Power Supplies Keep Failing

beyonu78

Commendable
Sep 8, 2016
9
0
1,510
I am currently RMA'ing my third Corsair power supply for my son's computer. I have been able to find the cause of the problem. I am hoping someone can help me find the root of this problem instead of just keep replacing the power supplies.

The current system build (http://pcpartpicker.com/list/LmYYgL):
PSU: Corsair RM1000x [CP-9020094-NA]
CPU: FX-9370 [FD9370FHHKWOF]
CPU Cooler: Corsair H100i
GPU: GeForce GTX 750TI Superclocked [02G-P4-3753-KR]
Memory: 8GB (2x4GB) Crucial Ballistix Sport [BLS2KIT4G3D1609DS1S00]
Motherboard: Gigabyte GA990FXA-UD3 (Rev. 4.0)
OS: Windows 7 Home Premium 64-Bit
Optical Drives: 2xAsus DRW [DRW-24B1ST]
HDD: 1TB Seagate Barracuda [ST1000DM003]
Case: Corsair Carbide Series 300R Windowed [CC-9011017-WW]
Case Fans: 4x140mm and 1x120mm Spectre Pro LED Fans (w/o PWM)
UPS: Tripp-Lite 750VA 450W [Internet750U]

The original build had a FX-6350 CPU and Freezer Pro CPU cooler with a CX500 PSU. The PSU failed within two months and the replacement (cash warranty) was stepped up to a RM1000. This PSU failed about one year later. The replacement PSU (exchange warranty) showed up as a RM1000x. At this time, I also I had Gigabyte test my motherboard with "no failures in multiple configurations." I added the UPS although my Fluke showed no signs of voltage drops or spikes in any of my receptacles over a 24 hour period. I also upgraded the FX-6350 CPU with the current FX-9370 and the Freezer Pro cooler with the current Corsair H100i. It has only been one month and the PSU is failing again. First, the PC wouldn't turn on in the morning. After cycling the PSU switch I got it to turn on. About two hours later everything abruptly shutdown. The cycle repeats. The shutdowns do not seem to be affected by gamer load or simply sitting idle. The previous PSU's had the same symptoms.

So far I have tried disconnecting the case fans (while keeping an eye on the temps that never went up much). I tried disconnecting everything and "jumper run" just the PSU and it never shutdown. Dust inside the computer or on the case is nearly unnoticeable.

I have monitored the BIOS readings, HWiNFO64 (v5.34-2930), Speedfan, and CorsairLink readings. All readings appear stable and correct except the following. In SpeedFan and CorsairLink (BIOS does not show this sensor) I have seen the MB Temp#2 reach temperatures of 70 C (idle) while all other temperatures remain in the 30-40 C range. I'm not 100% sure but this is possibly the Northbridge heat sink?. With HWiNFO64 (set at 500ms or 1/2 second samples) I notice the MB readings are not what I would expect. The voltage out of the MB (VR VOUT) has a 31.978 volt max (occasional spikes). I don't think it is even possible for the voltage to get that high, so I tend to think it is likely a software glitch.

A lot of information I'm hoping someone is willing to to go through to help me with my reoccurring PSU failures. Any help would be greatly appreciated. I'm willing and able to consider multiple theories or to test out anyone's ideas.

P.S. I don't mean to be rude, but I do not need responses of what components or brands I should have gotten or should add-on to the current build. This would be in no way helpful. My wife has been enjoying almost the identical setup with a higher tier graphics card and less case fans.
 
Solution
You might want to check with Corsair and see if there wasn't a problem or a recall with the RM1000 model. 3 years ago, I purchased an RM750 and it slowly took out my system as it took out itself. Turned out there was a recall on certain serial numbers of which mine was one of them. A temperature autosetting switch (I'd forgotten exactly if it was a switch or setting) for the fan was on a wrong setting, so the psu slowly burned itself up. It had a very slight electrical smell (where it was coming from I couldn't tell) then randomly crashed. Crashing became more frequent, then failed completely. Most of it's end happened in a month.
I went to the Corsair site and looked up my psu to get an RMA and found the serial number had been on...

beyonu78

Commendable
Sep 8, 2016
9
0
1,510
Thanks for the quick responses. Good thinking.

For Someone Somewhere:
The PSU is mounted at the bottom of the case with the intake fan facing down directly above the original case filter. The case is on a small cabinet about 2-1/2 ft. above the carpet and a 1-1/2 ft. away from the wall. The software monitor (not sure how it can detect it, there is no CorsairLink plug on this particular model) shows steady temps of 30C all the way until half a second before shutdown. The heater is on the opposite side of the room and in each failure it has been during the late summer or early fall with the heat off. The ambient temperature in his room is about 25C.

For AppleGwava:
We actually had a power outage that lasted 12 hours about two weeks after getting the UPS and it worked as intended. The idle computer stayed on and I had plenty of time to shut it down. It was pretty annoying to hear the "backup in use" beep for 11 hours though lol. The first RM failed before I added in the UPS.
 
Uhh, you might want to get a socket/outlet tester, could be a wiring problem in the home, or a rat chewed on the wires. Modern and decent quality power strips usually have a feature that also alerts you if the socket you plugged them into isn't up to specs.

For example: https://www.amazon.com/P11VT3-Performance-SurgeArrest-Splitter-Protection/dp/B0012YLTR6/
 

Rexer

Distinguished
You might want to check with Corsair and see if there wasn't a problem or a recall with the RM1000 model. 3 years ago, I purchased an RM750 and it slowly took out my system as it took out itself. Turned out there was a recall on certain serial numbers of which mine was one of them. A temperature autosetting switch (I'd forgotten exactly if it was a switch or setting) for the fan was on a wrong setting, so the psu slowly burned itself up. It had a very slight electrical smell (where it was coming from I couldn't tell) then randomly crashed. Crashing became more frequent, then failed completely. Most of it's end happened in a month.
I went to the Corsair site and looked up my psu to get an RMA and found the serial number had been on recall. There was no notice of a recall I knew before hand and why Corsair allowed the serial number to remain on retailers shelf, I have no clue. But you might want to check with Corsair and find out all the info on your psu.
I may not be the most respectable guy in the point of view of many Corsair fans. I've had more than one Corsair product fail. They indeed advertise and package their products well.
Tom's Hardware Psu Tier list the RM1000 on the 1st and best tier but it also list the RM1000 and the RM750 on the 3rd tier. This I believe is because some are made by Channel Well Tech and some by Chicony. Channel Well Tech being the better.
http://www.tomshardware.com/forum/id-2547993/psu-tier-list.html
Hope this helps
 
Solution