CPU/Mobo PC Reboots Automatically On First Boot After Rest

[kidrow]

Hi all,

Would appreciate some help in troubleshooting this weird problem.

So here's what happens - The first boot every morning (or essentially after a period of rest where the PC has been shut down for a while - say, an hour or so) results in a reboot. The crash can occur anytime between Windows beginning to load to after it has fully loaded (i.e. while browsing or so) & within the first 5-10 minutes.

Booting into safe mode does not change anything. In fact, just keeping the machine at the bios setup screen (by hitting F2) does not prevent the crash either.

Once the reboot occurs, the PC is rock steady & runs without issues for hours (the longest I tried being about 20 hours). Gaming, 3d rendering & similar intensive tasks can be carried out without a hitch.

It has been this way now for more than a couple of weeks.

I've tried shutting down the system properly immediately after the 1st boot on a couple of occasions. But that simply means that the crash & reboot occurs on the 2nd boot.

My specs are -
Intel i5 2400, msi h67ma-e35 mobo, Kingston 8gb ddr3, Evga Gtx 660, Saga 400w Psu.

There has been no additional hardware installed since the Gpu which was installed more than a year ago.

Please advise. Thanks for your time.

 

[Crapmypants]

How old is your current psu &
is it possible to try another one?

 

[malcolm_cg]

did you try cleaning contacts and reseating most of those components?
also seems like your psu capacity is on the borderline of the power requirement of a gtx 660.
try another psu as Crapmypants suggested.

 

[kidrow]

Thanks for your replies guys.

How old is your current psu &
is it possible to try another one?

PSU is probably ~3 years old. I do have another generic one lying about but I'm not too sure whether it'll be suitable for diagnostics as it is of poorer quality.

did you try cleaning contacts and reseating most of those components?
also seems like your psu capacity is on the borderline of the power requirement of a gtx 660.
try another psu as Crapmypants suggested.

Will try a general cleanup & reseating exercise, hopefully today or tomorrow.

While the psu may seem borderline spec-wise, it's actually got enough juice to run the rig without any problems. When I first got the gpu, I'd thrown many stress tests at it, including iray (3d) rendering which stresses both the cpu & gpu to a 100%, & it handled it well.

Of course, the psu may be aging now, but it's strange that the subsequent boot doesn't exhibit any problems at all, even with intensive tasks such as iray rendering or gaming.

Googling the problem seems to bring up cases where the common thread is a particular ram module by g-skill. In most of those cases, a bios update &/or setting the ram voltage manually seemed to fix the problem. I'm not sure if it's something similar in my case. But just thought I should mention it to provide food for thought.

 

[thedeadghost]

Try with removing RAM one by one. Usually its the culprit if not the power supply. It can also be a bad cmos battery.

 

[kidrow]

Try with removing RAM one by one. Usually its the culprit if not the power supply. It can also be a bad cmos battery.

Thanks! I'll try isolating the ram sticks a bit later.

My current plan of action involves reseating the ram & the gpu. If that fails, I'll try one component at a time, & eventually the cmos battery.

Will keep you all updated. But it could be a while since I'll have to keep the pc powered off (generally only at night) for a while to gauge any improvement.

 

[Crapmypants]

There is the possibility that the psu may be dying.
Can you use the onboard video? if so then remove the gtx 660 (and extra drives/components) and then see if it restarts after the first boot.
the idea here is to try booting with as little load on the psu as possible.
if the psu is unable to handle cold starts then it could be near the end of it's life.
if you're not comfortable with using the other psu then that's fine. no point using a suspect part to diagnose another problem.

Also, best way to clean the ram pins is with a soft eraser. aside from hard-dried and stuck dirt it gently gets any rust off as well.

 

[kidrow]

Thanks for the help guys.

I went ahead & reseated/cleaned up the ram, gpu & battery. There hasn't been the reboot issue since.

However, I'm not out of the woods yet. I'm getting the "Display driver nvlddmkm stopped responding and has successfully recovered" error every now & then.

Immediately after I first reseated everything, I was getting the error on basic tasks like browsing. So I thought that I must have not installed some component properly. I reseated everything a second time. The error then stopped popping up until I actually started playing a game.

Next, I decided to re-install the nvidia drivers & installed some older drivers by following a guide for a clean install (which included deleting folders & registry entries manually). I then ran the Uniengine Heaven 4 Benchmark for an hour. There were no errors & the benchmark ran glitch-free.

However, I again had the error pop up in-game. So I next disabled aero & uninstalled phys-x. That seemed to resolve matters, because I could game for a couple of hours without any issues.

Today though, I've encountered the error again in-game after about 30 mins of gameplay. Will probably try the steps here next - https://support.microsoft.com/en-us/kb/2665946.

Sigh! Woe is me, . Well, on the bright side, at least the reboot issue got fixed, .

Cheers! Again, thanks for your time.

 

[Crapmypants]

it seems to be a windows issue as people with amd cards have also reported similar problems.
here are 2 more links you might have already come across them though:
https://forums.geforce.com/default/...-tdr-errors-display-driver-nvlddmkm-stopped-/
http://www.instructables.com/id/How-to-Fix-the-Nvlddmkm-Error/

you mentioned that you've tried different driver versions and a fresh os install (i'm guessing with full updates). what about the bios and motherboard chipset drivers? check if the manufacturer has any new drivers available online.

 

[kidrow]

Well, I haven't re-installed the OS. Just the drivers. (The 2nd time around, I used the Guru3d DDU which detected my older 8600gt as well. Yet there was no difference.)

I ran the Furmark gpu stress test for about 45 mins without any problems, just to re-check whether the gpu itself is alright. Next, I decided to give ccleaner a try (before committing to prime95 & memtest). I also manually went through the add/remove programs & uninstalled unnecessary programs. Here, i realized that I had 2 versions of the Logitech Gaming Software installed. I uninstalled the older one.

A reboot later, I tested out a couple of games. I'm glad to say there weren't any problems for the couple of hours that I put in.

So it was either some junk that was cleaned up by ccleaner or a conflict between the 2 logitech gaming software versions. Since the problem was happening only in-game & not during stress tests, maybe the latter is more likely. (Though I've been running both versions for quite some time now). (?).

Am I speaking too soon? I hope not!!

The silver lining is that it provides me with an excuse to test {read game, } some more. If the folks at home have a problem, I can always say "I'm testing!" ...

Will update soon. Cheers! & thanks!

 

[kidrow]

So it was either some junk that was cleaned up by ccleaner or a conflict between the 2 logitech gaming software versions. Since the problem was happening only in-game & not during stress tests, maybe the latter is more likely. (Though I've been running both versions for quite some time now). (?).

Am I speaking too soon? I hope not!!

Well, I did speak too soon, lol! As it turns out, it wasn't either of the reasons I've stated above. I had the driver crash on me again on the day following my post.

So I next edited the registry & put in that new value as suggested by the Microsoft article I'd linked to a few posts above. Spent quite some time the next couple of days gaming without any glitches.

Just as I was starting to celebrate though, another spanner in the works! My PC wouldn't boot this afternoon -

So I killed the power & rebooted. Now the PC did boot but without the display. (I could hear Windows' welcome tune). I rebooted properly, though blind, via Start menu>Right arrow>Enter. No change.

Removed the gpu & connected to the onboard chip. This worked & I had a working display. Rebooted successfully a few times. Put the gpu in again. This time I got the display back. Rebooted a few times & found everything in order.

Reconnected usb devices, audio connections, lan etc. & closed up the side panel of the case. Tried to power it on & again ..... nothing.

This time around, even after connecting via onboard vga, the unit wouldn't boot up.

Gave it the ol' "Hold down power button for 30 secs with the power cord out" treatment. That worked.

Re-installed the gpu. Held down the power button as earlier & then booted. Met with success this time around.

Will be running an iray test render overnight to max out the cpu & gpu. Let's see if there are any issues.

At this point, I'm hoping (nay, praying!) that the boot problems were just an aberration & have now been taken care of with that power button maneuver. Fingers crossed!!

 

[kidrow]

Well, I ran the iray render for 6 hours straight without any problems. Used Task Manager & EVGA Precision X to make sure that the CPU & GPU usage were both 100%.

So, it's probably safe to assume that the PSU, CPU & GPU are okay. (Unless there is some intermittent malfunction which does not become apparent at all times). That leaves the RAM & the Mobo. Or maybe there's something else that causes excess static electricity to remain in the system while the unit is off. (Which is what is drained apparently when using the power button without the PC being plugged in, from what I've read).

For now, it's wait & watch. I'll probably make it a point to do the power button thingy every time I boot up. That way if no problems occur (or even if they do occur), it might help narrow down the possibilities.

Will update. Ciao!

 

[v.Na5h]

I had the exact same issue..
Thought mobo was the culprit ..
Since it used to run fine once it's on

Even spend 8k getting a new mobo + CPU
Found out later that it was the psu

Gotta get a cabinet and RAM and psu now

 

[kidrow]

I had the exact same issue..
Thought mobo was the culprit ..
Since it used to run fine once it's on

Even spend 8k getting a new mobo + CPU
Found out later that it was the psu

Gotta get a cabinet and RAM and psu now

Yeah, that's exactly what I'm trying to avoid - Spending unnecessarily on something that may not be actually faulty.

But wouldn't you say the PSU is not the culprit here based on the stress test involving simultaneous use of both the CPU & GPU for ~6 hours? I'd imagine any weakness that the PSU might have would be exposed by such a torturous exercise.

I actually have a new theory. This is an off-shoot of the idea that some static electricity is the cause of my woes.

I hadn't mentioned that my Creative T10 speakers have been crackling & popping for some time now. I just didn't think there was any connection, until now.

I'm now starting to think that the malfunctioning amplifier in the speaker is not only contributing to the crackling sound but also causing some sort of static electricity issues for the PC.

I may be completely wrong, but I haven't had any issues today with my speakers disconnected. I have booted the PC up twice today (after brief periods in which they were kept shut down) without issues. The first of these was done after the power button maneuver. The second was done without any such special measures.

Of course, it's too early to say. So I'll update after a couple of days.

PS: I have ruled out the PSU sending out dirty power as the cause for the crackling speakers because the speakers exhibit the same issues even when not connected in any way to the computer. Just the one connection to the power board is enough for it to start crackling.

 

[v.Na5h]

Yeah, that's exactly what I'm trying to avoid - Spending unnecessarily on something that may not be actually faulty.

But wouldn't you say the PSU is not the culprit here based on the stress test involving simultaneous use of both the CPU & GPU for ~6 hours? I'd imagine any weakness that the PSU might have would be exposed by such a torturous exercise.

I actually have a new theory. This is an off-shoot of the idea that some static electricity is the cause of my woes.

I hadn't mentioned that my Creative T10 speakers have been crackling & popping for some time now. I just didn't think there was any connection, until now.

I'm now starting to think that the malfunctioning amplifier in the speaker is not only contributing to the crackling sound but also causing some sort of static electricity issues for the PC.

I may be completely wrong, but I haven't had any issues today with my speakers disconnected. I have booted the PC up twice today (after brief periods in which they were kept shut down) without issues. The first of these was done after the power button maneuver. The second was done without any such special measures.

Of course, it's too early to say. So I'll update after a couple of days.

PS: I have ruled out the PSU sending out dirty power as the cause for the crackling speakers because the speakers exhibit the same issues even when not connected in any way to the computer. Just the one connection to the power board is enough for it to start crackling.

I took the plunge since I needed a htpc for TV anyway..

Even I ran prime 95 and ran perfectly for three days.

 

[kidrow]

I took the plunge since I needed a htpc for TV anyway..

Even I ran prime 95 and ran perfectly for three days.

Hey, thanks for your response. Prime95 for 3 days! Whoa!

I haven't yet run Prime95 or Memtest. Just the iray test render (which I personally believe is a pretty good stress test). If the problem does turn out to be the PSU, it'll probably be some flaw which manifests itself when it has been powered off for a while, rather than when under duress.

Right now, it's wait & watch for me. The PC hasn't misbehaved since yesterday. Wish me luck!

 

[v.Na5h]

Hey, thanks for your response. Prime95 for 3 days! Whoa!

I haven't yet run Prime95 or Memtest. Just the iray test render (which I personally believe is a pretty good stress test). If the problem does turn out to be the PSU, it'll probably be some flaw which manifests itself when it has been powered off for a while, rather than when under duress.

Right now, it's wait & watch for me. The PC hasn't misbehaved since yesterday. Wish me luck!

No not three days continues..
I meant the pc was functioning normally for three days...
Then it started giving the restart issues

 

[Samkong]

yes, it could be the PSU oand/or the BIOS BIOS tend to get corrupt due to sudden/Improper Shutdown so please check both. QUOTE="Crapmypants, post: 2041976, member: 58162"]How old is your current psu &
is it possible to try another one?[/QUOTE]

did you try cleaning contacts and reseating most of those components?
also seems like your psu capacity is on the borderline of the power requirement of a gtx 660.
try another psu as Crapmypants suggested.

it could be[DOUBLEPOST=1438277294][/DOUBLEPOST] After checking the current requirement by the Graphix and other components, [also the ability of the PSU] and updating the latest BIOS please update where lies the problem to let us know

 

[kidrow]

No not three days continues..
I meant the pc was functioning normally for three days...
Then it started giving the restart issues

Thank God for that!! I was feeling sorry for the lil' ol' chip

@Samkong I'm pretty sure that the PSU (when working at it's optimum capability) is enough for the rig, because ever since I've purchased the gpu, I've done quite a few iray renders & gamed a fair bit without any issues. Of course, I'm not ruling out a failing PSU yet, because as Nash's experience shows it might run well temporarily despite being faulty.

Yes, the BIOS upgrade did cross my mind. But I'm holding off for now, as the rig is running fine atm. If it does misbehave, then I'll probably run memtest first & then try out a new BIOS later.

Will update soon.

 

[kidrow]

Update- The problems, I believe, were a result of excess static electricity. While I'm not sure whether that description is technically accurate, I do know that my boot-up problems were resolved once I started to drain the static electricity (supposedly) from the system before the first boot-up. (By draining static electricity, I mean holding down the power button with the main power cord removed).

So to anyone that stumbles across this problem, do explore this option after you have stress tested all of your components and found them to be okay.

In my particular case, once I realized that draining the static electricity was working, my suspicion immediately fell on my crackling & popping speaker (a Creative T10). I disconnected the speakers & found over the next few days that the PC was back to normal & did not require me to discharge static electricity every morning.

I got my speakers repaired & now have been using the PC for the past 2-3 weeks with the repaired speakers connected. I'm happy to say that none of the problems have cropped up since. Hopefully, this problem has been resolved for good!

A big thank you to everyone who chimed in. Much appreciated! Cheers!!