Increasing PSU output capacity?

[GnatGoSplat]

I have an old PSU I was thinking about using. I don't know its actual maximum output specs, it is just a PCB with no marked specs, but I could probably calculate a theoretical max based on values of the output rectifier diodes. I know in its original purpose, the +12V rail only saw around 550mA and 1A peak.

I only have a very rudimentary understanding of how SMPS work, so this may be a rather silly thing to ask: Would upgrading the input rectifier, switching transistor, and output rectifiers allow the PSU to supply more than its rated current before blowing up? I haven't checked my desired load with an ammeter, but I'd guess maybe 1.25A average and 3A peak.

[mariushm]

Re: Increasing PSU output capacity?

The output diodes are just a small part, the transformer is also important, as is the mains switcher, the mains capacitor..

Generally, old psus (p3, p4) had about 8-10A on 12v, about 20-30A on 3.3v and 5v. So if your peak is 3A on 12v. Somewhat closer to the current times, the psus with 4pin cpu power connectors usually have about 10A or more on 12v...

Why don't you actually check the label on the power supply? I have near me an old 100w power supply from the times of Pentium 90 Mhz and even that one has 13A on 5v and 4.5A on 12v.

[GnatGoSplat]

Re: Increasing PSU output capacity?

It's actually a PSU in an old Mac Classic so it has no labels and no published specs. I would like to run additional peripherals using its existing PSU if possible.
Transformer is unusually large and uses large gauge wiring, but I've read the higher the switching frequency, the smaller the transformer can be. So it could be this PSU switches at a lower frequency and therefore the size reflects its input frequency and not its power output capacity?

[999999999]

Re: Increasing PSU output capacity?

It might help a lot of we had a good top-down picture of the PCB, and the part # on it. Ancient PSU using an unusually large transformer and low switching frequency tended to be steel laminations core type instead of solid ferrite core.

Without more info we can't know what the weakest link is, but generally speaking it's not usually worth the bother to try to wring much more power out of a PSU, particularly a pretty old one.

Calculating a theoretical max from output diodes isn't going to work well most of the time, those diodes are generally rated for significantly more than the current they actually handle except in very low end junk grade PSU. In low powered OEM systems the same is usually true for the input rectifiers, most often it is the transformer itself that's the limit BUT years ago when a PSU was rated for a particular wattage, odds were better it could run that wattage 24/7, more margin was built in.

What you could do is replace most of the capacitors as it probably needs that done anyway then put the desired load on it and see if anything blows up and if so, replace the failed component.

It is odd you didn't mention what the desired load is and what voltage it needs. If it needs 5V, certainly the PSU should be capable of 3A. If it needs 12V, odds are fair it is regulating based on a measurement of 5V rail voltage level so you might need to place a load on the 5V rail to keep the 12V rail from drooping in voltage, possibly even causing the PSU to shut off.

Otherwise, just about any PSU meant capable of powering drives from 12V output can manage 1.5A and around 3A peak at initial power on to spin up a drive, given enough of a load on the 5V rail as mentioned above.

[Logistics]

Re: Increasing PSU output capacity?

You know, your first priority is just a full recap of every electrolytic in the supply. It has to be at least 21 years old now which means all the caps are elderly, now. If you give it some new 105's, including any BIG mains caps it may have, it will probably do just fine with whatever you're trying to do.

IIRC, it's been stated on the boards before to avoid recapping something like a PSU with lots of low-impedance caps as it changes the way the circuits react and might cause problems for the other hard-parts in the supply. Just use general purpose or perhaps high-ripple caps, but not low impedance.

[goodpsusearch]

Re: Increasing PSU output capacity?

Never use general purpose caps for psu recap. You should use entry level low esr capacitors, for example, Panasonic FC, FK, Rubycon YXF, Nichicon HE, Chemicon KY. Panasonic FM / FR may be good too, if the original caps have close esr and ripple.

Rubycon MBZ/MCZ, Nichicon HM/HN/HZ and Panasonic FJ/FL are motherboard grade capacitors, not suitable for power supplies.

[999999999]

Re: Increasing PSU output capacity?

It hardly seems worthwhile to recap this ancient low efficiency PSU. What additional peripherals do you need to power that there isn't sufficient reserve on the system PSU to handle? If it has that little reserve power it should probably be upgraded instead.

[Logistics]

Re: Increasing PSU output capacity?

I thought it was the nature of this board to fix, improve or otherwise restore old or broken hardware. I also share this sicknes... err, this love for old things.

@GnatGoSplat: If you're restoring this old Mac then thumbs-up! Can't wait to see the outcome.

[GnatGoSplat]

Re: Increasing PSU output capacity?

I did try to restore the old Mac, but the motherboard is shot. The PRAM battery had leaked out and eaten away a bunch of vias and traces. I repaired the traces and vias which made it boot up, but it has weird errors, so something is not quite right.

Anyway, I had thoughts of mounting a low power Mini-ITX motherboard in it and running that off the original PSU in an effort to recycle as much original hardware as possible and also a cleaner internal appearance with everything mounted in stock locations. Even if I could reuse the PSU, I'd still need a 5V to 3.3VDC regulator and a 5VSB circuit to make it ATX compliant, not to mention the ATX connector, plus recapping, would end up costing more than buying a MicroATX PSU and rigging some way to fit that in there.

It does for sure need recapping, they are leaking a little and stink when I heat their leads with the soldering iron. They also leave a greasy spot on the PCB when I removed them. Somehow, capacitance and ESR still test good, however.

I'm probably going to just gut it, but for some reason I kind of liked the idea of making the original PSU work.

[999999999]

Re: Increasing PSU output capacity?

I thought it was the nature of this board to fix, improve or otherwise restore old or broken hardware. I also share this sicknes... err, this love for old things.

@GnatGoSplat: If you're restoring this old Mac then thumbs-up! Can't wait to see the outcome.

There is also value in knowing what is worth repair and what is not (in my opinion, though technical details of the PSU and pictures are not present). Some things fail because they were junk, is making junk live forever a good thing or an exercise in masochism? I could get a Ford Pinto from a junkyard and fully restore it but would I want to drive it even if it were free? No. It may be a stretch to compare that to a PSU, but when you factor the time, parts cost, and how little a low wattage new surplus PSU costs that is smaller, lighter, more energy efficient... It's just one opinion.

[Logistics]

Re: Increasing PSU output capacity?

If we were going by what is cost and time efficient then half the stuff on this board wouldn't get fixed. Honestly, a large portion of what is done on this board is novelty or special interest which throws concerns of time and money out the window.

[999999999]

Re: Increasing PSU output capacity?

^ True, I suppose I should state that projects aren't worth it more often than I do.

What do you think though, is it reasonable to pick what is suggested to be a 550mA, 1A peak PSU and modify it for 1A, 3A peak, or better to pick a PSU rated for _more_ than the expected load? PSU aren't exactly hard to come by in first world countries.