Weird power up problem in new system

BDEE

BDEE

BDEE

Soldato
Joined
21 Apr 2011
Posts
3,210
Weird power up problem in new system *Updated - Memtest complete*

Just built my new system. Specs are:


Asus P67 Sabertooth Mobo
i7 2600k 3.4GHz
Asus HD6950 Direct CU II
16GB Corsair XMS3 DDR3 (thanks to help in previous thread re changing to this memory!!)
Cooled with a Coolermaster CMV6GT

PSU is Enermax Galaxy EGX1000
Using one DVD R/W, one Samsung 1tb SATA drive and one older 80GB WD Raptor SATA drive.


I built yesterday, had two random lock ups. No BSOD, no anything....just total freeze.

Secondly, when powering up, the power comes on for about 4 seconds, turns off for about 2 seconds, then comes back on again. It does this every time I power up. It also does it when pressing reset button - just powers off then back on again.

Never come across this before. It wasnt doing it before chaning the hardware.

I have the DVD drive and both HDDs running on the same power cable which has three power sockets on. 2x 8 pin connectors both attached to GFX card and both sets of power cables connected to mobo.

I did a brand new install of Windows 7 Ultimate (64bit obviosuly). Nothing overclocked (yet!) - just running vanilla at the moment.

Any ideas?
 
Last edited:
No not tried to boot without them all plugged in - Will give it a go tomorrow and see what happens. Hope it isnt that as I am trying to keep cabling to a minimum.

Yes it is fully up and running - using it now.

The random lock ups are just that - Totally random. It has only been 2 so far and ive just been on t'internet when they happened. No BSOD on either occasion, it just locks and is non responsive to any input.

Will post results of the power cable tomorrow - bit late to faff about in the case at this hour.
 
Is this something new?

My previous three mobos have all been asus and none of them have done it, my most recent before this being a P5E3 Premium.

Good stuff if that is what it is :)

Will also run memtest tomorrow
 
Yeah its three seperate sockets on one original cable, not a splitter.

Just trying to work out how to use this memtest program.....got it onto a USB stick....cant find any instructions!
 
OK so had another random crash -

Ran memtest - Crashed.
Ran again - Lots of red - over 10,000 errors

I am now running again, but im only doing it with one memory module in at a time. Looks like its going to be 40mins to test each module, so in about 2.5 / 3hrs I should have full results, but definately does look like one or more of the RAM modules is fubared.

Anyone know if OC is open tomorrow to raise an RMA or will I have to wait until saturday?

Second point..... when I do RMA, is there any point in getting a replacement and sticking wih 16GB or even having 8GB on hand spare for the future, or should I just stick with 8GB and get a refund on the faulty stuff? Its a gaming PC, so I understand 6GB is enough, but with RAM being so cheap I went for the futureproofing option with this.


**UPDATE**

OK, didnt take as long as expected. I ran memtest on each module individually. I had one crash with one of them in, but on a second time of running it all came back fine, so 100% OK tests on all 4 modules individually.


Any Ideas?
 
Last edited:
Freddie589 said:
Hmm...I don't like the sound of them errors on the secound run, have tried error checking the hard drive, this is what i would do after ram, just right click the drive in my computer then tools then error checking.
Click to expand...

Well i've left the memchecker running will all 4 modules back on the board for the last 3 1/2 hours now - Its done two complete test cycles for all 4 moudles and passed with no errors, so christ knows what was going on when those errors came up.

Totally baffling. Hard drives next then - Would be suprised though, as never had a lock up / crashing problem on my old build, its only since upgrading.
 
Bit of an update...

It has just randomly reset itself whilst attempting to play COD4.

Im thinking it wont be HDD related.

So options now for me are PSU (seems like a slim chance I think, but could be), or there is a problem on mobo or graphics card causing it. Going to play a different game and see if there is issues once more load is placed on the gfx card.

I have got another gfx card still from my old machine, but it had a fault when it came out, so not sure I can really use that to test.

If I end up having to RMA the lot I will be very unpleased, especially as it came from two different places. Will see how this game test goes on Left 4 Dead 2
 
Great - Wont even let me play L4D2. Says something about the gfx card not being in the database, loads up the title screen, then just kicks me out.

Facepalm.
 
Got L4D2 working now.

verdict - Played all first campaign and as soon as the ene sequence started total power out and system reset, so I would say GFX card or PSU is the problem. If only my old card wasnt bust so I could test. I dont want to pay £100 for a new PSU just to find the gfx card is the problem.

Anyone live in county durham want to lend me a card to test?!

Grr things like this really pee me off.

Surveyor - Thanks for that. At least I know that isnt part of my issues now.
 
I didnt make any changes apart from boot priorities - Nothing is overclocked. The RAM is running at 1.65v AFAIK. Ive just opened up the case and reseated everything, pulled out the extra cables which I had hidden behind the mobo tray etc. Getting a bit late on to do much more testing now so will run another game now and see what happens then pick it up tomorrow.
 
Its meant to run at 1.65v apparently according to everything I am reading?

Ive tried putting everything to default. Last thing to try for tonight, I was going to bed 2hrs ago! It appears I may have changed something in BIOS for an "Automatic" overclock, so will see if this makes any difference. If not then tomorrow im going to try running on half the RAM and see if that makes any difference.
 
OK so latest update.

I have played about in BIOS and it seems that the Power on / off / on was OC'ing the RAM - Interesting since it was already running at 1.65v, so its very possible this was the problem.

Since altering settings to "normal" it no longer does the power on / off / on and just boots straight up, so one assumes it its just running at normal spec. Fingers crossed that will sort it. Going to test and see what happens.

I have ran passmark performance test, but it didnt really give much info, just an overall baseline score of about 4300 or something, which is pretty meaningless, so will look at furmark.
 
Well, it would appear to be OK now since disabling that "Auto Overclock". Annoyed that it came with that enabled as default.

Played a number of hours of Crysis 2 on full settings with no reboots, no crashes, all fine and dandy. Looks awesome too :)

Only thing now is to actually get teh RAM performing to its best. Not sure I need to overclock the CPU as 3.4GHz is pretty fine enough as it is I think.

Thanks for all the help and suggestions. Final part of the build left to do now : Cable management :D My stuff arrived today but got my daughter all bank hol weekend, so will be a job for Monday night after she goes back to her mothers :)
 
You must log in or register to reply here.
Back
Top Bottom