# Power supply circuit!!

Discussion in 'Homebrew and Kit Projects' started by 2E0GGO, Sep 18, 2008.

Not open for further replies.
1. ### 2E0GGOHam MemberQRZ Page

Hi there guys.
I wonder if anyone can provide a circuit diagram for a 12 volt 60 amp power supply that I can build.
I am trying to build a supply from scratch.
Any help will be an advantage.
I am also looking at swapping a few items for a 24 volt 50 to 60 amp transformer.
Thanks a lot
73's
2E0GGO
Gareth.....

2. ### VK2TILHam MemberQRZ Page

I built the 1986 ARRL Handbook design many years ago; I re-wound a large old TV transformer and the capacity is probably 40 amps.

Here is its interior;

If you can get the article (I can send it if you PM me) it has good design information including advice on heat dissipation.

Your "24 volt" transformer probably produces 24 volts RMS; depending on your rectifier design it will probably produce over 30 volts DC.

So, to produce 13.8 volts DC, at least 16 volts will have to "disappear"; it will be dissipated as heat.

16 volts at, say, 20 amps is 320 watts; it will be proportionally-greater at higher currents.

This is a lot of heat and you will require lots of pass-transistors & substantial heatsinking.

Note that this is a crude analysis; it's only intended as a guide.

If possible, unwind some of the 24-volt turns from your transformer; there's a good chance that it's three turns/volt which seems to be common in large transformers.

Please don't let this advice deter you; it's intended to guide you in a very interesting & useful project.

3. ### KC4YLVHam MemberQRZ Page

Or you can get a 24v CT transformer and probably eke +/- 12V out of it though I would personally go for a 30V transformer in that case, not enough headroom if you want +/- 13.8V.

4. ### M0DSZHam MemberQRZ Page

As regulators, the cheaper, linear PSUs all seem to use the ubiquitous 2N3055 in parallel in quantities according to the current output. A generous value of electrolytic smoothing capacitor is in order but at 60A you may need several in parallel to overcome the individual ripple current limitations. This is again exactly what the commercial PSU manufacturers do.

5. ### VK2TILHam MemberQRZ Page

Just being pedantic; the required quantity of 2N3055s (or other pass transistor) depends-upon the power being dissipated, not the current.

At 60A our OP's proposed supply may well require the dissipation of a kilowatt or more; that is the heat output of a small bar-radiator and will require lots of 2N3055s and a monster heatsink.

The Handbook article that I referenced earlier provides a very lucid explanation of the dissipation problem.

In practice; my homebrew Handbook design uses six 2N3055s on a substantial heatsink;

I have never drawn more than about 25A from it (I can only talk on one radio at a time) and the heatsink gets warm but not hot.

Guided by the information in the Handbook article I rewound the transformer to the minimum required voltage (about 16V I think, I don't have a record);

The winding was trifilar using, I think, 1.5mm wire. I think it was good for 45A or more but I left the over-current circuit as in the Handbook design which means that it trips at about 35A.

There are four capacitors; each is a little larger than a beercan (an Australian unit of precision measurement). I think each was 190 000uF.

As DSZ said, this large value was necessary to "smooth" the rectified DC; if a larger transformer secondary voltage is used, less capacitance is required to prevent "sag" in the final output.

But the trade-off for higher "raw" DC voltage is heat dissipation in the pass transistors.

There is a "happy medium" which the designer must determine.