[megaimg]

Trying to fix a R4000 PSU....So far new capacitors on both board....on the small board, replace the BYV32E,LM337T,TIP30C and MJE350....On the large board, replace the STR81145.....so far no luck....

Has anyone have fix this PSU....any other part that I need to check....

[weblacky]

I've not gotten to these model yet...but the general advice I give everyone rebuilding SGI PSUs is...many didn't have MUCH in the way of "smarts" They had circuits to stick in constant restart and such to prevent OL. So without additional smarts...if your PSU "failed" on you...a recap will not be all that is required...something actually failed...like a semiconductor or a passive resistor (as well).

My personal experience with OTHER PSUs has been the secondary diodes and various other power-handling FETs will blow once caps go so far out of tolerance that the ripple (or whatever) blows the brains out of semiconductors.

Outside of a basic resistance check...assume you lost a semiconductor.

The EASIEST way (without advanced equipment like I cheat with). Is to do a resistance check across the DC output lines (when PSU power off, unplugged). Then reverse the probes from DC and ground lines and see if you get different reading (you should). If you get the same reading...you have a blown secondary diode(s).

After that, take suspected output wires and use a current-limited bench-top PSU and shove correct (by very low setting, like 1A current ) voltage with the correct polarity to the output wires into a powered-off PSUs via it's DC rails (5v=5v, 3.3v=3.3v, etc) and see if they are in short (that this is taking real power...should only take like milliamsp...not amps. Do not do this for long or you'll chance burning out your secondary transformer coils (like 2-3 seconds only) if you're missing your diodes. That's a VERY quick check for shorts on a secondary. For your primary...just measure the main filter cap leads when on for like 300v or such. If not, then you have a primary problem. A shorted main FET SHOULD cause the primary filter cap to have extremely low voltage...but not a guarantee.

Most of the time (I've only found bad resistors in the power-path twice) the was nearly ALWAYS a bad FET or DIODE body. So go with the assumption that's what you've lost.

[megaimg]

I think I found the issue....waiting for parts... will update soon!