5

I have a Commodore PC-1 with a busted power supply. I have no desire to fix it as

  • I don't have the know-how
  • it is already 30 years old
  • It is difficult to debug / find replacement components

So I would rather opt for a new power supply

So I stumbled upon this Mean Well RPT-60B 60W PSU.

Now the original PSU was rated at 100W and according to the original PSU schema :

enter image description here

I see the following current ratings

  • 5V : 3,5A
  • 12V : 2,2A
  • -12V : 0,2A

The new Mean Well PSU although only rated at 60W has

  • 5V : 5A
  • 12V : 2A
  • -12V : 0.5A

Would this PSU be a good replacement for the original Commodore PC-1 PSU ?

Also, I currently hacked together "an ATX -> PC-1 adapter" for hooking up an ATX PSU.

enter image description here

I did some power measurements (putting my multimeter in-circuit to measure the current on the 5V and 12V rails, and I noticed that I'm only getting 7mA on the 5V rail and 220mA on the 12V rail. This seems very low (especially on the 5V rail). Any idea what might be causing this ?

The computer is running fine, and even during benchmarks / gaming the current measurement hardly changes. Any ideas ?

10
  • The outputs you have listed total about 50W, but the circuit diagram is apparently also supplying power to some of the pins on connector K1. Does that account for the other 50W of the "100W" rating when K1 is being used? Presumably, your replacement power supply doesn't have any outputs corresponding to K1 and the "60W" rating is consistent with the three outputs you tabulated in the question.
    – alephzero
    Feb 9, 2019 at 13:03
  • 2
    I'm not surprised that the 12V current is low, since it probably isn't used for anything much apart from RS232 serial output. But 7mA on the 5V line for a system that is working properly sounds like "sorry, your measurement is wrong" to me.
    – alephzero
    Feb 9, 2019 at 13:12
  • @alephzero: exactly ... however, it is measured exactly the same way as the 12v rail. Now ... there are 2 5V pins / wires going into the mainboard. i can only measure one at a time (only 1 multimeter). I image the current will be distributed between the 2 cables somehow. Could it be that the multimeter is causing some kind of side-effect on this ? Cause I once tried it by only hooking up 1 5V cable to the mainboard, and then I measured about 1200mA, but the computer didn't start at that point.
    – ddewaele
    Feb 9, 2019 at 14:21
  • @alephzero: Where exactly are you seeing the K1 connector ? On the PSU diagram ?
    – ddewaele
    Feb 9, 2019 at 14:28
  • 1
    The meter will be inserting extra resistance in the cable you are measuring, so unless each cable is powering a completely separate circuit (unlikely) it will affect the way the current is split between the two 5V cables. If you don't have two meters, try connecting one cable from the PSU to the meter, and two from the other side of the meter to the two connector pins. That way you should be measuring the total current. You won't damage the cables by overheating - ATX cables will handle much more than 3.5A maximum.
    – alephzero
    Feb 9, 2019 at 14:39

2 Answers 2

4

If you tested with:

PSU -> ammeter -> 2-wires -> J9-1, J9-2

And the computer did not start (but works with the meter out of the circuit) then I would suspect the meter is not capable of passing the current, or something is wrong in the setup. Drawing 1200mA sounds more reasonable for the +5V line, but it is unclear how the meter would report this and yet the computer does not start.

The specifications for the Mean Well PSU you found look reasonable. It provides more power than the original in all cases except for the ".2" (200mA) on the +12, which I would not expect to be a problem.

I'm not sure I believe your first statement, "I don't have the know-how", to get the original fixed. Anyone who owns an meter and who is willing to start busting out alternative power supplies and wire them up, seems like they are ready to take that next step. ;-)

The original PSU looks fairly basic, and unless the main transformer is bad, you should be able to locate any failed parts via Mouser, Digikey, etc.

If the system works with the alternate PSU, and nothing seems out of the ordinary, i.e. you are not letting the smoke out of any parts, nothing is getting hotter than expected, etc. then why not just use the new PSU?

I understand wanting to make sure everything is working correctly, however, without a known base-line of what a working system should be consuming on each power rail, the best you can do is guess. I suppose you could identify all the active components and total up their typical power ratings from datasheets (which would be a lot of effort), and use that as a base-line. Or, find someone else with that system and ask them to make current measurements (this is typically a tall request for most hobbyists / collectors).

Alternatively you can invest in an adjustable power supply to test with, which will then show you exactly how much current is being drawn. This can help you make sure the new PSU will be sized correctly, and can help let you know if something is going wrong on the system board. You can also control the amount of current going to the system, which can help prevent blowing things up if something is going wrong on the board.

1

Side note: If anyone reading this wants to re-cap the Commodore PC-1 PSU, I made a BOM, as well as a rough schematic and board layout: github.com/nbolton/cpc1-psu

To answer the OP question, I imagine that Meanwell PSU will probably be fine. Probably the worst that can happen is the Meanwell PSU will shut off if it detects an over-current, but I'd be surprised if the computer needed that much power. I'm guessing that the schematic means "1.4 A optimal, 2.2 A max rating". But, it could also mean "1.4 A constant, 2.2 A peak/surge". It's a bit ambiguous. You can probably assume the computer won't draw over 1.4 A.

In terms of fixing, as Matthew H said, as long as the transformer is fine, you can fix these boards with some basic electronics knowledge (and I will add: mains voltage safety knowledge). I'd happily fix this board. It's pretty easy to rig up the Commodore PC-1 PSU to some test loads (can be found on eBay), like so:

Testing Commodore PC-1 PSU on the bench

I made a little adapter board just for fun, but you don't have to.

Here's what voltages and currents I got with a working Commodore PC-1 PSU:

Test loads attached to Commodore PC-1 PSU

The bottom left test load is actually the -12 V output. At max load, the +5 V output drops to 4 V. I imagine that's probably within tolerance. I didn't check the computer main board, but I'd guess it's regulated down to 3.3 V. At 500 mA, you should see 5 V, but then it'll decrease as you add more load. All working Amiga 1200 PSUs seem to do this too, so I think it's just a design quirk with Commodore PSUs.

I noticed that you need to attach a load to the +5 V output, otherwise, you only get 9.5 V on the +12 V rail. With no load on either +12 V or +5 V, I'm seeing -9 V on the -12 V output.

4
  • Sorry, but the device is from 1987 - I can assure you it does not regulate 5V down to 3.3V as there are no 3.3V components in it, and most 5V components are not specified to work at 4V so it will be out of tolerance.
    – Justme
    Aug 14, 2023 at 18:56
  • Fair point. I'll remove that. In any case, it seems that all of my Commodore PSUs tend to have this behavior; fine at low current, voltage drops at max current... yet they all seem to work fine with the computers. I guess the computers never draw anywhere near the max current for the PSU. Aug 15, 2023 at 21:32
  • 1
    To be fair, unless you measure how the computer draws current from each supply voltage, you can't really emulate it with dummy loads. The supply will likely not work well if loaded in an unexpected way, such as drawing max rated current on all outputs simultaneously where the power rating may be exceeded, or by e.g. having almost no load on some outputs so the other outputs are out of regulation as well. They work best when all outputs have some range of typical min/max load that may depend on each other, such as floppy motor turning on uses 12V and the 5V R/W logic also consumes more current.
    – Justme
    Aug 15, 2023 at 21:45
  • Cool, in the future I'll measure what current the computer actually draws from a good working PSU, rather than maxing out every output at the same time. Aug 15, 2023 at 21:53

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .