Persistent mysterious restarts, TS'ing has not resolved.

[slytx]

Hi everyone

At work, I have a computer that I built that has experienced shut downs intermittently. When this happens, the entire machine powers down, but then starts booting up without pressing the power button.

We thought it might be a CPU temperature issue, were running SpeedFan, and bought an aftermarket CPU heatsink and installed it. The temp's decreased, but the shut downs persisted.

The shut downs seem to happen often when a Save function is attempted. Often using Photoshop and Illustrator, when ctrl+s is hit to save is precisely when the restart will occur. Upon bootup, I'll be left with only a temp file for the Photoshop file I was trying to save and it'll be broken. This is indicitive to me as some kind of Harddrive read/write issue. I ran some diagnostics on the Harddrive and got conflicting results, sometimes bad someitmes fine.

I replaced the HDD anyways, and all seemed OK for about 4 or 5 days. But alas, the problem persists, once again precisely when ctrl+s is hit. This confirms to me that HDD probably is not the issue. However, there are times when the restarts occur during usage when an application is not receiving a command to save.

The full specs are the following:

GIGABYTE|GA-X58-USB3 X58 1366 Motherboard
Intel i7 950
160GB Seagate Barracuda (newest HDD)
EVGA Nvidia GT440
PSU LOGISYS CPT|PS575XBK 575W RT
CORSAIR 4Gx3 of DDR3

Im wondering if the RAM has anything to do with any of this, or possibly the power supply. Both seem unlikely, however is there a tool I can use to assess the integrity of both of them?

 

[alphaalphaalpha]

I'm thinking it's the PSU, not the RAM. RAM shouldn't make it restart like this, just crash. The only thing I could recommend is testing your system with another PSU. If it works, then you know your problem. If not, then you know one more part that isn't the problem.

IE chances are that it's the PSU. A faulty PSU can cause ALL other components to act up if it isn't running properly. If it is the PSU, the next thing to do is try to find out why it died. Maybe the PSU isn't getting cooled well enough, maybe it's over-burdened, maybe there was a static shock to a component and it got damaged, etc. Don't open it up unless you already don't have an active warranty or for whatever other reason, you can't send it back for a replacement. Also, Logisis is a crap PSU brand, they break easily and often. If it's broken, then the next one is just as likely to break early on too. In fact, even if it's not the problem, I'd replace it anyway because it probably will become a problem at some time, maybe taking one or more other components with it when it dies.

Buy for Antec, Corsair, or Seasonic for PSU brands. There were a few $40-$60 PSUs around your wattage needs last I checked from all three of them on Newegg.com

Remember, PSUs are not a part you should skimp on because poor quality ones can cause problems such as yours and end up destroying other components too.

 

[440bx]

As Alpha above mentioned, it is quite reasonable to suspect the power supply.

However, while reboot (due to defective RAM) without being preceded by a quick blue screen are rare, they are possible.

I suggest you google "memtest86", it is a free program to test memory. This will allow you to test your memory and either rule it out or finger it as the culprit.

Regarding your CPU temperature, have you checked in the BIOS if there is a maximum CPU temp specified (which might be set a bit too low perhaps ?) Too low a value would normally cause the CPU to shutdown but a reboot is not out of the question. It may be a bit of a long shot but, it is too simple to check to ignore it.

HTH.

 

[slytx]

Thanks so much for the advice genlteman. I have not checked the BIOS minimum temps, and have been meaning to install a second temperature monitor. The one I am using now is SpeedFan. Going to give memtest86 a try, for sure.