_102104.jpg is no longer available
The attachment 20221220_103439.jpg is no longer available

Reply 9 of 14, by Kahenraz

User metadata
Rank l33t
Rank
l33t

I was very bothered by the failed unit simply not working whatsoever. The error was always a "general failure", with other parts of the drive working, such as disk detection. Upon closer inspection, I found a spot of corrosion. My multimeter read a break between the center trace, and scraping the solder mask away revealed that it was on fact severed.

I'm now regretting having futzed with the head alignment on this drive, if a repair was as simple as a single broken trace.

The attachment 20221220_113943.jpg is no longer available
The attachment 20221220_113923.jpg is no longer available
The attachment 20221220_115701.jpg is no longer available
The attachment 20221220_114739.jpg is no longer available
The attachment 20221220_115303.jpg is no longer available

Reply 10 of 14, by Kahenraz

User metadata
Rank l33t
Rank
l33t

The repair was simple enough. I hope it works! Here is my thumb for scale.

The "wire" I use for repairing traces is a single strand of solder wick. These are great, as they can be obtained as pure copper and come pre-coated in flux.

The attachment 20221220_123207.jpg is no longer available
The attachment 20221220_123231.jpg is no longer available

Reply 11 of 14, by Kahenraz

User metadata
Rank l33t
Rank
l33t

Progress has moved on this drive from "general failure" to "sector not found". I also confirmed that neither head works when swapped between the two drives, regardless of this progress. It still can't find sector 0, but the error is different.

Reply 12 of 14, by Kahenraz

User metadata
Rank l33t
Rank
l33t

I wasn't happy with being able to read only 740k disks, so I kept tweaking the head alignment while running a diskette diagnostic that would validate the drive. I got pretty close until I noticed that the results started to drift between tests of the same adjustment. So I tried tightening the screws a bit more and... the plastic broke away.

This is a problem because I was only ever able to get these heads to work in this drive. The other heads didn't work, or at least one of the two, there's no way to tell. So I did the only logical next step. I disassembled the head mechanism from the non-working drive, transplanted the working heads onto it, then moved it back into the drive known to work with them. And after LOT of careful tuning of both the top and the bottom heads, I finally managed to get them realigned; not only for 740k, but also for 1.44mb!

I've been working on this all day and I'm so well tuned to this drive that I'm resonating with it. What an absolute nightmare! I NEVER want to do this again EVER. This is by far one of the most tedious and mind numbingly infuriating thing one could ever hope to do with retro tech. Good riddance. But thankfully, I have a working drive now!

Hopefully it will stay aligned now. At least for a while.

Here are some photos.

The attachment 20221221_025138.jpg is no longer available
The attachment 20221221_024844.jpg is no longer available
The attachment 20221221_024903_resize_7.jpg is no longer available
The attachment 20221221_030859.jpg is no longer available
The attachment 20221221_031214(0).jpg is no longer available

Reply 13 of 14, by Kahenraz

User metadata
Rank l33t
Rank
l33t

Here are some more photos I took throughout the process.

The attachment 20221218_010010.jpg is no longer available
The attachment 20221218_052205.jpg is no longer available
The attachment 20221221_033253.jpg is no longer available
The attachment 20221220_102104.jpg is no longer available
The attachment 20221220_100739.jpg is no longer available

Reply 14 of 14, by Bondi

User metadata
Rank Oldbie
Rank
Oldbie

Great job, Kahenraz. Yeah, sometimes a seemingly easy work becomes a nightmare. Really can understand your frustration while going down that rabbit hole. Yet, the relief in the end and a working drive was probably worth it 😁

PCMCIA Sound Cards chart
archive.org: PCMCIA software, manuals, drivers