MIKE ROSS wrote in a message to Roy J. Tellason:

 MR> "Roy J. Tellason" bravely wrote to "all" (25
Oct 03  04:06:11)  ---
 MR> on the heady topic of "power supply"

 RJT> Got a dead ATX power supply here...

 RJT> Anyhow,  fuse is good.  Thermistor is good (about 8 ohms cold),
 RJT> bridge rectifier is good,  caps are good as far as I can tell,  main
 RJT> switchmode transistors are neither shorted or open,  and I have *no*
 RJT> idea about the rest of it.
 RJT> Repair FAQ says "startup"...

 RJT> anybody have other suggestions to offer on this?  Besides tossing 
 RJT> it?  :-)

 MR> I only try to fix them for myself when I don't have another handy.

That's the situation I'm in now.  And even when I do get another one,  it
wouldn't be a bad thing to have a backup.

 MR> The main switcher circuit on psu's have pretty much settled down to
 MR> a standard form.

 MR> Here IMO are the 2 most common failures:

 MR> If you can follow the base drive of the switchmode Q's you will
 MR> find a large value resistor (between 220K to 470K) to the main
 MR> rectified supply. One of these tends to open up for no reason. I
 MR> think it may be due to AC line spikes.

Ok,  that sounds do-able.

 MR> If these resistors are good then keep following the switchmode Q's
 MR> base drive and you will find a diode, 22ohm resistor, and a diode
 MR> across a 2.2uF electro (maybe 1uF to 4.7uF). These tend to dry up
 MR> since they are close to a heat source and also must conduct a
 MR> substantial current.

Ok.

 MR> The TL494 is hard to kill.

What is that chip,  anyway?  Something application-specific?

 MR> The switchmode transistor can lose gain but that would show up as 
 MR> a lower power output. Good luck.

I'll fiddle with it when the mood strikes me,  and maybe I'll get lucky, 
you never can tell...

---