Testing a Dead PSU with JUMPER WIRE instead of FUSE -- Bad Idea?

Thread Starter

rosario99

Joined Apr 7, 2017
2
Just yesterday my Silverstone SST-ST1000 PSU stopped working. Everything was fine until my home experienced 3 sudden power outages (On-Off // On-Off // On-Off) within 20 secs. When the power came back on and remained on the PSU did not come on.

I am not an electronics expert whatsoever but still want to be sure the PSU is dead for good. I figured the internal fuse was to blame so i opened it up and finally located the fuse. 15a/250v it claims. (see photo below). I performed a continuity test = failed AND a resistance reading = 0L. Great it seems the fuse is to blame but i have no replacement. And im still wondering if the rest of the PSU is ok.

I decided to strip off some 12/2 wire (rated for 20a) and stick it where the fuse used to be. The logic being this should be fine since im sure there wont be some surge coursing thru my house at the moment of turning it on. I also jumpered ATX / MB pins 16/17 (the green and black) and added the PSU power cord. At this moment the PSU is still switched to OFF. No load attached.

With the above all connected the plan was to turn the PSU to the ON position to at least see a green led light (the LED is right next to the PSU power switch) indicating power is there but i didnt expect the PSU to kick in and turn on its fan since there is no load attached yet. I went to turn ON the PSU and i got a SPARK and POP. I quickly switched it OFF.

What did i do wrong? Was the temporary jumper wire a bad idea? Does this test prove there was something further down the line that was damaged during the original power outage OR did i unwittingly damage a likely working PSU since i didnt replace the dead fuse?

Thanks for explaining.

 

AnalogKid

Joined Aug 1, 2013
8,938
Something partially failed, or failed in a way or in a circuit such that the fuse popped before the rest of the failure could occur. When you applied power without current limiting, the rest of the initial failure cascaded. Now there are several dead components and more that were stressed by the two events. Time for a new supply.

ak
 

Hypatia's Protege

Joined Mar 1, 2015
3,226
What did i do wrong?
In defeating the overcurrent protection, you've neatly 'escaped' the damage control safeguard provided thereby:rolleyes:

Note that it's barely possible that the initial failure owed to a 'shorted' OVP (secondary to 'line spikes') - in which case your unadvised 'test procedure' mightn't have produced further damage -- It may be worth your while to examine the circuit for damaged varistor and/or 'gas fuse' OVP devices ...

Best regards and good luck!
HP:)

PS -- Please don't feel too bad! If the initial damage was more 'complicated' than a shorted varistor, etc - the PSU was likely already beyond practical repair (sans service literature)...
 
Last edited:

Thread Starter

rosario99

Joined Apr 7, 2017
2
PS -- Please don't feel too bad! If the initial damage was more 'complicated' than a shorted varistor, etc - the PSU was likely already beyond practical repair (sans service literature)...

Thanks. Yeah i was planning on throwing away the PSU anyway so figured what the heck. Maybe i can get lucky or learn something new. ;P
 

dl324

Joined Mar 30, 2015
12,450
Yeah i was planning on throwing away the PSU anyway so figured what the heck. Maybe i can get lucky or learn something new.
The steps I might have taken:
  1. Inspect the unit for any signs of obvious component failure, replacing any found.
  2. Replace fuse and apply power.
  3. If fuse blows again, get a schematic and start doing unpowered checks as much as possible, e.g. power semiconductor junctions.
  4. If nothing found, buy a replacement.
ATX switched mode power supplies don't have transformers which means line voltage and rectified line voltage are present when the unit is on. I had a computer that became flakey and traced it to the power supply. Once I verified the fix, I just put the supply aside in case I might have use for it's parts in the future.

I have 40 years of experience in electronics and it wasn't worth my time to attempt to repair. The prospects of thinking I repaired it and have it create another unreliable computer wasn't worth the bother.
 
Top