HELP!! Riddle me this! Can anyone even *theorise* how this could happen? [UPDATED]

DeVore

CAN ANYONE BEGIN TO EXPLAIN THIS HOBO-CURSED PoS COMPUTERS BEHAVIOUR? I've never seen anything like this in 30 years of tinkering with PC's

Ok, here's my tale of woe. I'll try to keep it brief.

My machine is a small mATX board, with a GE-Force 780 and an i5. Pretty much a standard build, no over clocking, nothing weird.

I went to put a second monitor in one day, machine was fine worked for a little while so I made lunch. Came back the machine had turned itself off. Refused to boot. Everything got power but the fans in the case. (ie: CD drive got power, graphics fans spun, but no visuals).
The motherboard seemed dead.

The CPU had been overheating before this (and cutting to black-screen but rebooting). Not too often and the heat wasn't too high but still, in heavy games it might do that. We figured it might be the cause. An autopsy revealed that the CPU heat sink cement had dried up. We figured case closed, ordered a new MOBO, installed it. Dav did a cracking job on the heat sink and all the temp numbers returned to normal. All working great.

Jumped into a heavy 3d game the next day and got a black screen. Reboot. Comes back up and works again for several days, including several long sessions in the 3D game. I figure it might just be coincidence. All the temps are staying low so maybe its nothing.

But it happens again. And restarts... so we plan to maybe transplant the whole machine into a bigger case (mine is quite small and packed).
The day before we do, it dies. Just like before, everything getting power, everything good, just no signal to the screens even from the onboard graphics.

THIS IS WHERE WE LOSE CABIN PRESSURE
I leave it 30 minutes and try again, nothing.
I figure we have blown the motherboard again, like the last time. Some sort of short circuit or power spike from the house or something.
Its dead, its very dead and its not coming back. Cursing my luck I go to bed.

Next morning I get up, boot the machine on the off chance it will boot and lo and behold. It does! Works fine!

3 days later I shut down windows gracefully, leave the machine. Go work in my workshop, come back and it wont boot. Absolutely nothing changed in the mean time. No interaction with the machine at all. Cant have overheated since I wasn't doing anything with it for a start and even the session beforehand I was only reading mail etc.

That evening, it boots again.


So, the TL;DR, we have a motherboard which has genuinely fried at the start, giving very specific symptoms. Replaced that. Same symptoms again from the new mother board, only its intermittent.
Its not temp. Its not the mobo. We tested the Ram in another machine.


I'm genuinely at a loss to diagnose this and it feels like I'm sitting under Damocles Sword here...:)

HELP!?

challengemaster

Sounds like your power supply has been the issue all along.

Also I've no idea why you even replaced the first motherboard. If I saw dried up thermal compound and the CPU had a history of overheating, I'd be expecting the CPU to have died (if thermal shutdown protection was switched off or somesuch), not the mobo.

If you really want to test it you can check the voltages/amps across the rails using a multi-tester. Or if you have a spare PSU lying around, just swap it out and see what happens. That would have been my first port of call with the symptoms described.

Myrddin

PSU would be my bet too

TerrorFirmer

Yeah, CPU overheating would result in abrupt shutdowns under heavy load, but refusing to cold boot is going to be either the mobo or PSU...and since you already replaced the board it has to be the PSU!

DeVore

The reason we thought mobo was because no matter what we did we couldn't get a VGA signal from its on board and it wouldn't post.

I think I agree with you about PSU being the next suspect and I'll replace it and see if we get any further symptoms. Its a 600W PSU and its powered everything else so I guess we thought it was fine. I still cant explain why it would decided not to boot after I was in the workshop or why leaving it over night would matter but I'll grasp at any straws going!

challengemaster

Fair enough. It's not unheard of for a dying/faulty PSU to take down other components when it's failing. Not sure what odds you'd get in the bookies but I wouldn't at all be surprised if it killed the first motherboard which is why you couldn't get it to POST.

DeVore

I moved the other mystery into its own thread, here: http://www.boards.ie/vbulletin/showthread.php?p=98671106#post98671106

We're going to replace the PSU and see if we continue to get issues...

Thanks for all the help, I'll be back if there is no improvement after the PSU transplant.

strelok

i had a weird issue with booting before, i forget the exact details of it but there was little rhyme or reason to it. I ended up dropping over a hundred on a new psu which didn't make any difference whatsoever

updated the bios on a whim and that fixed it completely.

DeVore

challengemaster wrote: »

Fair enough. It's not unheard of for a dying/faulty PSU to take down other components when it's failing. Not sure what odds you'd get in the bookies but I wouldn't at all be surprised if it killed the first motherboard which is why you couldn't get it to POST.

The new board doesn't make any noise when it posts either, so I cant be sure it WASNT posting, but the machine steadfastly refused to show anything on the monitor regardless of what display port we used so I'm pretty sure it was dead. Might just have been pining for the fjords but I've sent it back to H.E. anyway so that's that

Thanks for all the help lads and lassies, much appreciated.

DeVore

Got a new PSU... hasn't made a difference. Its a 650 where as the last one was 600 and if anything, the problem (having gone away for a few days) is now back and while its hard to tell, it seems more prevalent than before!

Gonna take out the Graphics card and run if off the On Board GPU for a while, see if its something to do with the card. Also gonna switch out the memory cos maybe its that.

Or maybe its ****ing haunted. At this stage I know *I* am.

Cuddlesworth

Get a speaker out of a old machine and attach it to the mobo, to make sure that it isn't posting.

Short the power pins so that its not simply the power button.

Your issues sound like a dead chip(expansion/contraction causing and removing the issue) but if you replaced the mobo, then it should be reporting on the issue using beep codes from the bios.

DeVore

TerrorFirmer wrote: »

Yeah, CPU overheating would result in abrupt shutdowns under heavy load, but refusing to cold boot is going to be either the mobo or PSU...and since you already replaced the board it has to be the PSU!

So now I've replaced both and its *still* happening.

So it has to be either graphics card, memory or chip.

Going to try to update the bios and see if that shakes something loose...

Johnboy1951

we plan to maybe transplant the whole machine into a bigger case

It is not clear if you did so or not ....... but if not I suggest you do ..... it will take the likes of the case connections and switches and the mount points out of the equation.

ED E

OP says they did try another case as the original was cramped.

• Overheating rig
• Mobo changed
• Weird behavior starts

Still sounds to me like it could be the new mobo being a lemon.

Johnboy1951

ED E wrote: »

OP says they did try another case as the original was cramped.

I quoted what was posted which was an intent only ....... did I miss something that said definitely the case was changed?

Here is the quote again ......

But it happens again. And restarts... so we plan to maybe transplant the whole machine into a bigger case (mine is quite small and packed).
The day before we do, it dies. Just like before, everything getting power, everything good, just no signal to the screens even from the onboard graphics.

DeVore

No, we didn't swap the machine into a bigger case. We had INTENDED to but the new case had been left near a radiator and had warped. Which is about the level of luck I'm having with this entire sh*t-sandwich.


So, where are we now:

1. We have replaced both mother board and PSU. Its still doing the same thing but *slightly* differently.
A. When we replaced the Mobo it went away for 3 days but then came back the same as before... that is to say whenever the machine was stressed (big cut scene in games etc) it was prone to completely cut to black screen. It initially would boot itself again and be ok, but within a day to two that turned into having to wait for the next day before it would boot again.

B. when we replace the PSU is worked great! ... for three day. Then it started to cut to black screen even when just browsing the web! Or in fact doing *nothing at all!*. Again, initially it would restart itself but again, within a few days it decided to wait until morning.

Today I read one of the posts above talking about the Bios for the motherboard and decided that it couldn't really hurt to update it. I went to their website and sure enougn they have a warning saying if you are running Win 10 you should update.
I've downloaded the update but due to reasons I really don't want to get into I don't have a USB stick at the moment. (My dog chewed my keys. destroying two USB keys and at the same time arming the alarm on my house, at 1am. Yeah.)

So, while I was looking at the bios settings I changed a couple of things, one being the "emergency stop if Mobo gets too hot" setting and also the fans setting (I had set them to Always On when I thought it was a heat issue, that made things worse imho so I've turned them back to normal). So far (and that was only today I should point out) no further issues.

So, yeah... right now i'm through the looking glass and trying just about anything that might either make this worse or better because its the middle ground that is killing me! Thanks for all the help though, I'm listening and paying attention!

DeVore

No, we didn't swap the machine into a bigger case. We had INTENDED to but the new case had been left near a radiator and had warped. Which is about the level of luck I'm having with this entire sh*t-sandwich.


So, where are we now:

1. We have replaced both mother board and PSU. Its still doing the same thing but *slightly* differently.
A. When we replaced the Mobo it went away for 3 days but then came back the same as before... that is to say whenever the machine was stressed (big cut scene in games etc) it was prone to completely cut to black screen. It initially would boot itself again and be ok, but within a day to two that turned into having to wait for the next day before it would boot again.

B. when we replace the PSU is worked great! ... for three day. Then it started to cut to black screen even when just browsing the web! Or in fact doing *nothing at all!*. So, thats a change from before, it doesn't require the stress in order to cut out now. Again, initially it would restart itself but again, within a few days it decided to wait until morning.

Today I read one of the posts above talking about the Bios for the motherboard and decided that it couldn't really hurt to update it. I went to their website and sure enougn they have a warning saying if you are running Win 10 you should update.
I've downloaded the update but due to reasons I really don't want to get into I don't have a USB stick at the moment. (My dog chewed my keys. destroying two USB keys and at the same time arming the alarm on my house, at 1am. Yeah.)

So, while I was looking at the bios settings I changed a couple of things, one being the "emergency stop if Mobo gets too hot" setting and also the fans setting (I had set them to Always On when I thought it was a heat issue, that made things worse imho so I've turned them back to normal). So far (and that was only today I should point out) no further issues. I'm watching the mobo heat with Speccy and its <31C so that's perfectly fine.

So, yeah... right now i'm through the looking glass and trying just about anything that might either make this worse or better because its the middle ground that is killing me! Thanks for all the help though, I'm listening and paying attention!

jhud

Had issues over a few weeks like you had turned out to be failing memory chip which eventually failed so RMA them and returned ones worked everything perfect now. I also had random reboots non booting at times and no graphics other times.

Johnboy1951

There is one other item that has caused me a problem in the past .....

the connection between the CPU pads and the socket pins.

A 'not quite solid' connection on one CPU caused me similar problems.

I eventually (after several attempts) reseated the CPU and heatsink so that all remained properly connected and no further problems.

I know this is unlikely because you changed the motherboard but thought to mention it while I remembered .......

DeVore

Dav has reseated the CPU a couple of times now and yeah, we've swapped mobo's too so I think its not that but I'll keep it in mind!

Failing memory might be an issue. So far though, having switched things around in the Bios, it seems to be behaving itself but the problem is that it will sometimes do this for days and then fail twice in the one day.

Thanks for the feedback guys and gals, fingers crossed...

Lu Tze

DeVore wrote: »

Dav has reseated the CPU a couple of times now and yeah, we've swapped mobo's too so I think its not that but I'll keep it in mind!

Failing memory might be an issue. So far though, having switched things around in the Bios, it seems to be behaving itself but the problem is that it will sometimes do this for days and then fail twice in the one day.

Thanks for the feedback guys and gals, fingers crossed...

Time to call a priest for an exorcism

challengemaster

I was thinking it might be related to the graphics driver - after seeing what happened with 364.47 causing BSODs and just pure black screens, amongst other issues.

https://www.youtube.com/watch?v=DXowhcfNpr4

However, it seems your problem predates the driver release by about a month....

Lu Tze

Maybe try memtest86 to test the ram, see if it throws up any errors. Its always difficult to narrow down the cause of an issue if it isn't consistently repeating itself

DeVore

Thanks, I'll try that! It looks pretty comprehensive! (If anyone else wants to know where to get it, its here: http://www.memtest86.com/ )

jhud

DeVore wrote: »

Thanks, I'll try that! It looks pretty comprehensive! (If anyone else wants to know where to get it, its here: http://www.memtest86.com/ )

Just to note on this when i had the issues i got the memory to pass 5 times on each stick(one in machine at a time) also moved them to different sockets but some motherboards are set with things like use dimm socket 1 and 3 ore 2 and 4.

When you had none boot were the fans spin up and no beeps and shut off again as that what happened to me it took about 4 weeks for ram chip to fully fail before i got to the end of it.

syklops

DeVore wrote: »

So now I've replaced both and its *still* happening.

So it has to be either graphics card, memory or chip.

Going to try to update the bios and see if that shakes something loose...

You said its got to be either the graphics card, memory or chip so why would you update the BIOS? That only introduces another unknown to the situation.

On a mobo that is suffering from random shutdowns updating the BIOS could be fatal. What if it shutsdown or crashes in the middle of the update process, you could brick the mobo.

Unless this behaviour is a known issue listed on the Mobo manufacturers page and they recommend a BIOS update, I'd leave that well alone.

Did you try remove the graphics card, use the onboard VGA port? Did you reseat the RAM?

Ive had these sort of random problems before and the RAM was the culprit. If you have two sticks try taking one out and swapping them around. Have you any spare, even older RAM which you could try?

DeVore

ae;oghj pawoiueryhqbn p[[q038945y n[toqie5;'yonqe;'5ohnq4e'5ohnq

*head desk*
*head desk*


Ok, so I made some changes to the bios (just settings not an update) and thought it was going well, got two days out of it... then black-screen-of-death and no reboot. Next day, I coaxed it back to life.

So, my next order of business is:

1. Remove one stick of memory. Wait for a reoccurrence.
2. Swap the other stick out and the first back in. Wait for a reoccurrence.
3. Remove the graphics card and run off the On Board and wait for a reoccurrence.
4. Update the bios on the mobo. (The machine fails quite sparingly so I wouldn't be too concerned about it failing during a bios update, but still... I'll leave this till last).


5. Burn the PC in a skip. At night. At a crossroads.

Johnboy1951

DeVore wrote: »

ae;oghj pawoiueryhqbn p[[q038945y n[toqie5;'yonqe;'5ohnq4e'5ohnq

*head desk*
*head desk*


Ok, so I made some changes to the bios (just settings not an update) and thought it was going well, got two days out of it... then black-screen-of-death and no reboot. Next day, I coaxed it back to life.

So, my next order of business is:

1. Remove one stick of memory. Wait for a reoccurrence.
2. Swap the other stick out and the first back in. Wait for a reoccurrence.
3. Remove the graphics card and run off the On Board and wait for a reoccurrence.
4. Update the bios on the mobo. (The machine fails quite sparingly so I wouldn't be too concerned about it failing during a bios update, but still... I'll leave this till last).


5. Burn the PC in a skip. At night. At a crossroads.

5. Try a different case
6. Donate it to me

SterlingArcher

Is the Hdd solid? And has the same been used throughout your tests.

DeVore

Yep and yep.

free_man

Hi DeVore,
I can imagine your plight having had same issue with my setup. The BSOD and Screen turning black does happen but only when I play 'Civ5'. Somehow the game hates my setup or viceversa. It doen't happen all he time but very randomly. Sometimes after 5min of running the game and sometimes after 5hrs.

The screen turning black actually doesn't even cause BSOD. I just cant see whats happening as I can still hear the game music. I also noticed that this causes my GPU fans to stop spinning. Its as if GPU has start a non-cooperation movement and shut itself down. Its not a overheating issue as I trended the GPU temp and its never above 70DegC which is quite normal for the GPU.

I found a sure way to get the computer working again. I unplug the power from machine for abt 5 min. Then reboot and and press reset button asap. If its resets, the computer starts running again. If it doesn't, rinse and repeat (i.e. remove power cable for 5 min). I only had to repeat the power cable removal twice.

My Setup is as follows
i7-940, Rampage III, 12GB RAM (6 sticks), OS on a 250GB Crucial SSD, Nvidia 560Ti and EVGA SuperNova 750. I have replaced the graphics Card, Power supply and battery on the MB. I also ram memtest but to success.
Now I just live with it.

Let me know if the RAM check works for you.