PDA

View Full Version : Bring my PC back to life! Power issues?



Justin
16-01-2011, 05:52 PM
I'm trying to resuscitate an old PC of mine. Every time I switch on it starts to boot and then goes off. When I subseqently switch on it stays on for 2 seconds and then powers off again.

I thought it was a PSU problem and have connected to another unit but problem still occurs. Checked the front panel power button and appears to be connected ok, nothing loose.

Pulling my hair out now, removed memory but makes no difference.

Anyone any ideas?

Paul
16-01-2011, 06:16 PM
how far it goes when booting?

Kaz
16-01-2011, 06:19 PM
Have you re-seated the CPU?

Justin
16-01-2011, 06:28 PM
Trying it now, it starts up for just a second and then powers off. Keeping the button held in it starts/stops over and over. I'll check the CPU now. I did re-seat the heatsink after applying new paste.

Kaz
16-01-2011, 06:36 PM
You've maybe put too much paste on the heatsink, which means the CPU will overheat as the paste acts as a barrier, hence why it's shutting itself down again

Justin
16-01-2011, 06:44 PM
Just re-seated cpu. I only put a small amount of paste on, always the size of a grain of rice and ensure it's spread evenly. I don't think it's a heating issue as it's powering off almost immediately. Every now and then it gets to the boot stage and then powers off.

Kaz
16-01-2011, 06:58 PM
Try stripping everything out of the case and build it on a table, short out the power pins to get it to boot and see how you get on from there.

It's easier to fault find when it's on a table instead of in a case

Justin
16-01-2011, 07:02 PM
I've been after a test chassis for a while but not easy to come by, never in stock. Might build something myself.

http://www.kustompcs.co.uk/acatalog/Kustom_PCs_Shop_Lubic___Test_Cases_105.html

Justin
16-01-2011, 07:25 PM
OK, might be getting somewhere. I removed the Graphics Card and rebooted. Obv. the PC halts with an error beep because it can't find the card but power stays on. Numerous reboots and power stays on. Tried a graphics card from my working PC and same error happens, boots/powers off immediately and so on.

I'm assuming from this that the AGP graphics slot may be faulty. Maybe the mobo has started to pack up on me.

Kaz
16-01-2011, 07:55 PM
Could be a faulty slot, but more than likely it's that the PSU isn't big enough to run everything

Justin
16-01-2011, 08:39 PM
Faulty slot appears most likely. PSU is more than capable for what's running, overkill if anything :-)

Justin
16-01-2011, 10:33 PM
Justin wondering how I can check if it is the AGP slot that's faulty, wonder if something like this (http://cgi.ebay.co.uk/PC-MOTHERBOARD-ANALYZER-DIAGNOSTIC-POST-TEST-CARD-/330505968494?pt=UK_Computing_DesktopComponents_RL&hash=item4cf3b0db6e) would help?

This (http://cgi.ebay.co.uk/Desktop-PCI-E-PCI-AGP-Slot-Memory-Testing-Board-Tester-/310286645601?pt=LH_DefaultDomain_0&hash=item483e867961) also looks very interesting.

JSR
17-01-2011, 12:11 AM
How old is this "old" PC? In the old days when AGP first came out, it was optional and you could put the more common (of the time) PCI graphics card in a PCI slot. If you can find one of them these days (not to be confused with today's PCI Express cards), it might be an option.

Justin
17-01-2011, 09:12 PM
It's a few years now but nothing too antiquated ;-) I'm also an idiot btw....it IS a PCI Express slot, not AGP...DOH!

bikertec
26-02-2011, 09:58 PM
Could be the reset switch shorting disconnect little wire jumpers on MB. Could be as simple as that.

Justin
26-02-2011, 10:31 PM
Nope, pretty sure its the PCI Express slot, going to install another mobo and see what happens :-)

PetPals
04-05-2011, 03:48 PM
Not sure if this helps... I had a similar problem a few years ago with an old PC that had not been used for 4 years. I found a battery inside - looked like a large watch battery. I replaced this and the PC worked again.

Matt Quinn
04-05-2011, 05:50 PM
First of all thermal paste does NOT act as a heat barrier! Exactly the opposite as its whole purpose in life is to AID thermal coupling to the heat sink!! If the clamps to your heatsink are working properly all excess will be squeezed out; which can cause issues if it gets onto I.C. contacts...

The number of students I see who try to build PCs and wind up frying a CPU because of lack of thermal paste is bewildering!

Firstly check all the fans are connected and running - many MOBOS will shut down if they don't get a 'tach' signal of some kind from the CPU fan. The other obvious thing is dust. Get an airduster Make sure everything is clean.

Remember; PC components are static-sensitive and you need to be grounded as does the surface you're working on. DON'T be tempted to take a feather duster, nylon pantbrush or similar to the components to clean them! - And yes; I've known various assorted Muppets do this! Killing their machine stone dead in the process!

- A airduster or airline from a compressor to blow out the dust! Any dirty contacts need a wipe down on an static-safe surface with IPA ; again, make sure you're grounded when doing this. Fingers of slots can be cleaned with special cleaning strips for the purpose or relay cleaning strips...

Next remove ALL peripherals... Disconnect ALL the drives... Power and data cables. Re-install your AGP card but and see what happens. If IT has a fan make sure that it is working. There's a VERY good chance the card IS the culprit.

If it stays on and goes into the BIOS you have progress...

If at all possible swap out your drive data cables for new ones just to eliminate them ...

Next, going through a careful cycle of re-connecting each drive's power supply in turn... powering down, next connector etc...

If none of them reproduce the fault (dead short/ PSU shutting down as opposed to MOBO) re-connect your Master hard drives data cable, then, sequentially any slaves, then any optical drives... Shutting down before connecting anything of course...


- And yes; some BIOS batteries will cause a shut down if completely dead. Often removing them completely gets the machine going temporarily. If REALLY old/dead they've been known to leak onto the MOBO causing issues... I've only ever seen this with the CR2032 type button cells though.

Matt Quinn
04-05-2011, 06:07 PM
Doh! Just looked at the O/P date on this thread and realised the problem was probably solved back in February!

Justin
04-05-2011, 07:32 PM
Appreciate you advice anyway Matt! ;-) I did finally get around to put another mobo in and this sorted the problem. It was the PCI Express slot that appeared to be faulty, not something I've seen before.

Matt Quinn
05-05-2011, 10:36 PM
I've certainly known the video cards to keel over... Sometimes taking out the mobo with them! But that's usually a 'dead set' scenario....

They often draw very significant amounts of power through what to me seem very small contact areas. Slots can simply oxidise over time. Radiospares used to do contact cleaning strips meant for relays that could help; but getting into these things properly can be an issue; particularly the higher density ones...

Cutting your losses and just replacing the mobo was obviously the way to go... Just out of curiosity... Did the video car work afterwards or was that replaced too?

Justin
05-05-2011, 11:15 PM
The video card was absolutely fine :-) I just re-cased everything (because it was in a full tower) replaced the mobo and now the kids play Red Alert quit happily...hour after hour....

Matt Quinn
05-05-2011, 11:50 PM
<LOL> Sounds like my place... Every time the edit suite gets upgraded there is a trickle down effect. - If I thought the cat could type she could easily have her own graphics station ;-) As for the 'parts graveyard' :D