PC re-boots randomly

[daw123]

I've just completed a new build and my computer seems to re-boot randomly when under load. When it reboots it also boots to a blank screen.

P.S. I have ran Prime95 for about 2 hours successfully, yet it some times does the above after 30mins. Hence it seems to be random. I am guessing that it could be a RAM issue.

I noticed that the timings have been automatically set for 9-9-9-24, yet OCZ say it should 9-9-9-20 - would this make a difference? I don't know if the voltage for the DIMMS is correct (it should be 1.7V). This is also something I will have to check.

I am going to run Memtest later today to see if that comes up with any errors.

The spec is as follows:
PSU: Antec TruePower Quattro 1000W
Motherboard: Gigabyte GA-X48T-DQ6
CPU: Intel Quad Core Extreme QX9770; 3.2Ghz stock
HSF: Thermaltake Big Typhoon 120VX with Thermalright bolt-thru-bracket
RAM: OCZ Gold Edition DDR3 1333Mhz 2x 1Gb DIMMS (in slots 2 and 4)
Graphics: Geforce 7100GS (to be replaced shortly by Sapphire ATI HD4870 X2)
Sound: Creative Soundblaster X-Fi Extreme PCI Express
Internal Hard drives: 2 x WD 200Gb 7200RPM SATA Hard drives
Optical drives: 2 x Sony DVD Re-writable drives
Misc: BT Voyager Wireless PCI card
Hauppauge TV PCI card

SOFTWARE

OS: MS Windows XP Home Edition


Am I on the right track suspecting a RAM issue, or does any one have any other ideas?

Thanks.

 

[Blain]

* Memory
* PS
* AC power source

If you're having problems, stick with the more relaxed memory timings.
Run PC Wizard to check your DIMM voltage.

 

[daw123]

Well a new development in the problem.

I ran Memtest after manually altering the timings in BIOS to 9-9-9-24 and 3:5 DRAM:FSB multipier. I.e. I manually changed the settings I had in Auto to the same with manual set up.

The screen went blank during Memtest, which I witnessed, but the PC did not re-boot. This rules out PSU / AC power source issues. It also rules out OS / graphics card settings interference - PC boots into Memtest.

I believe I may have been wrong previously when I said that it re-booted then the screen went blank.

Also checking HWMonitor, VIN1 (which I belive is DIMM voltage - correct me if I am wrong?) is 1.73V which is fine.

I believe that this could be MB related rather than anything else.

Note: I RMA'd a Sapphire HD4870 X2 because the screen was going to "No Input Detected" about 15 seconds after boot up.

Although, the current problems are a lot more random and there is still a signal going to the monitor, although the screen is blank / black.

I think it could MB related - prehaps I should update the BIOS?

What do people think - is my assessment of the problem correct? Or could it be something else?

 

[Blain]

Some people simply love OCZ memory for their fancy heatspreaders and big rebates.
OCZ has had a checkered past as far as quality components go. Personally I stay away from them.
My favorite DIMM companies are Corsair, Mushkin, G.Skill and for some reason A-Data.

 

[corkyg]

Random reboots are sometimes heat related. Even dusty DIMMs can cause it, because dirt impedes cooling. (Even with heat spreaders.)

Give things a thorough cleaning before making maujor hardware changes.

 

[daw123]

You are right that it could be dust on the DIMMS. However, the temps seem fine judging from HW Monitor:

TMPIN0: 34C (I believe this is the outside temp of CPU or HSF)
TMPIN1: 26C (motherboard or inside case temperature)

All 4 Cores Temp: Approx 43C at idle and about 53C under load

GPU Core Temp: 53C going up to about 60C under load

HDD Temp: 23C

So the temperatures seem to be ok.

P.S. CoreTemp shows the 4 cores at 25C at idle. One hell of a diffrenece between the HW Monitor and CoreTemp figures. I don't know which one is more accurate - hopefully the CoreTemp figure

 

[gplracer]

Give the ram more voltage. What is the rated voltage for the ram?

 

[daw123]

The RAM is rated for 1.7V although OCZ state that the EVP is 1.9V (max. VDIMM).

Am I correct in stating that VIN1 is VDIMM in HW Monitor?

 

[daw123]

Yet a new development.

Today after it being on for about 5 hours - computer went to blue screen - windows has had to close to prevent damage to your computer. Then it said some thing about memory and caching. No specific error report given which was (not) useful. Any ideas what this error screen corresponds to?

I have swapped the DIMMs around and also changed them from slots 2 and 4 to 1 and 3.

May be this will make a diffrence - I know how fussy my MB is about RAM.

By the way, there was very little dust inside the case on the MB. Just a little bit on the SB and the top of HSF.

Please can some one answer my question regarding VIN1 in HW Monitor. Does thie refer to VDIMM? If so I will increase the voltage by 0.05V to 1.78V.

 

[corkyg]

How about setting everything to the mobo defaults? That would eliminate trying to manage multiple variables. I would also go with one DIMM at a time. If the problem is one of them, that would help isolate the fault.

 

[daw123]

Sorry, double Post.

 

[daw123]

I have reset the BIOS to it optmised defaults and removed one of the DIMMS. I have only altered the BIOS settings to enable CPU temp and CPU fan warnings. I don't believe these settings should make much of a diffrence.

Currently I have 1 DIMM in slot 1. I have checked CPU-Z to make sure that the RAM has been recognised correctly.

If the PC passes Prime 95 for a few hours without any of the previous problems rearing their ugly head, then I will try the other DIMM in slot 1.

I have had DIMMS in all four slots and the same problems occured no matter which slots I used, so I don't think it is a MB issue.

If the RAM is knackered then I don't mind replacing it. I have had a look on the net and found this: 4GB (2x2GB) CorsairTwinX XMS3, DDR3 PC3-12800 (1600), 240 Pin, Non-ECC Unbuffered, CAS 7-7-7-20, DHX. It will cost about £200.00 (or $360.00).

This is the RAM I should have bought in the first place instead of using OCZ.

What are people's experiences of this Corsair RAM and are there any better DIMMS for a similar price (btw I live in England)?

Note: I don't think it is temp problems because the RAM was cool to the touch and the NB and SB were also fairly cool. The only hot component in the PC was the GPU, which is what I expected since CPU-Z stated that it was 55C.

 

[daw123]

Well, I'm having re-boots with the first stick of RAM installed. I will try the other one tomorrow

 

[daw123]

I am also going to buy the Thermaltake Dr Power power supply tester to test the PSU.

Does any one know if this tester is any good?

 

[0roo0roo]

always test stability after having issues with minimum setup. all extra cards out, perhaps using onboard video etc to better eliminate possibilities

 

[corkyg]

And, the reboots can also be related to software issues in youir OS. If you have a bootable CD or DVD, boot to it and see if reboots occur. If they do - the hardware is the right track. If it is stable, then the problem may lie in your system HDD.

BTW - only in UK do things get knackered.

 

[daw123]

I am trying the other DIMM of RAM at the moment.

You may be right about the HDD - something was making a strange noise the other day and I am pretty sure it was the HDD (its the first time that has happened). I have got a spare HDD which I can try out.

That said, would the HDD come into play when booting into Memtest?? It rebooted randomly from Memtest

I will also try booting straight from the CD (never needed to do this before so I will have to ask the software bloke in my house how to do it).

I think the most basic set up I could have would be: 1 DIMM, the graphics card (MB does not have on board graphics), and 1 HDD.

Just out of interest; the last PC I built was about 9 years ago (1200Mhz AMD Athlon). How time flies when you are having fun All the PCs I have had since have been shop-bought

 

[daw123]

Well, I don't think it is a RAM issue; I tried both DIMMs individually and I got the random re-boot. What are the chances that both DIMMs are shot??

Time to try booting into Windows XP from the CD. This will rule out a software issue.

After that, I will strip the PC down to bare bones and test that.

I am also going to buy the Mr Power tester (only $15.00) and test the PSU. I think it will be useful to have any way.

I suspect it is an OS or HDD problem.

Could it be also caused by a faulty MB?

 

[Roguestar]

What are the chances both DIMMs are shot? Memtest will tell you that. Download and burn a disc of memtest and give it a go overnight.

 

[0roo0roo]

i doubt a 15 dollar power tester does anything more than confirm voltages.
you already know it boots. how a psu performs under heavy load is not something a cheap tester can test.

 

[daw123]

I am kind of hoping it is the RAM, so I have an excuse to buy the 4Gb Corsair DIMMS

I will try Memtest on each individual DIMM over the next couple of days. I will keep you posted on the results.

 

[daw123]

Well I have tested each DIMM individually and here are the results:
DIMM1 in slot 1 - Ran Memtest for 6hrs and it re-booted at least once. 616 errors after 1hr, 40mins on test 8.

DIMM2 in slot 1 - Ran Memtest for about 5hrs. 401 errors, unexpected interrupt halt twice (I re-booted the PC once to get rid of this error and tried Memtest again and got the second unexpected interrupt halt).

DIMM1 in slot 2 - Ran Memtest for 4hrs. 86 million errors and counting in test 8 (is this a record? )

It is safe to say that both DIMMs are knackered. Time to get my debit card out and buy the 2x2Gb Corsair XMS3 1600MHz DIMMS for about $360. Is there anything better for this price (in England)? I welcome recommendations / suggestions.

btw. I realised that I hadn't disabled the on-board sound in BIOS when using the PCI-E sound card. This seems to have cured the blank screen issues (PC has ran for 36hrs without screen problems).

Edited for spelling / grammar