Any PC hardware experts?

Koupe

Senior Retro Guru
Feedback
View
So, until today my old(ish) dual-core PC (with boggo onboard graphics) was running faultlessly. This afternoon I had to move my router to another room and switched from a wired ethernet connection to a wireless card. Connected fine, but now it crashes 5-10 minutes after booting up - the screen freezes, then changes to tesselated pixels and remains that way.
If I take the top of the case off, it runs fine - until I activate the wireless connection, then crashes shortly afterwards. If I leave the case closed, it crashes after 5-10 minutes regardless of what I try and do. :x
I cannot figure out how the hell a network card can be causing such a massive problem (presumably overheating?). The same card has been used in my OH's antique small-form PC for months without causing any issues whatsoever (in fact that's the set-up I'm being forced to use again now). I'd rather not spend ££s on a 20-metre network cable unless I absolutely have to... any ideas?
Much gratitude in advance :oops: :)
 
A WiFi card doesn't generate enough heat to overheat your graphics chip. How does the PC run if you take the card out completely?
No matter how small the chance is, it COULD be possible that some other component started to fail just as you inserted the card. We also need to make sure that static electricity didn't damage the motherboard or card.

As such, I would suggest re-inserting the card into the OH's PC and checking that it works there. Also make sure that your PC works flawlessly without the card.
If both these tests pass, at least we know that the problem doesn't lie with the components themselves.
 
Sounds to me like something is overheating...

Corrupt gfx on screen would usually point to the gfx card (or chip if onboard)

Is the wireless card internal (PCI or PCIe card)? - it may be as simple as it's blocking airflow in the case, cleaning out the dust bunnies from the fans (or replacing them) or adding another fan if there's room might cure this.

The only other possibilities I can think of is that you were running close to the max output of your PSU and the extra card tipped it over the edge (very possible with some of the el-cheapo PSU's fitted in PC's these days) or there's dirt in the connector causing data corruption on the PCI/PCIe bus.

running fine for a while before shutdown does suggest overheating, try running this to eliminate CPU overheating as the culprit: http://www.alcpu.com/CoreTemp/ (you can switch logging on and it'll save a file with your temps in it that you can view after the shutdown.

the equivalent info on your graphics card can be obtained with this: http://www.techpowerup.com/gpuz/
 
O/S? Exact make and model of system?

First thing to do is remove anything you have recently added.

If it runs stable after doing so, at least you know where the issue is.

How big is the wireless card?

I can't see it causing overheating, nor current draw sufficient to crash the machine.

Try disconnecting power from any secondary non boot drives if you think it might be power related.

Do you have a graphics card you can insert to see if is happier?
 
Thanks for the responses!

R_B: I've already replaced the card in the older PC (I'm using it now) so I know the card is functioning OK. In order to properly check my own PC without the card, I'll have to set it up in another room, which will have to wait until tomorrow.

xerxes: I'll try and investigate the temp situation tomorrow - I'm using Linux so I guess hwmonitor won't work, but I'll find something similar. I know the temp sensors are working, as the CPU fan speeds up noticeably when I replace the case cover.

Elysarian: Yep, it's an internal card, and I had considered airflow problems - there's only one fan mounted on the actual case, so I might try and improve that. I gave the PC a thorough clean with an air-duster last week, so dust/dirt shouldn't be an issue. I hadn't thought about problems with the PSU - unfortunately I can't splash out on a decent one at present, so I'll just hope that isn't the issue!

hf: I can't recall the make/model, but it's an Athlon dual-core so not quite prehistoric. I'm running #! (stripped-down Debian-based Linux distro) so there's no problems with resource-draining. Incidentally, the exact same OS is on the older PC which I'm now on, so driver issues don't figure. No other boot drives, and I don't have a graphics card to hand - all I can do is reconnect via cable tomorrow to see if that resolves it.
 
This really does sound like an airflow problem to me. Quite small differences can produce very different results if strength of airflow is marginal.
 
Could the graphics card be on the way out? This is how my old Dell started to die, perhaps swapping things around opened up a soldered joint just enough to cause problems - to give you an example I used to have to take the back off my laptop and blast it with a hairdryer to get it to boot, after a while the fans would kick in and it would die again. The opposite it true with yours but it was the heating/cooling cycle that buggered the solder...
 
I would guess that the extra draw on the PSU is not allowing the CPU fan to run fast enough to cool the CPU when needed, causing it to overhead and shut down.

After it shuts down how quickly will it turn back on again?

Or it could just simply be power related and nothing to do with heat.

You should be able to check the CPU temp in the bios after the system crashes, if it is close to the upper limit then its overheating.
 
Back
Top