Bipolar PWM homebrew power supply

[Madscientist267]

Anybody either have in their possession or know where a schematic for a power supply that outs +/-32VDC at about 2A peak (~1A sustained) from a single 12V input? Tight regulation isn't exactly necessary, but would be a plus. Complexity and cost are more important factors than regulation or raw efficiency. Anything drawing less than 10 amps with the outputs both pouring 1A at the same time would suffice.

I've been looking all over the place, and nothing. I'm considering tearing hard into the average power inverter, and making it do what I need, but this almost seems a bit extreme for what I need it to do.

Any ideas would be greatly appreciated, thanks ahead of time...

Steve

[Nothing40]

Find an old car stereo amplifier (Jerry might even have something you could use?).

Most of them step up the voltage to somewhere around +/-30-35V for the audio amplifier circuitry.The only thing you may need to do is to rewind the transformer,if the voltage is too far-off for your needs,and/or twiddle the voltage regulation feedback resistor(s).

I have actually done this..It wasn't terribly hard,if you're familiar with electronics a bit,and can solder.It worked like a champ also,The SMPS from the amp I used could supply about 400W continuously.

[dnix71]

Here's one in a box. http://www.zahninc.com/specD.html

[sPuDd]

Steve,

   below is a link to a very old but reliable 100W Car Amp design.

It is designed to give +-27V rails, but you can easily get +-35V

by using more turns on the secondary, higher voltage output caps etc.

100W Car Amp PDF

sPuDd..

[mikeyduk]

Roll your own .

http://sound.westhost.com/project89.htm

Regards

       Mike

[Madscientist267]

I ended up building one, and it works great, but it seems a little hungry.

It uses a center tapped toroid and 2 sets of 2 IRFZ44N MOSFETs in push-pull, driven by a TL494 chip. I don't have a schematic just yet, as I basically modified a cell phone charger to use as a controller.

Test conditions are ~12.7V input, output is +/- 32.4V. MOSFET temp held to a nice and cool ~75 deg F (AMD Athlon 64 heatsink with the fan running full speed from a separate supply).

One concern seems to be that it draws ~350mA with only a ~50mA load, and at ~120W (4x60W light bulbs in parallel, rail to rail), it was blowing 15A fuses, and even stressing and then eventually blowing 20A.

I'm thinking the inefficiency is related to a resonance issue, as the toroid gets rather warm at 120W after 10 minutes or so, and so does the input side's downstream filter electrolytic. I didn't use small value stiffeners on it, so I figure the heat in the cap is just from transients making their way back to the input caps (via the freewheel diodes in the MOSFETs?)

The heat in the toroid though is a bit troubling, but I think it wants to ring naturally somewhere around 25KHz, and the drive output is at about 4KHz. I haven't thrown it on the scope to verify this, but these are the 2 numbers that are coming from my meter. One thing is for certain, the toroid can certainly handle much more power than this (assuming everything is running in tune); it is one of two that came from a 1200W competition car stereo amplifier.

Anybody know how much of an issue this actually is, and whether it is worth trying to resolve? It takes playing with SMT components, and my hand isn't that steady anymore. So far these are the only problems I have had with the build; regulation is reasonably tight, approximately within 0.1% from 0-120W. I'm sure it could hang with much higher power levels than what I am feeding it, but I don't need it to at the moment. I have the taste in my mouth now though, and will need to go to the next level for my next project...

Pics as soon as I process 'em for the board.

Steve