Overloaded Horrible Antec SmartPower SP-500 to death

[kw71]

I don't think I need to introduce this power supply, as it has been featured in this forum before. I'd like to share what happened to mine. I am new to electronics and maybe someone will help me understand how this works.

The PSU was bought new and ran for years with its bad caps. Then, I decided to try overloading it. I added 20-25 amperes of 12 volt load. This ran okay for several months and the PSU began to shutdown under such load.

I looked inside, found the Fuhjyyu that everyone loves (except at the 12v filter, it was a different brand that nobody here likes, but was the only large cap that tested near rating on little Fluke meter.) They looked and tested as you expect, all electrolytics were replaced. At this time I noticed the burning around the optoisolator, I guess this is the 5vsb area.

I also dropped a dab of solder over the thermistor that controls the fans. Whee! The PSU ran again, no more hissing, even with the horrible overload. I ran it again for some more months. It stopped working with the heavy load, so I took the horrible load off, and then several months later, it died with a normal load. I reduced the load yet again and it ran for several hours and died again. Now it will not support a useful load.

Opening the PSU this time I noticed a new burn mark around the 12v output. Seems a jumper has changed color through some kind of metallurgic process unknown to me. The capacitor and inductor are OK.

I took the heatsinks out and as far as I can test with my little fluke meter I don't see anything wrong with the big devices there. The big diodes test all the same. The big bipolar transistors test as has been described they should on the diode tester, and I can use the ohmmeter to charge the gate of the FET and see it turns on. To me this proves they are acting reasonably like any device of their type and not that they will perform well. In other words, I believe that they are transistors.

When I first saw the burn mark by the optoisolator I figured it was a symptom of the bad caps -- these 5vsb filters have gotten attention here before -- and that it would not burn anymore having rectified "the problem." I didn't make any photographs at that time, so I'm not certain that the burn hasn't grown. Now, here is where I know I am over my head. I think I want to find out why this burned. Seems the neighbors of this opto are the transformer, surfacemount transistor and zener and a few passives around. But it seems the closest part to the evidence of the heat is the opto. Why would this get hot? Duty cycle too long, or too much current? Transformer is ruined? Looks like this could end up driving the FET.

Is there a schematic floating around of this or another Channel Well supply so that I might try to understand better? I'm thinking that I want to reassemble it and observe it on the bench with some load I can control, and maybe I will get a clue. But I also am not sure if I should remove any more components to test them out of circuit.

Thanks for reading. Hope you enjoyed this story, and I'm grateful for comments.

[KeriJane]

Hello.
Welcome to the forum!

"You did WHAT?"

You added 240+ watts of load to 12v only and ran this thing for months like that? On top of the usual loads? Of what? A Prescott?
Yes, things are going to get warm and probably change color.



I'm impressed. Sounds like an SP-500 isn't so terrible after all.

How did you dissipate that much extra heat?

Have Fun!

[dmill89]

Originally Posted by KeriJane

Hello.
Welcome to the forum!

"You did WHAT?"

You added 240+ watts of load to 12v only and ran this thing for months like that? On top of the usual loads? Of what? A Prescott?
Yes, things are going to get warm and probably change color.



I'm impressed. Sounds like an SP-500 isn't so terrible after all.

How did you dissipate that much extra heat?

Have Fun!

The Antec Smart Power and True Power series really aren't bad power supplies, they just used garbage (fuhjiyuu) caps which hate heat and a "silent" design which slows the speed of the internal fan and stops the external fan under most loads leading to almost guaranteed cap failure. I've recapped several Smart Power/True Power PSUs (Mostly with UCC KY and KZE on secondary, Panasonics on primary) and have yet to see a single failure of these recapped PSUs. One of them, a SP350 has been running for over 5 years since the recap and still works fine(all rails in spec, no failed caps).

Originally Posted by kw71

When I first saw the burn mark by the optoisolator I figured it was a symptom of the bad caps -- these 5vsb filters have gotten attention here before -- and that it would not burn anymore having rectified "the problem." I didn't make any photographs at that time, so I'm not certain that the burn hasn't grown. Now, here is where I know I am over my head. I think I want to find out why this burned. Seems the neighbors of this opto are the transformer, surfacemount transistor and zener and a few passives around. But it seems the closest part to the evidence of the heat is the opto. Why would this get hot? Duty cycle too long, or too much current? Transformer is ruined? Looks like this could end up driving the FET.

Honestly I'm not sure I would even try to repair it after the abuse it has suffered. Running it with an overload and excessively hot for an extended period of time (months in this case) has the potential to damage or at least weaken just about every component in the PSU. To truly repair it to its previous condition could easily exceed the value of the PSU(remember you can get a good 500w PSU for around $50 these days) and I would be hesitant to trust any system to a PSU that I knew had been damaged in that manner. If you want to repair it have fun it will likely be a challenge, however it may be best to use this PSU for parts, but that's up to you.

[nickt1862]

Yes I know this thread is months old.

I have an SP500 which still "runs" with one of the two main filtering caps (I think it's a 200v - hard to read it) a little bulged on the top.

I use this PSU for testing mobos and such I find on the curb, even temporarily assemble non-high powered computer systems to see how well it works.

lol!

[ratdude747]

Originally Posted by nickt1862

Yes I know this thread is months old.

I have an SP500 which still "runs" with one of the two main filtering caps (I think it's a 200v - hard to read it) a little bulged on the top.

I use this PSU for testing mobos and such I find on the curb, even temporarily assemble non-high powered computer systems to see how well it works.

lol!

better recap it... unless you want to start killing motherboards...

[999999999]

^^ Is it really bulged or is there just a plastic sheet on the top that is bowed to make it look bulged but the cap can under it isn't?