386_junkie wrote:

PeterLI wrote:

The PS/2 40 is an ISA machine.

It is... I wonder what they were thinking in the manufacture for this one, possibly an experiment to see how it is taken by the market of the time? Though dropping their own prop bus for the widely used ISA... is not very IBM is it?

I suspect that they needed to have both MCA and ISA in their line of PS/2s simply because the market was predominantly ISA. Getting an ISA based PS/2 meant that people would become familiar with the PS/2 product in general, and IBM was betting that people would simply stick with it and get an MCA based machine next. At least, thats what I've heard over the years...

For what it's worth, both my PS/2 Model 40 and my PS/1 Model 2121 are ISA based. Yay for soundblaster support!

@386_junkie:

Yay! My IBM Model 40SX (2.88 MB Sony) and IBM Modell 30 8088 with 1.44 MB Mitsubishi need new Electrolytic Caps and my 40SX Mainboard also 😀

Greetings

orinoko wrote:

For what it's worth, both my PS/2 Model 40 and my PS/1 Model 2121 are ISA based. Yay for soundblaster support!

Ha... yea, now where did I see that MCA soundcard....

Dubinius wrote:

@386_junkie:

Yay! My IBM Model 40SX (2.88 MB Sony) and IBM Modell 30 8088 with 1.44 MB Mitsubishi need new Electrolytic Caps and my 40SX Mainboard also 😀

Greetings

Greetings,

It sounds like you've got a busy afternoon some day! 🤣

I love to solder... I could do all day, every. Unfortunate that there is no money in it. 😒

All the best with the repairs...

I did the same thing with my PS/2 P70's ALPS drive, but my desktop 70's drive (also alps) seems to not work.
I have another Mitsumi drive that has some mechanical problem, but a solder pad is ripped off from where a cap is supposed to be.
Pics or it didn't happen:

I also fixed one ibm floppy drive, by replacing the caps. Have a other drive that needs new caps, switched out 2 i already had, and the drive started spinning again. Need to by a lot of caps, since i need to repair a few other things.

I do have also ALPS drive on my ibm model 95XP(i guess this is very rare pc as haven't seen anyone sell one..) and now i came conclusion after many tries it just not read disk.
so changing caps is needed.

armankordi please can you tell me the right value for the cap rated 10 ? is it just 10uF?
and there is 3 of 4.7 caps without any polarisation, any tips?
nice job you done on your P70 (also have P70 i guess same thing there is needed to do..)

riku wrote:

I do have also ALPS drive on my ibm model 95XP(i guess this is very rare pc as haven't seen anyone sell one..) and now i came co […]
Show full quote

I do have also ALPS drive on my ibm model 95XP(i guess this is very rare pc as haven't seen anyone sell one..) and now i came conclusion after many tries it just not read disk.
so changing caps is needed.

armankordi please can you tell me the right value for the cap rated 10 ? is it just 10uF?
and there is 3 of 4.7 caps without any polarisation, any tips?
nice job you done on your P70 (also have P70 i guess same thing there is needed to do..)

Ok, have since bought another drive for the (desktop) 70, and replaced all caps on the board.
Did not work.
About the no polar caps (solid silver ones) for the P70 I ended up just not replacing them, drive functions normal. I've also noticed that some alps drives despite being manufactured in the same year have different designs. The one I bought has some mechanical differences and no doubt electric ones too. Keep your eye peeled, and buy all new caps for the drive just to be safe.
You'll want
3x 4.7uf 16v non polar
1x 4.7uf 25v
1x 10uf 16v
The new drive I bought, also has a weird issue where the motor spins TOO fast, and sometimes doesn't spin.( After I replaced caps)

Can I use standard through-hole electrolytic capacitors instead of aluminium SMDs?

Miphee wrote:

Can I use standard through-hole electrolytic capacitors instead of aluminium SMDs?

Yes, so long as ratings / specifications match, there is no difference as they do the same thing.

Make sure you solder with correct polarity!

I have another faulty ALPS 1,44 drive that doesn't read disks.
I recapped it, cleaned and greased it but still nothing. Head is also clean.
Drive is detected at startup and the head is moving fine but when it comes to booting the head only flinches once, disk rolling continuously, LED stays lit forever and nothing else happens. System also hangs like it's hopelessly trying to read the disk.
I compared it to my other PS/2 drives of the same model, they are 100% identical and nothing is missing. I didn't take apart or modify any moving parts.
I've had a drive before that did this (LED stayed on forever) but recapping worked on that one.
Any tips?

386_junkie wrote on 2016-01-25, 11:42:

P.S. If anyone is in the same situ, let me know and I can help those wishing to restore their drive.... it beats $50+ for another one!

Hi, I just now stumbled upon this post and I wanted to ask if you could help me. I just spent 4 hours debugging my IBM PS/2, model 50 with some friends. We weren't able to get the system to boot off of a reference disk (found at http://www.walshcomptech.com/selectpccbbs/, under the "Model 50, 50z" section) and suspect it's an issue with the floppy drive itself. Do you know of anything else we can try before trying to repair/replace the floppy drive?

I did find on the Internet that Windows messes with the "magic byte signature" needed on the reference disk, and I used the REFSTAMP tool found at http://www.tavi.co.uk/ps2pages/refstamp.html to try and make sure the magic signature was there. However, I'm not sure if that really worked, since I'm using REFSTAMP from within a DOS emulator (DOSbox) and I used a USB-based floppy drive.

We also found that by removing the floppy drive/hard drive, we could get the computer to boot into BASIC from the ROM instead. However, there didn't seem to be much in the way of debugging to be done there.

That sounds about right. Welcome to the IBM club!

I use a windows machine and a program called WINIMAGE to make my disks. I am quick to remove the disk after writing before windows has an opportunity to poke its fingers into it. You should not need the refstamp tool with winimage. at least in my experience.

It is very common for these drives to need a proper full service. cleaning the heads at a bare minimum. The re-cap procedure is 'relatively' new and I have a literal pile of drives waiting for me to service them.

I would recommend practicing your soldering on something else though. ^.^

Gotcha. I figured the re-capping procedure wouldn't be easy, but I, stupidly, decided to go ahead and try it anyway. It almost went great, but I managed to tear a pad on the last capacitor, so I'm pretty sure the drive is shot now (unless it can function without the capacitor). Do you know of any place I can find a replacement drive (preferably one that's confirmed working)?

A cursory eBay search reveals a guy selling a bunch of different IBM PS/2 floppy drives, but they're all listed as "untested, sold as-is due to being 30+ years old". I'd rather not spend $45 on a maybe-working item, unless I could reliably fix it.

Alternately, is there any repair service for these drives? That way, I could get the maybe-working drive and let a professional handle the restoration, instead of messing it up myself...

I'm doing a little more research now, and it looks like there are relatively cheap (compared to the $45 mystery drive) adaptor boards that convert from a more modern IDE connector to the PS/2 connector, like this one: https://texelec.com/product/ibm-ps2-to-standa … floppy-adapter/. Do you think that might be a good option? Has anyone used that before, and could you tell me us if it worked OK for you?

I have a few as backups. I have not used them yet.

Don't forget a sound card from that site if you wNt anything better than the pc speaker. ^.^

Well, I went ahead and ordered the TexElec adaptor and an IDE floppy drive, and installed them both. I also wrote a brand new reference disk using the Linux "dd" utility, then flipped the write-protection switch and used my Windows computer to extract a disk image. I then compared that disk image with the original image file, and the hashes (CRC-32 as well as SHA-256) matched. In theory, this means I have a perfectly working reference disk, right?

Or so I thought. The PS/2 still shows the exact same behavior--turns on, runs a memory check, beeps twice and shows the 161 and 163 codes, then takes me to the "refer to the manual" screen. I'm really lost as to what else I can try here. Any suggestions? Otherwise I might have to shelf this project indefinitely.

EDIT: I did one more thing, I ran the REFSTAMP utility I mentioned earlier on this "working" reference disk (in write-protected mode) and it's not stamped!? I thought writing the image would take care of that for me. I'll see if there's a way to do that in Linux too.

EDIT 2: I'm not really seeing anything on Linux that can do this. I can't use the REFSTAMP util in my DOS emulator to stamp the disk, because that would allow Windows to come through and mess with it. I'm really at a loss for what to do now, because unless a DOS machine shows up at my doorstep, I can't make the diskette bootable.

I have two 1.44mb ALPS drives and can’t get either to work. Both had ugly capacitor leakage One ended up having some torn pads on the motor board so that one is hopeless.

The other the LED lights up and the disc spins, but no movement from the heads whatsoever. The worm shaft moves easily so not seized up. I even tried switching the motor and controller board with the other drive just to see but no difference. I’ve spent way too much time trying to get these to work but I’m still annoyed and don’t know where else to go when the head assembly won’t even try to move.

Could be a problem with the signals that tell the drive to step the heads, and direction. Possibly some traces got eaten away by the corrosion you mentioned. Did you try running IMD and testing the head movement in alignment test?

polishvito wrote on 2023-03-08, 14:19:

I have two 1.44mb ALPS drives and can’t get either to work. Both had ugly capacitor leakage One ended up having some torn pads on the motor board so that one is hopeless.

The other the LED lights up and the disc spins, but no movement from the heads whatsoever. The worm shaft moves easily so not seized up. I even tried switching the motor and controller board with the other drive just to see but no difference. I’ve spent way too much time trying to get these to work but I’m still annoyed and don’t know where else to go when the head assembly won’t even try to move.

I just recently repaired an ALPS drive (DFP723D33A, 34 pin, media-sense) on my recently acquired model 35SX. It was just a recap and cleaning job, essentially, and now the drive works perfectly.

Couple things to check. Like previous poster said, there might be corrosion from the electrolyte on the traces so it might prevent the drive from working. Another thing to note is that electrolyte is conductive, so if you havent't cleaned that crap from the PCB, it may short some stuff and drive doesn't work because of that.

Torn pad isn't necessarily a deal breaker. In fact, I managed to tear one during my recap work due to my mistake: I didn't want to use hot air for desoldering (so much plastic and rubber near the capacitors) so I went with a twisting method and screwed up with one cap. However, it was relatively easy to trace from the pad to the next SMD resistor and I just soldered the new cap to that point. One thing I almost missed before starting the recap is that in my drive there are three bi-polar capacitors on the motor board. "Normal" polarized electrolytic caps won't work there, but this should not affect the head movement at all, naturally. But your drive might be of a different type and there may be bi-polar caps on the control side too, so check that you have used correct caps. If you replaced bi-polar caps with polarized ones, that may cause the problem.

Also, I specifically used electrolytic caps as replacement, like the originals were. Reason is that I wanted to be sure that the drive works after recap 100%. Changing cap type to tantalum or ceramic changes capacitor ESR dramatically to lower value and I have no clue how sensitive that circuit is for ESR. So, to be sure that circuit still has 100% identical specs after recapping, I went with electrolytic caps.

One thing worth checking is the possible ribbon/flat flex cables. They might have fractures and that might be the reason why yours isn't working.