As promised some info on my home made power supply.
As I said it's based on Elektor's lab PSU design from 1982, there's nothing special or fancy about it at all. I chose this design because I had the ideal transformer for it to hand. My former employer had a few of those ancient monster C core trannies that they sold for a really good price, it's got 2x29V/6A and 4x11V/1,6A secondaries which was perfect for the circuit published by Elektor in 1982.
The case is also home made as I could not find a reasonably priced case that suited my needs, this was supposed to be a low cost project and I didn't want to spend over 100 Euros only for the case. So I hit upon the idea to use the heatsinks as sides, that has been done before so why not? I had to buy them anyway as worst case dissipation is above 150W per supply if I recall correctly. They were the most expensive part of the whole design. I'm using 4x 2N3055 per supply, so this thing makes for a pretty good heater as well
- no switching of secondaries here and it is capable of 5A @ 1V over long periods, I have tried it.
Initially I built it using four analog meters, but as you all know good ones cost a pretty penny so I skimped and bought the cheapest chinese crap I could find - they claimed 2,5% full scale accuracy but turned out to be more like 3-6% all ofer the scale.
Later I converted it to digital displays using the good old 7106. That turned out to not be as easy as I imagined - there is a separate transformer and four DC/DC converters in there just to get the displays going
. Luckily the DC/DC converters were part of a load of huge boxes of electronic components I got for free.
About two years ago I wanted to power a car headlight (H4, 55W) and while doing that the one supply died. Magic smoke came out and that was it.
While investigating what had happened I found that it oscillated violently under heavy loads and some resistors in the original design were not specced powerful enough. I don't know if the original design using the proper PCB and cleaner wiring would behave the same but my tests lead me to think it's possible.
I ripped it apart, replaced the burnt out resistors, a died 2N3055 and placed small caps in the feedback loops of the two opamps to get rid of the oscillations. After some trial and error that worked and no matter what kind of load I hooked up it was now totally stable.
I also replaced my original hand wound current sense "resistors" (really just a coiled piece of copper wire) with precision resistors which cured some annoying inaccuracy with the amps meters
.
That was when I took the pics attached below.
Since then I also had to replace the low quality coarse voltage pots with high grade sealed units as the originals got a bit scratchy and this supply has the annoying habit to go full voltage when the wiper looses contact - not well designed!
All in all it works pretty well and apart from that one time it never played up. I've abused quite a bit - charging the car battery in a snow storm at -15°C with 2x5A and such things...
Total cost was an estimated 100 to 120 EUR.