PDA

View Full Version : CBM 8050: IN5402 diodes melted and fell off(!)



Kakemoms
May 1st, 2017, 03:52 AM
Hi

I just got another 8050 and it was almost working until two diodes (IN5402) on CR5 and CR6 went into IC heaven. One melted the solder around it and the other just cracked.

I have tested the two 6502s inside and they both work, so I may have been lucky. I have replaced all the capacitors (should have done so before I started using it) and I am waiting of new IN5402s. Both diodes were connected between the transformer and a LM323 (which is connected to a massive heatsink) which is capable of pulling a lot of current. The diodes are also connected to the positive side of a really large capacitor (just states 23K MFD 15V) that I probably can't replace.

I am not 100% sure what went wrong, but maybe someone here has seen the same. The primary suspect for now is the really large capacitor, but I have no idea how to find a spare one of these. You can see them in this picture (http://www.classiccmp.org/dunfield/pet/h/d8250i.jpg) of another 8050 (the large green ones).

Ah, and the MOS 6530-038 was also dead, but I suspect that was why the drive wasn't working properly (or it may have died with the IN5402). Anyway I have a batch of mini-6530 replacement cards in its way (design stolen from Ruud with the inspiration to miniaturization from Dwight Elevey).

What would you have done at this stage?

KC9UDX
May 1st, 2017, 04:12 AM
Disconnect the capacitors and measure their resistance. It may start out low but should rise to infinity as you keep measuring, the rate dependant on the impedance of your meter.

Replacing them should be no problem except that the replacement(s) will be considerably smaller.

KC9UDX
May 1st, 2017, 05:29 AM
By the way, I forgot to mention that you must discharge the capacitors before measuring resistance. Do that with a voltmeter.

Kakemoms
May 1st, 2017, 06:49 AM
By the way, I forgot to mention that you must discharge the capacitors before measuring resistance. Do that with a voltmeter.

Well, if I short it out it charges very slowly with the Ohm-meter. Using a voltmeter did not discharge it fully (or it was so slow I couldn't see it).

Is the 23000 designation for 23000uF?

KC9UDX
May 1st, 2017, 07:20 AM
Your voltmeter is high impedance, which means a large time constant for discharge (or charge).

Your capacitor is probably fine. If it didn't charge slowly, it would be shorted and would be your cause of failure. So you probably have a short somewhere else.

If it says 23000 on it, then probably, yes. Although that seems kind of high to me.

Kakemoms
September 12th, 2017, 11:01 AM
Some tips on Commodore IEEE disk repair.

Apart from the 8050 service manual which you can find around, I found this document quite useful to suggest additional components in the 8050 drive(s) (or 4040) that may need replacement:

http://www.vintagecomputer.net/commodore/IEEE_drives/CBM_IEEEDriveNotes.pdf

Scroll down to page 9 to read about 8050 suggestions or page 10+ for 2040/4040. There is also some notes on the 9060 and 1541.

Also, if you are not getting green LED or LED flashes, it means that the 6502 (in the 4040) or the 6502 at UN1 (in 8050) is not running. First thing is to check for +5V and GND, preferably with an oscilloscope to see that there are no ripples.

I have looked at the ROM code, and once the main 6502 starts, it will check ZP and start flashing LED if it fails. So (as the service manual says), check the clock if you do not have flashing. If that fails, change the 6502. If that fails, change the ROM at UH1 and/or UL1. If all that fails, it probably means you have a short somewhere (which would pull address/data lines low or high, or can leak data from one line to another), or in worst case a broken line or solder. In fact, many solders tend to form cracks on these old PCB's. A way to fix that is simply to remelt the solder joints.

All-in-all the old IEEE drives of Commodore are tedious to repair. While the flashing code helps, it doesn't always say the truth. Two of my drives had a bad 6530 which I had to replace with an adapter board+6532+27C64+7402 (see http://www.baltissen.org/newhtm/6530repl.htm). The Rockwell 6532 works ok in these drives, while Atari components didn't work for some reason (there is a version of the 6532 in early Atari machines). The 2114 SRAM chips also tend to die.

I hope this info helps if someone are trying to get one of these old beasts up and going again. ;)

MikeS
September 12th, 2017, 12:50 PM
Your capacitor is probably fine. If it didn't charge slowly, it would be shorted and would be your cause of failure. So you probably have a short somewhere else.

If it says 23000 on it, then probably, yes. Although that seems kind of high to me.
No, there really are two 23K caps in there and I agree, that's probably not the problem.
It might even just have been one of the diodes shorting out; if one shorts then the other will also blow, probably spectacularly.
You could always just replace the whole PS with a modern 5 & 12V switcher; less heat and certainly a lot less weight ;-)

m

RuudB
September 15th, 2017, 05:28 AM
... (design stolen from Ruud with the inspiration to miniaturization from Dwight Elevey). ...
That's the spirit :)