IBM PC XT 5160 5.25" drive not working...

[profdc9]

I am looking for a service manual for a IBM PC XT 5160 5.25" drive. However, it is NOT a Tandon drive.

It has a part # 0384-002 and 6181147. There are pictures of it in this thread

https://forum.vcfed.org/index.php?threads/ibm-360k-floppy-drive-not-tandon.70394/

The problem is that for a while it was "dying" and would only start working when the computer was on for 10 or 15 minutes, not responding, or only spinning and not reading correctly, until it had been on for a certain amount of time. I put it away in storage for two years, and now it appears to be completely dead, not even spinning or attempting to read.

I know that frequently the tantalum capacitors are to be suspected for old IBM equipment, and that they generally fail short. There are numerous 1/4" resistor-sized yellow components on the board, and tantalums are usually yellow, however I can not identify a band on them to indicate polarity, which tantalums generally have. I don't feel any of the components getting hot, however, and if a capacitor did fail short, I would expect that the power supply would trip if the current exceeded safe limits.

There are also two polarized caps near the 4-pin power connector, which could be tantalums, but they are blue-black in color. There is a small toroid inductor next to them, perhaps to filter noise from the power?

There is a Texas Instruments 40-pin socketed chip DCU8423, dated 1981. I have no idea what this is, perhaps a microcontroller? SN75478 driver ICs perhaps for the drive head stepper motor?

Before I go on an extensive teardown, are there any obvious things to look at, or common problems? Often a service manual will indicate those, but I can only find a service manual for the Tandon drive, not the drive I have. For example, there is a 2N6667 Darlington power transistor which is probably to turn the spindle motor on and off.

Thanks for any guidance!

 

[modem7]

I am looking for a service manual for a IBM PC XT 5160 5.25" drive. However, it is NOT a Tandon drive.

Well, perhaps a circuit diagram, not a service manual.
Navigation:

1. Goto https://minuszerodegrees.net
2. {Manuals} section
3. {Third Party}
4. {IBM}
5. {IBM Options and Adapters - 6322509 - 1984}
6. {IBM 5 1/4" Diskette Drive}

That document has circuit diagrams for 3 make-models that IBM supplied, the document describing them as types 1, 2 and 3.
Maybe your drive is one of those.

 

[Chuck(G)]

I believe that 3 different drives are in the O&A Techref--one is the Tandon, another is a CDC, the third might be Alps--I don't recall.

 

[profdc9]

Ok, I found out the drive is the 3rd type specified in the documentation. Unfortunately the scanned schematic is low resolution and the letters and numbers are very difficult to make out, and I haven't been able to successfully use it.

So I have left it on for an hour and the drive seems to be working intermittently now. It can even POST without a 601 error most of the time.

If the problem is a bad electrolytic or tantalum, would there be a way to determine that short of replacing all of them? I was trying to measure the voltage across them, and I actually fried two traces as the traces are so close together and so very thin that they are very easy to bridge even with a sharp-tipped probe. I had to solder in jumpers to replace the two traces. It is a mystery to me why someone would have such super thin traces (0.25 mm) for power on a board like that, but I suppose they had their reasons.

Often when I have repaired equipment, electrolytic capacitors tend to depolarize over time if the device is not turned on and the potential reapplied so that the dielectric oxide layer can regrow. I will probably leave the computer on for 24 or 48 hours to ensure that the capacitors regrow their dielectric if possible. Nevertheless, if there is a faulty one, I would like to identify it.

Otherwise, what other kinds of parts might be intermittent, only working after the computer has been on for an hour? I am currently running the computer with the cover off and the floppy drive sitting on the case, so I don't think there's some sort of heating going on, as would happen if the heat was confined to the enclosed case. But I could be wrong about this.

Thanks for any advice.

 

[SomeGuy]

Have you tried a different power supply? Have you tried it with as little installed in the machine as possible? (disconnect any hard drive, for example) I've had power related issues cause problems with floppy drives before.

If it is a thermal issue, you might be able to use a fan or compressed air to narrow down the fault. I've seen failing electrolytic capacitor fall out of spec and cause problems as they heat up. But other components can have thermal issues too.

I'm assuming this is a belt driven drive. It might be a good idea to gently clean off the drive belt to make sure nothing is slipping.

If it were me, I'd set up a boot disk that runs something like a disk test in a loop and leave it running for a while when it works. I've had a couple of drives that just needed "exercise" before they worked reliably.

 

[profdc9]

Have you tried a different power supply? Have you tried it with as little installed in the machine as possible? (disconnect any hard drive, for example) I've had power related issues cause problems with floppy drives before.

If it is a thermal issue, you might be able to use a fan or compressed air to narrow down the fault. I've seen failing electrolytic capacitor fall out of spec and cause problems as they heat up. But other components can have thermal issues too.

I'm assuming this is a belt driven drive. It might be a good idea to gently clean off the drive belt to make sure nothing is slipping.

If it were me, I'd set up a boot disk that runs something like a disk test in a loop and leave it running for a while when it works. I've had a couple of drives that just needed "exercise" before they worked reliably.

Thanks for the advice. Here is what I've tried:

I used a hair dryer to heat up the board to see if that would cause failure, or cause it to start working if it was not working. It did not seem to change anything, even though the board was probably heated over 40 C, and I didn't want to go any hotter.

I also removed all of the components except the hard disk drive, but I did remove the hard disk controller. I could not remove the hard disk drive because it's kind of jammed in there and I couldn't get it out to unplug it. Nevertheless, with all of the other cards removed except for the floppy controller card and Hercules graphics card (removing the 256k RAM card, serial port, network interface, hard disk card, parallel port card, and a serial port card) the floppy drive still did not work.

I measured the voltages on the floppy drive, and it measures 5.020 V and 11.990 V, so close enough to work. This is with none of the hardware removed.

When I first turn it on, sometimes the floppy drive fails, and the floppy LED does not turn on and the spindle does not spin. The computer POSTs with a 601 error. After I leave it on for a few minutes, it might spin the drive when it starts up with the LED on, however, it still reports a 601 error. If I leave it on for an hour, it will POST most of the time without the 601 error. I am looking for a spare 360k floppy I can sacrifice for this, they're not easy to come by, but I am going to borrow one to try and perform an endurance test.

The belt does not seem dirty. I suppose I could clean it with some cotton swabs and isopropanol?

Perhaps you could help me identify what some of these components are? I have attached a picture of the PCB. So there are axial-leaded yellow components. Are those tantalum capacitors? There appears to be disc capacitors as well, also yellow. There are what clearly are electrolytic capacitors, which are axial leaded and blue-black and have a taper on one side. I somewhat suspect the two capacitors near the power connector at the lower-right hand part of the picture. There's also a silver colored 22 uF, 35 VDC electrolytic near the heatsinked transistor which I believe is for the motor. Perhaps I should replace that one?