Shutting down without warning, driving me crazy!!

[snape]

sojrner, I thought about what you said and decided to take my power supply out of the case last night and leave everything else in to eliminate some sort of short to do with the power supply and case.

I realised when I looked that two of the metal connectors on the end of the power cables of my modular power supply were touching the metal guard of the case's cieling fan. Could this have been causing static build up as you mentioned...?

When I ran the memtest with the psu out of the box it didn't shutdown at all. I ran the test twice for a total of 1.5hrs then tried the DVD player for about an hour and it still didn't shutdown. Thats all I had time to test last night but I have left it running all day so hopefully when I get home later it will still be on!!

My hopes are high but it could just have been a fluke last night. Fingers crossed.

Thanks for all the great help everyone!

 

[jap0nes]

hmmmmm could be. basically, what you got was that you were injecting current at your motherboard through the grounding. Your motherboard makes contact with your case, which is supposed to be grounded. If your grounding does not work, and there's current circulating in your case/motherboard, you may expect any sort of malfunctioning. The weird thing is that it happens only after some time...

good luck man

 

[Fox_granit]

I think that looks to be the problem, I've had something similar happen to me. I had a cable that would ground out after a few minutes becuase of heat expansion. After the cable would heat up, it would touch the steel edge of the case and short out the power supply. It would take a while for it to cool off, but once it did, the machine would start up again just fine. Took me a month to find it. Good luck and hope the PS short was the problem........

 

[snape]

Well, I hope so....Ill let you know what happens

 

[sojrner]

rockin man, that is what i thought. I have a case I built that is metal and plastic... (made the case from a tool box) and while the mobo and most components are on the metal grounded part, my actual power switch is inside the plastic and I was getting occasional shorts on it just like you were describing. Hopefully that was your problem.

The reason (i think...) that is takes a while before it fails is that it is building up a charge and once the capacity of the case is reached it discharges and shorts out somewhere. it would do that if you are close to the switch, but if not then it takes a bit longer and shorts out something on your mobo.

I could be wrong on all of that (I have been before ) but if you are running now then grounding (in some form) was your issue. rock on man.

 

[Wulf]

I've ditched several Abit boards for the ASUS or Intel boards due to Abit's caps problems. Unexpected shutdown is one of several characteristics coming from the faulty capacitors.

Personally, I've desoldered and replaced the faulty capacitors on the out-of-warranty boards. I have a donated Systemax with several blown caps on it's Gigabyte board, the system is still running dependably for almost two years since.

 

[sojrner]

that is an old problem that is not only abit, but many board makers both cheap and high-end that used the caps w/ bad electrolyte that resulted in them not regulating correctly and either bursting or just failing. Abit was actually one of the few companies that owned up to it and replaced boards that had them. I give them props for that, and depending on how far out of warranty the board was you prb would have gotten it replaced. The caps were even in items like ps2's and vcrs... it was widespread a few years ago. afaik Asus was one of the few that did not use those caps and that resulted in their better rep over some other makers. jmo though. Hopefully snape has his prb fixed.

 

[jap0nes]

as far as i remember, this was a problem with a shipment from a japanese electronic components maker... every company who bought from this shipment was affected

 

[sojrner]

yup, some dude stole the electrolyte formula (from a company he used to work for i think) but forgot a key ingredient. He then made a ton of 'em w/ that flawed formula and sold 'em on the cheap. Once it was found out of course, there was a ton of product out there and the dude's company had made a bunch of cash... very few companies took the blame but abit was one of them. I had a soltek kt266A mobo w/ the failed caps and they told me to essentially piss-off, and I got no help (just a few months out of warranty) but a friend of mine got his abit replaced and it was almost a year over warranty.

 

[sailer]

You seem to have tried a lot of things, but get the same result. When you use the DVD player, the computer shuts down, then won't restart for a period of time. Two things I see; one, that the DVD player has a problem and that triggers the shutdown, or two, the power supply has a problem. My guess is that its a power supply problem. The power supply is big enough technically, but it may not be able to stand the load, therefore has an internal problem. This seems to be verified by your statment that the computer won't restart for half an hour or so. The only thing I can see to cause that is the power supply. Something in it is overheating at too low a temperature or too low a power draw, and it is not resetting until it gets very cool.

Just a guess, but after eliminating all the other things, that's what's left. Sure, there's possibly a motherboard problem, but it doesn't sound like it at this point.

 

[snape]

Damn. Its shutdown again. That was when I had put the psu back in the case and left it to run all day. When I got home it was off.

As it had seemed to be working again last night when I had the psu out of the case I thought I had figured it out. Because two of the metal ends of the psu's power cables had been touching the case's ceiling fan's metal guard that maybe it had been causing some kind of short.

I removed the metal guard and put the psu back in the case to test (using the dvd software again) and left it to test while I was at work and as I just said, I got home and it had shutdown again.

Again I tried another test with the psu outside the case and after an hour or so the computer froze.

That is another symton that I have not mentioned yet; it freezes sometimes and then eventually shuts itself down if i leave it frozen for a long time.

Is the freezing accosiated with the same problem? Probably.

Sailer, in response to what you said; it seems to have a problem with all dvd sofware and other processes but dvd software seems to reliably shutdown the computer from what I have noticed. It happens when I run from the dvd drive and straight from the hdd...

Also I have tried my freinds psu who has the exact same one as me and that crashed in about 5mins so I am reluctant to think it might be my specific psu but as you say maybe the spec is not right. Doubtfull though as it is a high end psu.

I have the option to take my board to Abit and change the mobo as i have spoken to them but I have a nagging feeling it will have the same problem.

It really seems like something is overheating as Im sure lots would agree.

Probably the reason it was working better when I had the psu out of the case was because the whole system was soo much cooler without the redhot psu in the case and it just took whatever is over heating longer to overheat.

Just to summarise as this thread is getting quite long so far I have tried:

swapping out the ram
swapping out the graphics card
swapping out the psu
swapping out the hdd

buying a surgemaster power extension

replacing the mobo when I first bought it (6 months ago or so)

running the computer with the psu out of the case

stripping down the machine (inc. unmounting the mobo from the tray) and rebuilding to eliminate the shorting from the psu

re-installed windows

temp check right now whilst online and everything is back in the case with side on: CPU: 44 SYS: 34 PWM: 55

-----------

And the symtons are:

shutdown with out any warning whilst using the pc to run non-intensive software - especially dvd software for some reason

freeze with out any warning whilst using the pc to run non-intensive software - especially dvd software for some reason

shutdown whilst running memtest in dos - which eliminates windows

-----------

My spec is as follows:

MOTHERBOARD - Abit N18 SLI
CPU - Pentium D Dual Core 3.0GHz
CPU fan - Coolermaster Hyper48 Cooler
COMPOUND - Arctic Silver 5
POWER SUPPLY - Hyper HPU-4B580 Type R 580W Modular ATX2.2 PSU
MEMORY - 4 x 512Mb Corsair DDR2 XMS2-5400 675Mhz memory
GRAPHICS CARD- NVIDIA GeForce 6800GT Series GPU
BOOT DISK - SATA Western Digital Raptor 36Gig (WDC WD360GD-00FLA2)
OTHER BOOT DISKS - 2 x Western Digital Caviar SE16 250GB 2500KS SATA-II
CASE - Lian-Li PC-60 with 120mm front fan, exhaust blower fan + 2 exhaust bearing fans.
FLOPPY -1 x Floppy Drive
DVD RW - 1 x Sony DW-U14A
OP SYSTEM -Windows XP Pro SP2

Does anybody else have any new suggestions?

Thanks

Spencer

 

[sojrner]

well, first I would say that if abit is willing to give you a new one I would get it, that would eliminate it as an issue... you are right that it will prb have the same issue, but it for sure removes it as a possibility if you get a new one.

second: I still am wondering about the grounding on the wall outlet in your home. have you tried another outlet like I suggested earlier? Another option is to try taking your system to a friends and running tests there, or even better is to try it in another case (friends?) while over there. I realize this all sounds kinda weird, but you may be on to something when you had it all out of the case. Perhaps a combo of grounding and poor cooling is causing it all?

Lastly have you tried running a game on a loop yet? I know that the dvd app kills it reliably, but does anything else? a game is intensive and should do the same thing. If the game does not then perhaps it is the dvd app(s) having a conflict.

I am just offering suggestions that are logical steps I would take when I do not know what is going on... These are just ideas that would systematically eliminate (for sure) what is not the problem. Eventually you will find out what is the issue.

Some may seem redundant and you may be certain that it is not the issue, but confirmation makes it a sure thing. It is easy to think something is not it and miss it when it really is the problem. I have done it many times. Hope some (any) of this helps man, good luck.

 

[dokk]

sojrner, I thought about what you said and decided to take my power supply out of the case last night and leave everything else in to eliminate some sort of short to do with the power supply and case.

I realised when I looked that two of the metal connectors on the end of the power cables of my modular power supply were touching the metal guard of the case's cieling fan. Could this have been causing static build up as you mentioned...?

When I ran the memtest with the psu out of the box it didn't shutdown at all. I ran the test twice for a total of 1.5hrs then tried the DVD player for about an hour and it still didn't shutdown. Thats all I had time to test last night but I have left it running all day so hopefully when I get home later it will still be on!!

My hopes are high but it could just have been a fluke last night. Fingers crossed.

Thanks for all the great help everyone!

HI there: Got a dumb question here after reading thru to the bitter end.
How are you mounting [installing/fastening it to the case/backplane] your mobo..Had a similar problem my self ,worked fine outside the case but,put it in and no go,would not stay lit for more than a few minutes,turned out that one of the mobo's fasteners was too short and was bending the board ever so slightly ????,you shoulda heard the new four letter tech terms that day... Also do you have any way to monitor the DC voltages from your PSU,maybe something on the mobo is getting too much power,I suppose you have checked/smelt the mobo for any untoward overheating/discolouration????
Keep us posted.Thus we can all learn.......

 

[snape]

Hi sojrner,

Sorry I forgot to mention that I had taken my pc to work about a month ago and it was shutting down there too. I have assumed this would rule out a problem with the grounding at the wall outlet.

Yes I agree I probably should change the mobo whilst ive got the chance just to eliminate it for good.

I havent got any games to test on it, could you suggest one that is easy to get hold of...or should I just download the free version of 3d mark?

I also have an intersting update having run another test...on saturday I unscrewed the psu and pulled it halfway out the back of its bay in the case and continued to use the computer as I normally would. Since I have done that it has not crashed once, and it has been on now for almost 24hrs since my last voluntary restart.

During that time I have downloaded some large files over the net, ran defrag on the boot drive (raptor 36gig), and this morning played a dvd successfully for 1h30mins.

To me this suggests two things, either the psu is touching something it shouldn't when it is all the way in, or the psu is overheating a component it is near to on the mobo when it is in.

I really cant see that is touching something it shouldn't but then I am no expert and I may miss it if it is.

Here is a link to my case just in case it helps: http://www.overclockers.co.uk/acatalog/Online_Catalogue_Lian_Li_Cases_225.html and it is the Lian-Li PC-60 Silver Aluminium Midi-Tower Case (No PSU) (CA-041-LL)

Thanks

Spence

 

[sojrner]

3dmark is ok, but if you can spend the little bit of money you can pick up ut2004 for cheap right now. It is a great shooter game, and you can just set it to a specator match and it will loop indefinitely on that until you stop it. It scales the best of any game out there. Personally I have run it on a athlonXP 1500 w/ a gforce3 ti200 and up to my current system. w/ the setting all the way up it hits the video card very well, with them all the way down with lots of bots (comp controlled players) it is a great cpu test. You should find it for around $20 now. (get the dvd version if you can, otherwise it is 5 cds!)

ya, you are right... if you ran at work that would eliminate the ground at your place being an issue.

With it running w/ the psu slid out you are right that it is either cooling or (still) grounding. My $ is still on grounding (but dont rule out cooling yet)

It could be that one of the mount screws is grounded badly inside the psu, like it is touching something inside. If you try sliding the psu back in but not mounting the screws and it still works then it is prb not cooling and more likely the mounting. If it fails inside there (w/o the mount screws) then it may be cooling. If cooling then you need better airflow and should test it w/ the side open and a table fan blowing on the mobo/psu. If it works then you need to re-do the airflow logic in your case. (remove/rework cable/component layout)

If it worked w/o the screws then the mounting is bad. I.E. the psu mounts on the case/psu are not right. I am not sure here what to do, as I have not worked on that case or psu. (nor that problem in general) Perhaps someone else here?

Otherwise it could be like dokk said and the mobo is mounted w/ either the mount flexing the pcb or a mount grounding the board wrong. Not sure why that would be fixed by moving the psu out but at this point anything is possible! lol

 

[snape]

I took my pc to Abit yesterday where they very kindly helped me work out what was going wrong.

It works out that the hyper type R psu is sending fluctuating voltages which is causing the mobo to work a lot harder and ultimately overheat.

When we tried a different psu it was a lot cooler and I could see in the bios that the voltage readings where a lot more stable. It also seemed to handle the dvd test for over an hour without a problem.

Unfortunately I could not leave my computer at Abit for longer testing but they were fairly confident this was the problem.

When I swapped the psu out with my mates it did crash within about 5mins but then it was the same model psu.

The confusing thing is my mate has got the same psu, case and mobo as me and his never has this problem, mind you he has got a 2.8Gig processor whereas I have a 3.0Gig and this will no doubt stress the board more.

My next action is to change the psu which overclockers have kindly agreed to do for me without charge.

Sorjner, I will test using the game on loop failing the psu swap. Thanks for the suggestion.

Spence

 

[sojrner]

that rocks, and only gives more respect to abit as a company in my eyes that they did that for you. Looks like most ppl on this thread were right about the psu being the issue, and I am sure that many would agree when I say that the psu is the most overlooked component on a computer. Whenever possible spend $ on it and you will not be dissapointed. This does not just mean for more wattage, but for quality brands.

small inconsistencies in mobo components or cooler ambient temps could attribute to your buddies same system holding up better. I would still let him know what you found out and get him to change psu's as well or eventually his will dump on him too.

Glad to see all is getting better man, rock on.

 

[lyssmarie92390]

Hey, my computer does the same thing (I have a Toshiba Lap Top).. it has two cooling fans on the bottom and one burnt out, if yours has cooling fans, check the bottom and see if the fans are spinning, if not.. there's your problem. If the fan is not spinning, you need to go and get that fixed, otherwise it'll keep on doing it to you.. Sorry if this doesnt help, but it seems like the only reason why it would do that.
-Lyss