CPU LED lit on motherboard

righteousass · 2 Sep 2010 at 23:46

EDIT: incorrect image, added text spec

Hello

Hopefully someone could suggest the best way to go about troubleshooting this problem.

I had a clean shutdown last time I used my PC. When I've turned it back on I got no POST screen so I checked the mobo and there's a big red LED marked CPU lit up. Uh oh!

My spec is:

HIS iCooler V ATI Radeon HD 5850 Turbo 1024MB GDDR5 PCI-Express Graphics Card
Asus Crosshair IV Formula AMD 890FX (Socket AM3) DDR3 Motherboard
BenQ G2420HDBL 24" Widescreen LED Monitor - Black
Crucial RealSSD C300 64GB 2.5" SATA 6Gb/s Solid State Hard Drive (CTFDDAC064MAG-1G1)
AMD Phenom II X4 Quad Core 955 Black Edition "125W Edition" 3.20GHz (Socket AM3) - Retail
Antec TruePower New Modular 650W Power Suppl
Corsair XMS3 4GB (2x2GB) DDR3 PC3-12800C9 1600MHz Dual Channel Kit (CMX4GX3M2A1600C9)
Corsair H50-1 High-Performance CPU Watercooler (Socket LGA775/1156/1366/AM2/AM3)
Samsung SpinPoint F3 1TB SATA-II 32MB Cache - OEM (HD103SJ)
Antec 300 Three Hundred Ultimate Gaming Case - Black
Sony Optiarc AD-7260S 24x DVD±RW SATA ReWriter (Black) - OEM

I've been running it absolutely fine since Tuesday evening - left on downloading it all the way through to Wednesday evening, then played WoW for 4 hours. Same again today, all day downloading then played WoW for 4 hours. All this time I've been carefully watching the CPU and GPU temps. CPU has been around 25-27c idle and 31-37c playing WoW (which is basically all I've done since I got it!). GPU was around 55c, can't remember mobo temp exactly but it was definitely low 20s.

The PC was off for all of 10 minutes when I remembered I needed to check something on my work email and it did this.

I don't have another AM3 CPU to test. I think my first port of call will be to disconnect the case fans and remove the GPU to listen for the H-50 fan and the water pump motor, though I'm not sure if the pump is noisy, haven't really heard much except the case fans.

Help

righteousass · 3 Sep 2010 at 00:06

Panic over temporarily. Reseated the RAM and I watched the LEDs on the mobo cycle, then I got the POST. In Windows now and temps seem normal - CPU 28, mobo 28, GPU 35.

Could it really have been as simple as the RAM reseating? Bearing in mind this thing has run solid since Tuesday evening and I've only opened it once to take a pic!!!

alex24 · 3 Sep 2010 at 00:13

Short answer, yes. I've had that several times before.

righteousass · 3 Sep 2010 at 02:14

Ran fine since I last posted, and shut it down again. I thought 'what if?' so turned it back on. Exactly the same thing, except this time reseating didn't work.

I've reset the CMOS, though the only setting that was changed in BIOS was memory clock which was defaulting to 1333 so changed it 1600.

alex24 · 3 Sep 2010 at 08:01

Nice PC btw. You may have something set wrong in the BIOS then, perhaps a voltage or timing that's throwing it off. Either way, a board of that quality should not throw a hissy fit over that. Are you overclocking the CPU? The RAM running at 1600MHz is technically an overclock of the CPU as the Phenom II IMC supports up to 1333MHz on paper. Mine manages 1600MHz fine, but something to bear in mind.

righteousass · 3 Sep 2010 at 09:53

Hmmm, could be onto something there Alex. I manually changed DRAM setting from 1333 to 1600. It's up and running again now and going cool at that. Still only a POST problem from cold start, restarting from in Windows doesn't reproduce the problem.

I'll experiment with it this evening when I get home.

Thanks for the comment on the PC. Check it again as I posted the wrong pic and left the GPU off there too - HIS HD 5850

Runs WoW smoothly at max everything, not exactly the right medium for showboating though so I hope to get some other games onto it this weekend. once I've got my perfect setup I'll be ghosting it and starting the overlocking tests.