Can switching supplies be paralleled?

CounterRotatingProps

Over the last year or so, we've received a bunch of donated PCs.·Most weren't more·than·scrap.

So now I've got a pile of·switching power supplies that I'd like to use for bench power.·
I've read that·they aren't as reliable for this as a linear supply.

But·was wondering if this·could be fixed -·would it be OK to:

1.·stack their outputs in parallel to get the amps up, and

2. run these outs·through filters and a couple of power transistors to clean it up?

please advise.

Thanks,
Howard


▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

Philldapill

Howard, it's not a good idea. What will end up happening, is one SMPS will be doing all the work - unless you can PRECISELY tune the volatge reference for each so they are the same. In a way, it's a little like paralleling diodes, but not at all. When you parallel diodes, one diode will have just a SLIGHTLY higher voltage drop, causing the lower voltage drop diode to pass the majority of the current. Same thing with paralleling SMPS. One SMPS will be fairly content at, say 4.97V, while the other one will want to push the voltage up to say 4.98V. However, if you loading them fairly heavily, BOTH will be well below their "content" voltage, and both will be working together.

CounterRotatingProps

Ah, there's always a wrench in the works when you try to do something non-standard [noparse]:)[/noparse])

Is that voltage reference easy to cut into... in other words, could I hack a feedback loop into them for·common V. references?

> loading them fairly heavily, BOTH will be well below their "content" voltage, and both will be working together.

Interesting --- could I 'trick' them into thinking they were being loaded?

I've got like 20 of these things, so I don't mind toasting a few ...

thanks for the reply!
- Howard

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

LoopyByteloose

Filtering is possible, but may not be required. The best place to start is to provide load in the form of some 20watt resistors and take a look at what the output is doing on a scope. If there is good performance over a couple of days, just dust out the interior with a large soft brush and use it.

These days when I go shopping at the local electronics store, I see very few power supplies over 3amp in new stock. The older switchers are sitting on the shelf and they go up to 20amps, but nobody seems to be buying them. So I really wonder if you will need much more than a 300watt switcher for general bench use. You may occasionally need a 3amp wall wart switcher for 3.3v or 5volt, but you can now get those cheaply.

I have a 300watt ATX stablized with two 10watt resistors - one is on the 5volt and one is on the 12volt. The values are 15 ohms and 68 ohms. That seems to work for me. I just have to connect the green wire to ground to get it to run.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Ain't gadetry a wonderful thing?

aka G. Herzog [noparse][[/noparse] 黃鶴 ] in Taiwan

Post Edited (Loopy Byteloose) : 6/26/2009 3:40:38 PM GMT

CounterRotatingProps

Thanks Loopy!

Many years ago, someone gave me a switcher with big wattage resistors strapped on - I used it for a bench supply but never questioned why those resistors were there --- now I know, thanks.

Here's a real power-supply N00B question: since these are 'good enough' for PC's, would they be reliable enough for MC protoboards, for example - if properly loaded as you describe?

> If there is good performance over a couple of days ...

wow, how interesting ... can they really drift that slowly?

- Howard

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

LoopyByteloose

Testing for a couple of days provides you with some idea if there is an overheating condition or faulty caps. Sometimes there is an accumulation of black dust [noparse][[/noparse]really carbon] inside the supply that will cause intermittent shorts.

One can use 6V and 12 light bulbs as well.

My 300watt runs 5v at about 4.86v and 12v at about 12.78v. The 12.78 is a bit problematic, so I generally avoid using it. There is some latitude -- but if the voltage remains stable, one can live with a certain amount of overvoltage or undervoltage. +/- 5% is generally considered okay.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Ain't gadetry a wonderful thing?

aka G. Herzog [noparse][[/noparse] 黃鶴 ] in Taiwan