This PSU has been powering my main system for more than 4 years now. It powered my old P4 system, my Pentium D system, and now my core i7 system. I've run the pentium D and core i7 systems 24/7 under full load for months on end, so I've run this guy pretty hard over the years. Anyway, earlier this week my computer started having very odd errors (USB system not responding, LAN shutting down, unable to open programs, sound not working, etc) so I began troubleshooting. It seemed odd to me that so many different systems would be messing up all at about the same time, especially when most of the system is only 4 months old. And so I thought to myself, well! it could be the power supply! I had cracked it open this summer, out of curiosity, and saw it full of Teapo, but it all looked good, and so I put it back together. I was planning on doing a preventative recapping on this when I was hope over Christmas, as well as the PSU in my server (Rosewill green 420W or something).
So I pop it open today, and find that one of the Teapos is bulging! So to make a long story short, I recapped it
Here's the label - the 2 12v rails do have separate OCP protection. I couldn't read the part numbers, but each rail has 2 TO-247 package schottky packs, so I trust that the ratings are good.
Here's a sort of overall before shot. The input filter is crazy! Total (including this and on the AC receptacle) there are 3 X1 series caps, 4 Y2 series caps, 3 MOVs, 2 line filters, fuse, current inrush limiter, and bleeder resistor. The big cap is a Teapo 330uF, the top was flat so I'm not worried about it. That giant wrapped coil is for the active PFC.
Here's the secondary side. The big cap farthest to the left is the one that's bulging (obviously). The 5vsb filter is kind of visible in the lower right. The little crooked cap to the left of the main filter coil was very odd, I'll get to that later.
Here are the improvements I made to this PSU:
+5vsb
1x Teapo SC 1500uF @ 10v 10mm and 1x Teapo SC 1000uF @ 6.3v 8mm -> 1x Panasonic FM 1500uF @ 10v 10mm and 1x Panasonic FC 1200uF @ 6.3v 8mm
3.3v
1x Teapo SC 2200uF @ 10v 10mm and 1x Teapo SC 2200uF @ 6.3v 10mm -> 2x Panasonic FC 2700uF @ 6.3v 10mm
5v
1x Teapo SC 2200uF @ 10v 10mm and 2x Teapo SC 1000uF @ 10v 8mm -> 1x Rubycon MCZ 2200uF @ 10v 10mm and 1x Panasonic FM 1000uF @ 10v 8mm
12v
1x Teapo SC 2200uF @ 16v 10mm (bulging) and 1x L-tec TK 47uF @ 25v 5mm -> 1x Nichicon HE 3900uF @ 16v 12.5mm and 1x Panasonic FM 680uF @ 16v 8mm
The little L-tec was the one I mentioned in the other picture. I have no idea why they added a 47uF cap to help filter the 12v rail, I doubt it did much of anything. The TK series is just general purpose, 1000h rated. Lame. And since I wasn't able to stuff the giant 12.5mm HE all the way to the board (it's about 3/8 inch above the board), I decided to stuff an FM in there to help with things.
Here is the 5vsb filter after I got done with it.
And here we have the output filters. The huge HE was only a few mm shorter than the heatsinks.
Here's a top down shot of the whole thing. The main transformer is huge, the same height as the heatsinks.
After I recapped it I plugged it into a couple power resistors, and let it run for a few hours, to make sure nothing was wrong. Then I threw it back in my system, and everything has seemed to be fine since (it's been about an hour). The voltages look good, both the 12v and 5v are right where they were, and the 3.3v is a hair lower. The CPU is also running a few degrees cooler, but that's probably for some other reason.
What d'ya guys think? It's too early to tell if any similar problems are going to happen. I'll reply back if this didn't solve my problems.
So I pop it open today, and find that one of the Teapos is bulging! So to make a long story short, I recapped it
Here's the label - the 2 12v rails do have separate OCP protection. I couldn't read the part numbers, but each rail has 2 TO-247 package schottky packs, so I trust that the ratings are good.
Here's a sort of overall before shot. The input filter is crazy! Total (including this and on the AC receptacle) there are 3 X1 series caps, 4 Y2 series caps, 3 MOVs, 2 line filters, fuse, current inrush limiter, and bleeder resistor. The big cap is a Teapo 330uF, the top was flat so I'm not worried about it. That giant wrapped coil is for the active PFC.
Here's the secondary side. The big cap farthest to the left is the one that's bulging (obviously). The 5vsb filter is kind of visible in the lower right. The little crooked cap to the left of the main filter coil was very odd, I'll get to that later.
Here are the improvements I made to this PSU:
+5vsb
1x Teapo SC 1500uF @ 10v 10mm and 1x Teapo SC 1000uF @ 6.3v 8mm -> 1x Panasonic FM 1500uF @ 10v 10mm and 1x Panasonic FC 1200uF @ 6.3v 8mm
3.3v
1x Teapo SC 2200uF @ 10v 10mm and 1x Teapo SC 2200uF @ 6.3v 10mm -> 2x Panasonic FC 2700uF @ 6.3v 10mm
5v
1x Teapo SC 2200uF @ 10v 10mm and 2x Teapo SC 1000uF @ 10v 8mm -> 1x Rubycon MCZ 2200uF @ 10v 10mm and 1x Panasonic FM 1000uF @ 10v 8mm
12v
1x Teapo SC 2200uF @ 16v 10mm (bulging) and 1x L-tec TK 47uF @ 25v 5mm -> 1x Nichicon HE 3900uF @ 16v 12.5mm and 1x Panasonic FM 680uF @ 16v 8mm
The little L-tec was the one I mentioned in the other picture. I have no idea why they added a 47uF cap to help filter the 12v rail, I doubt it did much of anything. The TK series is just general purpose, 1000h rated. Lame. And since I wasn't able to stuff the giant 12.5mm HE all the way to the board (it's about 3/8 inch above the board), I decided to stuff an FM in there to help with things.
Here is the 5vsb filter after I got done with it.
And here we have the output filters. The huge HE was only a few mm shorter than the heatsinks.
Here's a top down shot of the whole thing. The main transformer is huge, the same height as the heatsinks.
After I recapped it I plugged it into a couple power resistors, and let it run for a few hours, to make sure nothing was wrong. Then I threw it back in my system, and everything has seemed to be fine since (it's been about an hour). The voltages look good, both the 12v and 5v are right where they were, and the 3.3v is a hair lower. The CPU is also running a few degrees cooler, but that's probably for some other reason.
What d'ya guys think? It's too early to tell if any similar problems are going to happen. I'll reply back if this didn't solve my problems.
Attached Files
Comment