Autopsy Report: Corsair 650 PSU Smoked

CompuTronix

Intel Master
Moderator
Guys,

Last night my Corsair CMPSU-650TX instantly, and without any warning, smoked and tripped the breaker on the UPS. After opening the PSU, (thus voiding the warranty), close visual inspection showed D301 on the main board, which is a 1N5406 rectifier diode, (600 Volts, 3 Amps), was split longitudinally and partially melted.

I lifted one leg from the circuit and Ohmed it, which indicated a dead-short forward and reverse. Without a schematic diagram, (which I'm sure is proprietary), it's difficult to determine D301's exact function, but a little reverse engineering shows that it feeds a switching FET bank on the input side of transformer T001.

Since this PSU was less than a year old, and was well ventilated in an Antec 900, my questions are:

(1) Has anyone encountered this problem with the CMPSU-650TX?

(2) Does anyone know if the 1N5406 is a high failure rate component in this PSU?

(3) Are there any other components known to cause D301 to fail?

Thanks,

Comp :sol:
 
Solution
The 1N5406 has been around for many years and is a standard recovery 3A, 600V rectifier. Because it's standard recovery speed it was most likely in the AC (50/60Hz) input section and not in the higher frequency power conversion section unless as a protection function. Without seeing a schematic it's purpose is unknown and a thorough failure analysis is needed to know what happened. Often the part that has the obvious damage is not the only failure or the root cause.
Smoke and smell incidents are being taken very seriously nowadays especially in Japan.
I'd think/hope Corsair would be very interested in understanding and correcting any quality or design problems.

CompuTronix

Intel Master
Moderator
Johnny,

Thanks for the input. Actually, I was quite astonished when it happened, since Corsair PSU's have earned good standings. :D I then thought, how damned inconvenient to have my rig shut down at 10:43 PM, right in the middle of a movie! :pfff: Since this made me feel a little peeved, :fou: I thought, screw the warranty :kaola: ... I want to know what the hell failed, :heink: so I punched a Phillips screw driver right through the warranty sticker, and backed out the screws! :lol:

Comp :sol:
 

CompuTronix

Intel Master
Moderator
I read Newegg's 1 egg reviews, some of which accurately described the failure I experienced, so it happens. Whether these particular groups of failures are all the result of a bad batch of 1N5406's, a variety of other component failures, or an inherent design flaw, I suppose I'll never know, unless I can find a Tech Support Engineer willing to discuss it with me. I can only hope that Corsair has an ongoing program to debug and improve their PSU's through analyzing the RMA's they receive.

Regardless, I like the supply, and have already installed a new unit, so my rig is back up and running. :D Let's see how it goes. :sol:
 

leon2006

Distinguished
Put it back to together and call corsair. You will never know they may still replace it. Don't say a word put back and send it.

Sometimes all they care for is if they receive the RMA material. Once confirmed that the RMA came in they ship a replacement unit ASAP.


 

ogoB

Distinguished
Oct 27, 2009
1
0
18,520
The 1N5406 has been around for many years and is a standard recovery 3A, 600V rectifier. Because it's standard recovery speed it was most likely in the AC (50/60Hz) input section and not in the higher frequency power conversion section unless as a protection function. Without seeing a schematic it's purpose is unknown and a thorough failure analysis is needed to know what happened. Often the part that has the obvious damage is not the only failure or the root cause.
Smoke and smell incidents are being taken very seriously nowadays especially in Japan.
I'd think/hope Corsair would be very interested in understanding and correcting any quality or design problems.
 
Solution

CompuTronix

Intel Master
Moderator
Since I troubleshoot electronics for a living, I couldn't agree more, which is why I asked above, "Are there any other components known to cause D301 to fail?"

Regardless ogoB, welcome aboard! Based upon this excellent 1st post, I'm certain that you'll be a valuable asset, and a credit to the community here at Tom's! Incidentally, thanks for the "Best Answer"! :D

See you around the Forums,

Comp :sol:
 

XMSYellowbeard

Distinguished
Nov 2, 2006
933
0
19,060
Forwarded link to the PSU guys. Thanks for the input, we are always willing to look at ways to improve our quality and QC procedures. That said, I am not aware of any systemic failures in the 650s and we sell a LOT of them.

Sorry you had an issue and I'm glad it did not take any other gear with it when it died.