Simple regulator sanity check?


2009-03-27 12:31 am
I've put together this schematic of a power supply with three voltages, +/-12v, and +5v. I don't need more than 1amp from any regulator. Does this circuit look correct, or am I missing something? Other than a fuse, I just feel that something's wrong here. Just not sure about running the 5v regulator like that.

Thanks for looking!


  • power.png
    36.5 KB · Views: 118
The 4.7μF caps are somewhat superfluous IMO. The 7805 could be paralleled with the 7812 input, with adequate heat sinking of course.

Agree completely. I'd also recommend a pass transistor for the 7805 since it will have to dissipate more due the the higher input voltage, and I'd rather dissipate that heat with a transistor.

IMO I'd recommend using a pass transistor for all 3, since I usually don't run more than 500mA in those kind of regulators to keep them well within SOA despite temperature.
I tend to disagree with the use of pass transistors if the supply isn't required to deliver more than 1 amp at any of the outputs. Any of those regulator chips can easily provide greater than 1 amp with adequate heat sinking.
At the stated power levels, the heat sinks needed don't need to be humongous. The 12V supplies are somewhere around 5W dissipation. The 5V supply is dependent on where it's connected.
Using the pass transistor for the 7812 will allow the 7805 to remain as shown the first configuration. Then you could just connect the 2 regs (7912 & 7805) and pass transistor to identical heat sinks that will each dissipate about 7 watts. Those aren't big sinks, and can probably be had for a dollar or two apiece.

edit1: oops, the 7912 might have a lower current rating, which does change what I just wrote.
Rats, I hate when that happens.
edit2: no, the MC7912 is spec'd for 1A like its 7812 complement. I was looking at an old LM7912 datasheet. I hate when that happens, too.
Last edited: