Kitz Forum

Computers & Hardware => PC Hardware => Topic started by: silversurfer44 on November 08, 2011, 08:24:28 PM

Title: PC strange behaviour.
Post by: silversurfer44 on November 08, 2011, 08:24:28 PM
Overview: After shutting down for the night and starting up in the morning the PC will suddenly shut off completely after a couple of minutes. Not after a given period and not after a particular event.
I have temperature monitors displayed on the screen permanently and they show no abnormalities. After a shut down when I press the start button nothing happens for approximately 5-6 minutes and then the computer will start up and stay running for the rest of the day.
I have checked the power lead, the USB leads and any other thing plugged in. In fact I have booted up with each item disconnected all to no avail.
I have run an extensive RAM test overnight with no faults shown. What did happen was I closed down to do a cold boot after the test and the PC would not start immediately.

The specs of the pc:
CPU                       AMD Athlon ll X2 250 Socket AM3(300GHz)
RAM                       4Gig Dual-DDR
Motherboard           Asus M4N 78-AM
Graphics                512 Nvidia Geforce 8400GS
Soundcard             Onboard 5.1 High Def

Anymore info required I will gladly post.

Apart from leaving the machine on 24/7 I am at a loss as to what it could be. Suggestions please?
I have just done an update to my Mageia Linux operating system and when I did a cold boot I pushed the start button and nothing happend for a few minutes. Just as on a morning.

Thank you.
Title: Re: PC strange behaviour.
Post by: sevenlayermuddle on November 08, 2011, 09:08:10 PM
There's no guarantee, but personally I'd say that behaviour would be strongly suggestive of a failing PSU. 

If you have an accurate DVM, you could try measuring the PSU voltages to see if any are on the edge of (or beyond) their normal tolerance.
Title: Re: PC strange behaviour.
Post by: silversurfer44 on November 08, 2011, 09:56:58 PM
Thank you 7LM you have just confirmed my suspicions. I do have a spare unit which I will try.
Title: Re: PC strange behaviour.
Post by: exo on November 08, 2011, 10:14:26 PM
A failing PSU would also be my first best guess. It is under most stress when cold and is when problems are likely to show themselves.

exo
Title: Re: PC strange behaviour.
Post by: HPsauce on November 08, 2011, 10:17:34 PM
+1 for PSU.
It's relatively so cheap and easy to replace that you should just do it anyway as step 1 of diagnosis.
Title: Re: PC strange behaviour.
Post by: burakkucat on November 08, 2011, 10:39:02 PM
And if it doesn't turn out to be the PSU, my second choice would be the power button or the connectors between it and the leads to the PSU.  :)
Title: Re: PC strange behaviour.
Post by: silversurfer44 on November 09, 2011, 07:31:36 AM
Thank you everyone.
I have put my spare unit in this morning and she fired up without any problem.
Another remedy in the fault finding kit. I have not come across that before as I usually leave the pc on 24/7, but being a bit planet conscientious I started switching off overnight. I do have one old slow donkey that runs 24/7 with databases & a web server on it.

The only problem I have now is the unit that I have put in has two fans and they are noisier than the one that is in the failing unit. I'll get used to the noise level.

Thanks again.
Title: Re: PC strange behaviour.
Post by: sevenlayermuddle on November 09, 2011, 09:12:15 AM
Glad to hear it SS.

I'm guessing just a little, but I'd bet a pound to a penny that if you peer into the old PSU you'll find the electrolytic capacitors are bulging.  It's the one component the electronics industry has failed to get right, despite the fact they've been making them for over half a century.

In switch mode power supplies, the circuits stop working properly if the parameters (ESR in particular) of the capacitors drift too far out of tolerance, often manifested as lazy start-up and easy overloading.  Failure of other components tends (usually) to lead to a more absolute failure.  Even then, the failure of other components can often be traced back to increased stress caused by a failed cap' or two.

Since this PSU's not completely dead, replacement of these cap's (a few tens of pence each) is probably all that is needed to restore health, though with labour costs it's unlikely to be a cost-effective option unless there's any special reason you want to keep it.   You may also struggle to find anybody willing do do the work, unless you can DIY.   On the brighter side, even though a capacitor may be well within it's temperature rating, it's life doubles for every 10 degrees further reduction in temperature so these noisier fans may be doing more good than is apparent.

Cap's are an ongoing gripe of mine, just thought I'd get it off my chest  >:(
Title: Re: PC strange behaviour.
Post by: roseway on November 09, 2011, 10:27:14 AM
You're quite right about electrolytics, 7LM, but good cooling doesn't have to equate with high noise. The PSUs with large slower-moving fans in the bottom of the unit are generally quieter while still moving just as much air through the unit. I can't stand noisy computers, so that sort of PSU is top of my shopping list when I'm putting together a new PC.
Title: Re: PC strange behaviour.
Post by: sevenlayermuddle on November 09, 2011, 11:19:57 AM
but good cooling doesn't have to equate with high noise.

Agreed, I've one PSU with a nice big fan that keeps everything cool and is quite literally inaudible.  The only way to convince yourself it's spinning is to shine a torch through the grill.

I'd still though argue that the other way around...  high noise usually equates with good cooling.  Some of the industrial server systems that I've worked with, for which cooling is paramount, but noise not a consideration, can literally sound a bit like a jet aircraft on full take-off thrust.
Title: Re: PC strange behaviour.
Post by: silversurfer44 on November 09, 2011, 11:48:49 AM
I can certainly agree about noisy fans in a domestic situation.
The fan for CPU cooling is as big as the fan in the PSU but runs in quite mode and does the job. On boot up it does sound like a jet taking off, but it's only for a second or so.
I will have a nosey inside the duff power supply when I have nothing better to do, one of those tuit jobs.

No shut downs so far, will find out tomorrow on a cold boot.
Title: Re: PC strange behaviour.
Post by: silversurfer44 on November 10, 2011, 05:54:32 AM
Machine started up fine this morning. I think it went through the bios post quicker as well.
I had a look inside the old power supply unit and sure enough there is a cap looking like the top of Mount Etna. I would need to de-solder it to find out its value, which I may do. If it's repairable it will come in as a spare.

Thanks once again for the quick and accurate responses.
Title: Re: PC strange behaviour.
Post by: sevenlayermuddle on November 10, 2011, 10:10:10 AM
If you decide to repair it, as you may know, power supplies usually require very low ESR specifications, shops like maplin are unlikely to have suitable components, or even to publish the spec's in enough detail.  Panasonic capacitors are highly regarded, particularly the 'FM' series, which can be ordered from RS ( http://www.rswww.com ). 

And take a close look at the other capacitors too ... multiple failures are very common.

- 7LM