SCOTT JONES wrote in a message to ROY J. TELLASON:

-=> ROY J. TELLASON wrote to JIM HOLSONBACK <=-

 RJT> I know it'd probably be a good thing to be moving up to some newer
 RJT> stuff,  but from what I hear the newer stuff isn't always that much
 RJT> better,  or that much more reliable,  as it turns out.  I keep hearing
 RJT> about problems with ATX power supplies seemingly blowing up so much
 RJT> easier (and sooner!) than the older AT supplies,  and there's that cap

 SJ> As always, YMMV, but the only times I've seen any problems with ATX
 SJ> power supplies were in hostile environments that probably have
 SJ> killed an AT supply just as easily.  Meanwhile, this machine is on
 SJ> its third AT power supply in two years.

I guess I'll find out when I get there.  

 RJT> problem you mention,  I remember reading about it here and elsewhere,
 RJT> but don't recall particular time frames with regard to that.  I'm going
 RJT> to have to research it a bit and see what I can find out,  as the
 RJT> effects of that one are pretty far-reaching as I understand it.  And it
 RJT> seems to me at this point that if I can get a rough idea of what time
 RJT> frame that problem happened in I'd be better off staying away from it
 RJT> altogether...

 SJ> IEEE had an article on the problem back in February.  I found it
 SJ> archived on their web site a couple of weeks ago when I had to fix
 SJ> the year-old GigaByte mb in one of my wife's machines that had two
 SJ> bad caps. That was cheap to fix.  A couple of bucks for one cap and
 SJ> another stripped off a board I had sitting around.

I did run across some info once,  I could probably find it again.

 SJ> OTOH, my MIL's husband gave me a bunch of parts to help 
 SJ> troubleshoot the wife's board, basically a whole computer, 
 SJ> including a year-old Soyo mb. The Soyo has *11* bad caps. It'll be
 SJ> a whole lot easier and cheaper to buy a new board than to replace
 SJ> those, unless anyone knows where I can get 2200uF 10V high-temp
 SJ> caps for cheap.

On looking closer at some of this stuff I'm amazed at how small some of
them are getting for the values that are in there.  But then,  that's been
happening for as long as I've been into this stuff,  several decades now.

How are you testing those caps?  I see lots of references to ESR meters, 
but that's not something I'm likely to want to invest in anytime soon.

---