The Backstory

Recently I had become involved with this b-bit revolution that has been going on. I remember, that perhaps the first computer I had before getting the IBM PS/2 model 70 was a Commodore SX-64, and man it did everything! From the dot matrix printing to programming of certain radios of Midland and Maxon to actual programming of the old cellular from GTE Mobilenet, that Commodore surely did quite a lot! Did I mention I loved Zaxxon an Impossible Mission that were made for that system?


In the later 2000's I had sold it off for what I thought to be a considerable amount of money, games an accessories all, and figured time and technology needs to move on, and for me that was the end of the I thought....


Stumbling upon the local gaming convention here in Cleveland an as well The 8-bit Guy, I was surprised to see a retro revolution with 8-bit computing and how it's been re-innovated to do more than it did 36 years ago--in seeing and knowing that I was hooked and had to find a machine again that I sold decades earlier. I had found one on Amazon Marketplace and it looked to be pristine--it was an unbelievable find (even the original keyboard cable)--so I went ahead to purchase it, only to know I should have asked one question; was this system modified in any way? Sadly, I had to open her open to find a mess inside.


Don't get me wrong it worked for a time when I got it, but it was riddled with toggle switches in its storage compartment, only for powering on and off the selected devices within the machine, and then one for JIffyDOS and the standard pre-loaded BASIC. The other thing I notice was that there were bare wire splicing covered loosely in electrical tape that had been melted away by the heat of the machine; I was surprised that nothing had shorted out from the exposed bare wire, especially near the CRT. I was not happy what I was seeing, since the expansion cable was torn to bits for these toggle switches and the sloppy splicing.


After a few weeks of have it here, I was going to use my SD2IEC device to play some disk images and was re-assigning the device ID. The commands were not a problem, as they went through, but as I rounded for home, the CRT flickered and rolled and blanked out.....obviously I don't have to tell anyone here that was not a good thing!


So rewind to January.....


I contacted a Commodore guru out of Washington state and he had told me that if I had replaced all the chips that were known to fail (VIC, PLA, MPU) it would be ideal to use the multi-meter and see if the correct voltage was being applied to said chips (5 volts). Taking some initial readings I found out that it was rather inconclusive seeing that what I thought was the VCC looked normal, but the one dead giveaway that there was no 5 volts was because the disk drive light never came on! The spindle and drive motor ran on 12 volts and the drive light was independent of that--it was then I knew I had to chase into the power supply.


My first Switch Mode Power Supply


I have never messed with a power supply like this before. Mostly a linear supply may have been all that I had dealt with and maybe some AC rectifying, but this was a whole new beast and it was 36 years old too. Finding parts I knew was going to be a hard task as well, if not the equivalents to the original parts I wasn't sure would guarantee that it would work neither.

So I am going to fast forward a bit here and just say what was done and what was wrong when it was removed completely from the system. At first it seemed that the 12 volts was still being supplied but after a time that even quit. It was observed that the fuse kept blowing and that was because the one power transistor after the full bridge rectifier would short out and take out the PWM IC on the 12 voltage board as well! Those have been the three components that have been replaced consistently with no real answer as to how come. All electrolytic capacitors and mica one have been checked with an ESR meter to test for resistance, and everything comes back fine. I have tested that actual resistors and diodes and still no shorts or erratic resists readings at all!


If you are wondering what about the 5 volts in all this, well, there was one inductor that was removed from circuit to isolate the 5 volts from everything else. It was found out that the 5 volts was not producing because of a PNP pass-through transistor, which for the moment we are using a NTE replacement part for testing purposes. There is no load to the supply so this transistor has not shorted out on us at all. The correct ones I do have on hand and did come from element14 lol.


So to this point, about 2/3 of this supply is working. There is still no definitive reason as to how come I keep loosing the fuse, power transistor and PWM IC. With all that has been tested and done one would have thought the problem would have reared its head already, but still everything is not OK just yet. The last thing I suspected but have no proved yet were the cold solder joints of the 12 volt daughter-board seeing one was able to rock it back and forth with no effort at all. I had touched that up and it's solid now to the main circuit board. Everything has been put back into place and ready for another test run, I am just unsure that this will solve the issue or burn up more components.


I was told from the start that troubleshooting a power supply even like this is very tedious and the issue could exist just about anywhere. In the whole of the Commodore community, there is no real documentation or any others out there whom talk or speak about the SX-64 power supply, seeing from what I have found out it just keeps running. I see tons of issues with the power bricks to the breadbox Commodore 64 and tons of solutions that way, but none to the effect of solving a switch mode supply and how come it would fail.



If anyone has any further suggestions or happens to be an expert in this field, I appreciate any feedback at all. The only thing left on my mind is that this may be a lost cause and the supply is beyond saving already.