Sign in with
Sign up | Sign in
Your question
Closed

Anyone like mysteries? Extremely erratic behaviour!

Last response: in Systems
Share
October 15, 2009 11:48:33 AM

I'll apologise in advance for the disorganised, and possibly vague nature of this post, but the problem to me seems pretty disorganised and vague...

I built a computer recently, and had a few problems with it, which after a lot of hunting I traced back to the power supply. After replacing it, the computer seemed to work fine. However, it had a few problems: firstly it would only detect 4gb out of the 6gb of RAM, and secondly, it would occasionally randomly restart or shut down, and when I powered it back on, it occasionally would fail to POST. The next attempt to turn it on would usually yield a successful post, but with a reset bios (times and dates reset etc).

So, I took all the RAM out, and re-inserted it stick by stick, restarting after each stick I put in. And after inserting the third and final stick, it was registering all 6gb.

However, since this happened, the random restarts have been worse: sometimes it will just restart. Sometimes it will freeze completely and not respond at all, sometimes I'll get a blue screen, but have been unable to keep it there long enough to record the stop code, and sometimes the screen will tear before suddenly shutting down. Also, more and more frequently the computer is refusing to POST. When I switch it on, I get a blinking _ on a black screen, no beep code or anything.

I've tried removing all the RAM and testing them one by one, and the same problems persist regardless of which combinations I try.

Before all of these problems occurred the system worked fine under reasonably high load for several hours, so I'm relatively confident that its not a heat issue on either the CPU or the graphics card.

My thoughts are that it is most likely to be a RAM issue... but how likely is it that all 3 dimms are faulty?


For reference, here is my spec:

intel i7 920
asus p6td deluxe
6gb ocz gold RAM
radeon HD4870 1GB
WD Caviar Black 1TB
OCZ Vertex 60GB
LG GGW-H20L blu-ray drive
Zalman CNPS10X Extreme


Can anyone suggest anything?

Sorry this has been so long and disorganised, if you need any further information, please ask and I'll do my best to answer promptly.
October 15, 2009 4:52:22 PM

To me, it sounds like the problem might be your motherboard. I'm not really sure what I would do to test/confirm this if I were in your situation...
If you've got a friend or someone you know who also has a LGA 1366 PC try and convince them to let you test out all of your components on their motherboard and see if the problems persist.
You could also breadboard the PC piece by piece to rule out any other component.
a b B Homebuilt system
October 15, 2009 4:57:53 PM

The problem is most likely not faulty RAM, but incorrect RAM settings. You need to manually set the RAM speed/timings/voltage in the BIOS. The correct settings for the 1600MHz OCZ Gold RAM are 1600MHz with 8-8-8-24 timings at 1.65v. Setting the voltage to 1.65v is critical. Your motherboard owners manual should tell you how to set all the RAM settings in the BIOS.
October 15, 2009 10:30:23 PM

@kufan64 - Unfortunately no-one I know has the right stuff for me to swap test my equipment. Not even the local pc-world has the capability to do it! I'll try breadboarding over the weekend.

@shortstuff_mt - I did notice that the settings in the BIOS were set for 1.50v, but when I adjusted that to 1.65v it warned me that it could potentially damage the processor, so I decided not to go through with it. Is there a real risk of processor damage, or is the BIOS just erring on the side of caution?

Thanks for your help. Any other suggestions or advice would be greatly appreciated.
a b B Homebuilt system
October 15, 2009 10:34:29 PM

Don't worry about the warning. Intel says up to 1.65v is safe. You can't expect your RAM to run correctly if it's not receiving enough voltage. Try manually setting the speed/timings/voltage, I think you'll be happy with the results.
!