Hardware·18,997 discussions

Hardware troubleshooting needed

Locked

Hey all, I have a serious problem with my computer which I cannot fix, which is slightly embarrassing as I support end users also with their hardware.

First problems arose a good month ago, when I left my computer on as so many times before when I went to the gym, and came back to find it in hibernation. It wouldn’t wake up, so I rebooted. I didn’t get any visual, though I heard it booting. Now it seems I have 2 ports on my graphics adapter, and only one of them works. I then got the visual back, but when working longer than 5 or 10 minutes, the computer froze. No blue screen, no black screen, no event log, simply freezing. The lights on my keyboard also went out, leading me to suspect it was a power problem.

So I took the computer back to the store, and they fixed it for me. I was hoping they’d replace the PSU (if that was problem), because that was the only part still in guarantee. However, they didn’t replace anything, they overclocked my machine. Specifically, they set the DRAM timing in my BIOS on manual, all with the default values. I looked into the manual of my motherboard (ASUS PQ5), but it isn’t very informative.

What scared me most was they didn’t connect the PSU fan to the motherboard. When I did, I noticed in the BIOS it would be marked in red, giving around 500 RPM, unlike the others, who were at 2000 RPM and not marked in red.

Now the pc ran fin a couple of days, but I have games frequently crashing, and most notably my display adapter drivers crashing a lot, and the problem is getting more and more frequent.

List of components:
– ASUS P5Q motherboard
– Intel Dual Core processor (2*3 Ghz)
– Nvidia GForce GX2
– 4 Gig Corsair RAM DDR2

What component do you guys think is causing this trouble? I have not really access to spare parts, and I don’t want to go to the stores (it’s a long and thus expensive trip) unless I’m more or less certain.

Only the PSU is under guarantee, anything else I’d have to buy again, which I don’t want to do unless I’m also pretty sure. Could it be the memory? I wouldn’t mind buying that again, it’s rather cheap, and an excuse to get me some DDR3. I hope it’s not the graphics adapter, that would cost me as much as a week holiday.

Topic

Clarifications

In reply to Hardware troubleshooting needed

Clarifications

check the ff

In reply to Hardware troubleshooting needed

1. Memory – Run memory diagnostic test to each of your RAM seperately if you have more than 1 pc of RAM and see to it that all are working fine (I assume you have memory test software or boot CD since you’re a tech support).

2. CPU/FAN – Open the BIOS and check for hardware monitor, observe the temperature of your CPU and check if its in normal temp or above, if your CPU temperature increase in above normal then you might need to upgrade/change your CPU fan, there are lots of new technology for CPU cooling system to choose. Try also the CPU diagnostic tool if possible to detect if CPU is failing though its very rare based on my experienced.

3.Video Card – Have you tried removing your video card and using built-in video if ever you have built-in in your motherboard? If memory, CPU and CPU are all working fine then videocard is possible to be damaged, try using your videocard to other PC and observe if it has the same effect as with our PC.

4. PSU – I cant really tell exact things how to determine if PSU is defective but it should be OK if your system unit is working, for further assistant in PSU I think you can coordinate to the store where you bought it since its under warranty..

UPdate for result of every test..

Will try that thanks, some remarks

In reply to check the ff

1) I have not done that. I actually have four separate modules, so I’m assuming I’ll have to test all four of them separately. Will give it a go this evening and post back.

2) Hardware monitor does not show abnormal temperatures (at least I think it’s not abnormal, can’t really find a reference online), but I have three fans, one on the chassis, one on the CPU, and one on the PSU. Only the one on the PSU is marked in red and thus not running fast enough.

3) I don’t have built in video, but I think I can borrow an old AGP card from work. Will try that after the memory tests, and post back. As for testing the card on another PC, that’s more difficult. Will have to find another PC with PSU powerful enough, that could be a problem. I think my friend has one, will try that, but it can take a while. Though it’s likely the video cards laying around here won’t be powerful enough to work in vista.

EDIT: silly me, my graphics adapters is pci express, I cannot find another graphics card. have to wait till my friend can test my card then. Seems the only way to find out anything.

4) Neither can I, but from the experience I have, some pretty weird errors can come from faulty power supplies. We have a few PSU laying around at work, but none powerful enough to power my graphics adapter (it also requires a special slot on the PSU). If I still get weird errors after testing with another graphics card, I might try that, at least it’ll narrow down the problem a bit until I can “force” my friend to test my graphics card.

Thanks for the tips so far. Part of the problem is all spare parts I have access to are several years old, while my system is relatively new.

Update: memory and cpu

In reply to check the ff

1) Memory check doesn’t give back any faults.
2) CPU temperature around 30?C/86?F, motherboard temperature around 25?C/77?F.
3) No built in video card, so it’s waiting till I can get my hands on another card.
4) PSU: also no spare part handy, I’ll wait to see what the result with the other card is.

I went to the store, and they overclocked my machine. Now I don’t want to go again unless I’m more or less sure what the problem is, cause I’m sure they will do anything in their power to convince me it works instead of broken hardware.

With all respect, a pc SHOULD work without manual settings in BIOS. And it works indeed, unless I’m doing “heavy” things, like gaming, but also some other things one can consider demanding. Strangely enough it works for sometime now, but I’m doing nothing but browsing for the moment. If I revert my BIOS back to automatic, it falls out. Should I try to flash BIOS with later version ?

I know it’s not software related, cause I tried diffent OS on different hard disk, same behaviour.

Before Flashing Bios

In reply to Update: memory and cpu

Read Carefully why there is an update to the Bios, the release notes of the update should say what it fixed.

I would avoid flashing the BIOS unless the update clearly sounds like it’s what you need.

definetly look at firmware flash for your video card if possible.

have you tested for conflicting software? msconfig, disable everything that is not needed and reboot to test.

BIOS revisions won’t work then…

In reply to Before Flashing Bios

…because none of those are exactly what I need.

Conflicting software is not an option, I’m a maniac in keeping that list always as short as possible, and to be safe, I even had my OS reinstalled after I got it back. Certain programs react a lot slower than they used to (before the incident first occurred), and I’m talking about standard windows stuff here, like messenger and windows update. Sometimes I have to manually kill the process, sometimes the only way to get them to work is restart. I have an additional USB disk with a kubuntu install, will check how that one works, but last time I checked (before they changed the values of my bios), that showed exactly the same problem, freezing after 10 minutes.

I am looking for firmware for my graphics card (NVIDIA geforce 9800 gx2), but no luck so far. Not on the original CD either. I already updated to the most recent driver I could find, but no success so far.

All games I have played with the same hardware for almost a year, with the same OS, and zero problems till that dreadful day.

Disconnect the CPU fan

In reply to Hardware troubleshooting needed

from the motherboard and if the Fan spins freely when you give it a spin it is probably OK. If there is resistance replace it.
Faulty Fans, Dust and Grime build up on the Heatsink restricting air flow. Remove the Power cable from the rear of the System and disconnect the Video cable. I would give the inside of the case a blowout with compressed air. When blowing air through the fans make sure that you physically stop them from spinning, as they may generate power and damage something. Use an Anti-Static Wrist Strap to ground yourself or by placing the back of your hand on the Power Supply Unit and not moving your feet. This will prevent electrostatic charge from building up and by not taking this precaution it is possible that you could inadvertantly cause damage to the PC from an electrostatic discharge.

Then remove the Heatsink, giving it a thorough clean and reseat the CPU applying new CPU grease.

Will give it a try…

In reply to Disconnect the CPU fan

The fan is no problem: I can see it spinning, and if I rotate it manually, there’s no resistance.

I have no CPU grease on hand atm, but I’ll see if I can get my hands on some. When I do, I’ll try reseating the CPU (and cleaning it in one go).

Another thing I noticed when powering on: I hear the fans spinning, and I get a light at the front, but then it immediately stops working, and falls silent. A second later, it automatically restarts, and boots up. Never seen this before, any idea what this can mean?

I almost wished this thing wouldn’t boot, then at least I’d know what to do and where the problem was (well, in most cases anyway).

I would be concerned with only

In reply to Will give it a try…

having 500rpm on that particular fan. There could be a problem with the power connector on the motherboard. You could get a CPU fan with a molex connector and connect it to the Power supply.

Swap the Power switch to the front panel connector with the Reset switch to rule out a faulty switch.

You can test your PSU using the instructions below

http://www.driverheaven.net/guides/testingPSU/

http://www.ochardware.com/articles/psuvolt/psuvolt.html

http://www.journeysystems.com/?powercalc

Also check the Capacitors around the CPU for swelling or bulging.

You could also try clearing the CMOS and setting the BIOS back to defaults.

Thanks again, will measure later

In reply to I would be concerned with only

Have to see if I can find another fan here, but first I’ll swap the switches.

Those links to testing the PSU are what I really was after, I’ll take the meter home with me and post back this evening with the results.

As for clearing the CMOS, I tried that already, no effect. And no swollen or bulging capacitors (now I finally now that word in English) either.

EDIT: putting BIOS back to default (there’s no fail safe default on this mobo) causes the overclocking to be removed, making the PC freeze after 10 minutes.

Also try this and let it run through

In reply to Thanks again, will measure later

Video Memory Stress Test 1.4

http://www.majorgeeks.com/Video_Memory_Stress_Test__d5896.html

Read the documentation.

readme.en.txt

Lots of failures

In reply to Also try this and let it run through

I get more then 70000 errors when doing the stress test on dirext x (though not on cuda). Unfortunately, I don’t know what it means. Perhaps I used an incorrect parameter? The readme test is not quite clear.

All this points me in the direction of the graphics adapter. I didn’t switch it yet, but I will try to arrange for it to happen this weekend.

Good luck :D {nt}

In reply to Lots of failures

😀

Voltages correct, BIOS working again (for now)

In reply to I would be concerned with only

I measured the voltages, and they are all correct. Though the PSU fan is still at 500rpm when booting, my colleague claims it cools enough because the PSU doesn’t overheat.

I reset BIOS to the default values, but still experience crashes, however, it no longer freezes up. Got a blue screen, going to read the dump to see what driver is causing the problem.

Sounds like

In reply to Voltages correct, BIOS working again (for now)

you are slowly getting there.

Check PS

In reply to Hardware troubleshooting needed

Check the PS, I found one person that had problems. Would not boot sometimes, freeze, all kinds of funny stuff. Bought a $30 PS and 2 months later not a problem.

More like $100+ PSU

In reply to Check PS

Yes, that’s also my experience. However, the graphic beasts from NVidia don’t run on a $30 PS, they run over $100 because they need a minimum of 580W. Something I could not get my hands on.

I have gotten to test my card in a friend’s computer, it works fine there. Apparently, my Direct X is corrupted, causing the video card problems. A reinstall of Windows (curse them!) will fix that. Something like this just had to step to confuse me even more.

More importantly, he also gave me a pcie card that does not draw such an amount of power from the PSU, so I can test a cheap PS I have laying around to see if it works.

I am now going to try the following things:
– Run BIOS default settings with new Graphics card, see if problem persists (from what I’ve seen so far, it does, got a sudden reboot yesterday, but need some more testing).
– Run BIOS default setting with other PSU, see if problem persists –> yes, new MB, no, new PSU.

My coworker thinks it’s the MB, because it works (longer) with modified BIOS. My friend thinks it’s the PS, because the double boot thing (the hickup when first starting) is an indication of a PSU that is slowly dying but not quite dead yet. Both also work in IT, and both know their stuff (better at hardware than me, I’m actually a programmer, however, since I work in a small business, I’m now a bit of everything).

That was again a longer post than I anticipated, thanks to everyone who helped me. Feel free to post any additional comments or things to watch out for, will post back one more time with the final result.

[Author]

Replies