We'd like to remind Forumites to please avoid political debate on the Forum... Read More »
We're aware that some users are experiencing technical issues which the team are working to resolve. See the Community Noticeboard for more info. Thank you for your patience.
📨 Have you signed up to the Forum's new Email Digest yet? Get a selection of trending threads sent straight to your inbox daily, weekly or monthly!
computer is insane
Options
Comments
-
The chances are it doesn't matter what proof you provide to them they'll deny it is their responsibility. I have fun when I return items under a warranty as they try and start talking technical and saying have you tried, this, this, this, this, this, this, this, this, oh and that! I reply with technical info as well and because I'm technically-minded most of the time I have tried all possible tests to ensure it is such a thing that is faulty and can't be (or highly unlikely) to be something else. The excuses these warranty companies use are unbelievable.
Anyhow, as mentioned by T4i, I'd contact t/standards for advice and specifically what laws, Acts, etc your warranty situation comes under Then contact warranty company again and if they're fobbing you off again say that you've been in contact with t/s and under such a law they are required to look into it, etc.0 -
thanks for your help, people. i'm just trying to build up a whole list of possible hardware failures that could be causing this. at the moment, i have mboard, cpu, memory, psu, graphics card... i'm amazed that i'm typing this on it now...
yeah, the graphics card certainly seems to be a hardware failure (although they will probably come back at me with some talk about compatibility, and user error on my part etc etc). i can hold my own on technical talk, but the killer is i have no way to prove it beyond any doubt. when i mentioned the voltages (11.13v on the 12v rail), he said it was within a 10% tolerance and thus must be fine.
i'm not sure i see what their problem is, to be honest... after all, surely they just need to return any faulty components to whichever wholesaler they got them from, and say that they failed within a year, which is surely not on...
would an error in the graphics memory cause prime95 to fail? and if not what are the chances that the memory AND the graphics memory have failed at the same time?
never has a thread been more aptly namedthanks for all your help
0 -
I've looked about various places and from what i've seen Prime95 does not test graphics memory.
Bit worrying that memtest fails though, something is not quite right.
You could try this program called BurnInTest.....http://www.softpedia.com/progDownload/BurnInTest-Professional--Build-Download-3299.html
It might give you a bette idea to what is actualy failing. It enables you to do a 2D, 3D, CPU, Memory test. Try the 2D and 3D's to see if it brings owt up.
Also, in BIOS you may have an option of Agp FastWrites or just Fastwrites, what are these set to currently? Many GPU's/Mobos have incompatability problems with fastwrites enabled. Another is Sideband Addressing which again can be disabled for more stability. You sohlud have a AGP Aperature size in the BIOS too, options are 32, 64, 128 and 256 from what I can remember. What is your set to?
Fastwrites and SBA have about 0.1% performance difference so disabling them is not going to slow your machine in a way that you will notice it.0 -
i'm downloading the program now.
fastwrites are not supported (according to everest, anyway).. sideband thingy is supported and enabled, so that might be a problem.
aperture size (whatever that is!) is 64mb...
i'll put more details tomorrow... far too tired to do extensive diagnostics tonight... i'll end up reverting to kicking the darn thing a la basil fawlty...
thanks!0 -
just a quick update... i ran the burn in tests on cpu, 2d, 3d and memory, and it didn't come up with any errors... however, the trial version only lets it run for 15 minutes, so it is a little inconclusive0
-
Been reading with interest. Does your graphics card have a seperate power connector? If so, might not be getting enough volts if the psu is being run at full chat. Just an idea is to swap the psu power connector on the graphics card if there is one. I have in the past had a molex connector fall apart and cause a whole lot of problems with the hdd it was plugged into.
To test the graphics card memory normally you would download 3dmark or similar and test it to check for sparklies when it renders 3d stuff. Overclocked cards normally show these when clocked too much or the ram is failing.0 -
Ferris wrote:i'm downloading the program now.Ferris wrote:fastwrites are not supported (according to everest, anyway).. sideband thingy is supported and enabled, so that might be a problem.Ferris wrote:aperture size (whatever that is!) is 64mb...
Did you say your memory fails at the same place/time in memtest86?0 -
T4i wrote:Sorry, I should have looked for the trial version specs. 15mins is no good
You could try disable Sideband.
I dont want to start an arguement about what Apertaure Size is/does but either 64 or 128MB is fine and the 'norm'
Did you say your memory fails at the same place/time in memtest86?
well, it always says it fails one specific test at something like 381.6MB, and the failing address codes all share the same last three or four characters (xxxxabcd) but i don't know if this indicates proximity in the way memtest deals with memory addresses..
i have some screenshots of both diagnostics if it would help to understand.
i've just had a thought... i haven't run memtest since the BIOS was flashed to the latest revision, which just *might* have made a difference. although it does still fail prime95. but not the memory test in MicroScope. hmmm... maybe i should try it again.
thanks all!0 -
Yeh, try with the new BIOS installed. Itermidant problems are a pain to deal with, they take so much time and patience to fix.
You have 2 sticks of 512MB RAM? Do you know the make/model of the sticks? Is your BIOS picking up the correct RAM tmimings?
Infact, look in your BIOS and tell me what your voltages are set to too. You should have a voltage for CPU, VDimm, AGP, VDD.0 -
I've seen this problem long time back when I worked as a network engineer. Can't say what's the real problem, but they are always realted to memory. Can't you try a chip from another pc, perhaps belonging to your friend? Could be a bad memory slot. Try changing the slot for the RAM chip. May as well work...
Also, may sound illogical, try changing the memory you allocate to the graphics card and would also suggest changing the size of the swap file.
Overheating is certainly not a problem from the temperatures you have mentioned here.
Hope something works.
Cheers
Terry0
This discussion has been closed.
Confirm your email address to Create Threads and Reply

Categories
- All Categories
- 350.8K Banking & Borrowing
- 253K Reduce Debt & Boost Income
- 453.5K Spending & Discounts
- 243.8K Work, Benefits & Business
- 598.6K Mortgages, Homes & Bills
- 176.8K Life & Family
- 257.1K Travel & Transport
- 1.5M Hobbies & Leisure
- 16.1K Discuss & Feedback
- 37.6K Read-Only Boards