> http://www.microsyl.com/projects/powersupply/powersupplysch.pdf

Ah, pardon my rudeness. I should've posted that link instead of making
a fella hunt for it. If anyone wants a go at one they can use my PCB
layout. Apart from the ZXCT1009 (Mouser) most semis would be
substitutable

http://home.clear.net.nz/pages/joecolquitt/sylvain_psu.html

> Worst case the whole dissipation will be in the 3055's. 25 V x 5A =3D
> 125W. I'd expect them to handle that with enough heat sinking

I'd expect so too, if they survive the initial short or heavy load. The
whole case is pretty much a heatsink, heavy gauge aluminium. I
acquired it somewhere, had a blown uA723 controller. If I'd made
it I'd have spread the transistors out a bit, as there actually is room
over the transformer down that end. Not run it in anger yet but the
load test showed it stays reasonably cool. Can add forced air if
necessary, plenty of little fans lying arond
=20
> If you mean "short is gradually removed by a polyfuse" then yes -
> I'd expect them to survive that well

I'll have to look into polyfuses and the mechanisms of transistor
demise

> BUT you can set thermal limiting on HS temperature easily

One mod I will probably make is to add a hefty relay for quick
disconnect if power needs to be removed from the load in hurry
or if T goes too high. The LCD PIC can have that as another job

My intuition is that a relay will be far too slow to offer protection
against transients though

It seems to me that you need a fast semi to protect a slower one

For example, if collector current was through a normally-on FET,
is there a technique to detect over-I and switch that FET off before
the over-I takes out the 3055s ? Or a big fat SCR to crowbar some
I to 0V ?
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.