Unstable new build [help]

[fangerz]

Hey guys, so I recently put together the following build:

Intel Core i7 2600K
ASUS P8Z68-V PRO Motherboard
CoolerMaster Hyper 212+ CPU Cooler
Seasonic X-650 80Plus Gold 650W
Western Digital Caviar Black 1TB WD1002FAEX
G.Skill Ripjaws F3-12800CL9D-8GBRL 8GB (2x4GB) DDR3
ASUS Radeon HD6670 1GB
OCZ Agility 3 240GB SSD 2.5"
Lite-On IHAS324 24x DVDRW

I had originally set it up with an ASUS P8Z68-VLE mobo and was receiving power surge warning error messages on start up saying "ASUS surge protection was activated due to an unstable power supply". It was also causing my pc to freeze and power off during normal use occasionally and instantly power off if I tried to run prime95 blended test. I then sent my PSU and mobo back to the retailer under warranty for testing they said they couldn't find any errors. I asked them to send my mobo to the manufacturer for further testing and bought a new mobo (ASUS P8Z68-V PRO Motherboard) while I was waiting to test my system with a new component. I just set everything up with the new motherboard and although I haven't noticed any power surge warnings, my computer froze maybe 40 mins after starting windows and then I ran prime95 blend test and my comp instantly shut down.

What else could it be? I figured the two main culprits were my mobo and PSU; however, it seems highly unlikely that I have received TWO faulty motherboards and I would think that my PSU is more than capable of supplying my build with power (also it seemed to be fault free).

With my first motherboard I was unable to run memtest as it froze every time I tried to run it. Prime95 also caused my pc to shut down. However, after changing my ram cas timings (to faster timings) in the bios it seemed to fix everything; I could run memtest for 6 passes 0 faults and prime95 ran fine on all test modes. I thought I had fixed the problem miraculously; however, my computer started freezing again and the power surge message would appear on start up. This is when I decided to send the PSU and mobo back to the retailer.

I now have the new motherboard and memtest runs fine but the prime95 test causes a shut down as I said.

In all the tests I ran my cpu temps were all low/normal according to realtemp (idle below 30 deg and max at 60 during torture test)

I'm thinking maybe it might be something to do with my voltage settings? Although I have not touched them so it would be strange that voltage settings are causing a problem if they are on stock settings right? Just looking at CPU-z voltage numbers it says core voltage is 0.9 right now which seems low but I think that is just because it is idling as my cpu is showing a core speed of 1600MHz (it jumps up to 3500MHz during cpu torture test and voltage up to ~1.1).

I am kind of lost as to what the source of my problem could be. Does anyone have any ideas as to what might be causing my problems?

Any help will be much appreciated.

Thanks.

ps. the way it freezes now is strange. It will lag out for maybe 5 minutes unresponsive and then all of a sudden everything happens quickly and the system is back running normally again.

 

[simon12]

Have you got a replacement PSU? If not I suspect this is the most likely problem (its enough power and a good make but could still be faulty. If you have got a replacement PSU I suspect another component is causing the problem. I suggest try using the intel onboard graphics and removing the GPU. Also try running with just the hard drive and motherboard connected to the power with no case fans or anything.

 

[fangerz]

Hmm my only other psu is 300W from my old dell set up, I don't know if I want to connect that to my new gear. Now that you mention disconnecting case fans etc. although this is probably trivial, in ez mode bios one of the cha fan monitors bugs out. It will show a normal speed and then every now and then it will instantly jump to maximum 109000 revolutions and then go back to 10000 rpm.

Then again I did have everything set up outside of my case with just the mobo, gpu, ram and cpu fan connected to the psu and screen and I was getting the power surge warnings then before any chasis fan was connected.

The people where I bought it from said they stress tested my PSU a few times and couldn't fault it at all; maybe they just didnt want to refund

 

[simon12]

Ok just try without the GPU, if this runs fine it either means the PSU is at fault and the lower load allows it to run fine or the GPU is causing power surges and it at fault. If you still have the problem then I think it must be the power supply unless there is a problem with your mains supply. I have never heard of it but i suppose its not impossible the RAM could do something. I don't consider myself an expert so maybe someone else can suggest something.

 

[fangerz]

update: I just took out my gpu, booted up using the onboard graphics and ran prime95... no change; instant power off :S

 

[gamerk316]

Could be RAM, though a BSOD would be a more likely outcome. I'm blaming the PSU on this one.

 

[jsc]

Prime95 small fft's tests the CPU pretty well. Large fft's and the blend test put more stress on the memory. If small fft's runs well, but the blend test fails, you most likely have some kind of memory problem - even if memtest does not report any problems.

 

[fangerz]

im running the blended test fine right now (no instant power off) with just one stick of memory in my comp. Does this mean anything? I will leave it for a while then try it with only the other stick see if there is any difference. What if it runs fine with my other stick as well? how will I know which one is faulty (if they are faulty).

 

[fangerz]

I might really love you guys... I dont want to jump to any conclusions but... I ran blended test with my first stick for ~20 mins fine no power off nice cool temps. Then I took it out and put in my other stick same slot and ran prim95 - instant power off. Restarted tried again - instant power off. Then put the other stick back in ran prime and its running fine approx 10 mins now.

It kind of coincides with the cas timings fix I had done before, i.e., ram related.

Is this meaningful do you think? could it still be my psu or something else?

 

[Uther39]

You sir, have a faulty RAM module, i see they are rated at 1.5v to 1.65v this could also be a cause of the problem as SB CPUs are known to like 1.5v modules only, RMA your RAM and buy some new modules rated at only 1.5v.... LIKE THESE

CMX4GX3M2B1600C9 - 4GB (2x2GB) Corsair DDR3 XMS3 Classic, PC3-12800 (1600), Non-ECC Unbuffered, CAS 9-9-9-24, 1.50V

http://www.scan.co.uk/products/4gb-(2x2gb)-corsair-ddr3-xms3-classic-pc3-12800-(1600)-non-ecc-unbuffered-cas-9-9-9-24-150v

 

[fangerz]

Happy days!

Out of curiosity, do you think that the faulty ram could have damaged my system in any way? Could the unstable power issues have caused any wearing or damage to other components?

Also, would I be advised to not use the stable stick either and send them both back? Or can I keep the working one so I can use my comp while I wait for new sticks?

 

[Uther39]

No other damage should of occurred, and you need to send both sticks back as you will need a matching pair of new ram modules like the above i posted.

 

[correon]

The faulty ram haven´t damaged you system i´m sure - but it´s the reason to all your problems i´m sure - get some new and you´re good

 

[fangerz]

Why wouldn't memtest show any errors I wonder?

 

[correon]

Hmm maybe the psu have made the damage on the Ram-stick?

 

[fangerz]

Just an update. I ordered a new cheap psu and ram modules to test things out. I put in the new ram and and ran prime blended and it didn't power off so I thought that it must have been the ram modules. I thought everything was fixed so I went about normal computer use and then I got a lag spike, comp froze up loading for about 30 seconds then back to normal. I have connected the new psu up and am running on it right now seems to be going fine...

Are there any other tests I can run? How do i test my PSU? (I had it sent in and the retailer couldn't fault it as I mentioned).

 

[fangerz]

Ran the new psu without problems for most of the day and reconnected the seasonic and within 40 mins or so, got a freeze up.

I also connected the faulty ram module while the new psu was connected and it powered off during prime. So it seems as though the ram module AND my seasonic are faulty...

Could the Seasonic have caused damage to the ram module perhaps as Correon mentioned?

 

[Uther39]

Yes it may well of done, or you may of just been very unlucky and bought 2 faulty units, RMA your RAM and PSU with a note on your findings, order a new XFX unit as in my 11 years of experience i have found these on average to be the most reliable.