Windows 10 Reboot Loop

[Mitsu3000gt]

Hi guys,

Strange problem I need some help troubleshooting. My computer was on for a few hours, just had it running for Plex. Woke up the screen, clicked on Firefox (don't think that matters) and I heard the power supply "click" and it rebooted itself. It then enters a constant reboot loop, stopping JUST before the windows 10 log-in screen. Just before it would change to allow me to enter a password, you hear the power "click" and it reboots indefinitely to that point just before the login screen would normally appear.

First I thought this was a hardware issue, but I can get into my bios screen and it will sit there indefinitely so it's definitely getting power. About 6 months ago I got a brand new EVGA Supernova P2 power supply as well so I doubt that is the issue. CPU temp is 25C so I don't think it has anything to do with voltage or overclocking.

I googled it obviously but the issue seems very new. A couple people seem to have the problem but no solution, or the solution involves using windows which is impossible. Might be a bad W10 update file. I am trying to get into safe mode but I read that safe mode doesn't exist on W10 if you're using a SSD like I am.

Any suggestions? I'm pretty sure it's windows and not hardware, but I cannot get into windows so the only things I can try are in the first 10 seconds of boot before the power cut. TIA


EDIT: By some miracle I got it into system recovery mode and it's currently trying to restore to a previous version. ..fingers crossed. If it works I need to then figure out what update is the culprit.

EDIT2: Nope, still reboots every time at the same spot even after trying to restore. I'm stumped. It says "Automatic repair could not repair your PC"

[spikerS]

happened to me too, had to do a system recovery to get it back from a restore point.

I am thinking there are some buggy windows auto updates out there...

[Mitsu3000gt]

So just make a recovery USB and try that? I'm just found a USB stick to try it with.

Did you have to do anything else? Or just recover from a USB?

I made what I thought was a recovery USB and all it did was give me the same options to roll back to a previous build OR restore from a restore point, both of which failed.

There is a last option to reinstall windows....is that what you had to do?

EDIT: I got into Safe mode and everything works! I just need to figure out what update causes the reboot loop....safe mode won't let me into windows update so I'm screwed again haha. Disabling automatic restart after system failure also does nothing.

I can also get to a command prompt if I can identify the problem update and uninstall it there somehow maybe.

EDIT 2: Now it reboots every time I try enter safe mode. Right after I enter my password and hit enter it reboots.

[spikerS]

weird, I know on mine, i just had to get it to the point where I could select a system restore point, and then let it do it's magic.

[Mitsu3000gt]

Originally posted by spikerS
weird, I know on mine, i just had to get it to the point where I could select a system restore point, and then let it do it's magic.

Yeah tried every restore method short of reinstalling windows and they all fail.

I thought it might be update KB308124 as there are recent articles about it, but I tried removing it via command prompt and it says the update was never installed so I'm running out of things to try.

I wish I could get into safe mode again but I can't, it reboots when I hit "enter" after my password.

EDIT: Tried to get into safe mode again (failed and rebooted the instant I selected the option) and now it wont even run for more than 5-6 seconds before powering off and rebooting, I can't do anything at all anymore. Something is totally fucked but I am out of ways to troubleshoot unless it starts up longer. I barely get to see the ASUS mobo startup graphic and it reboots.

[revelations]

Boot to WinPE or something similar - check the HDD SMART info as its likely dying or toast.

[Mitsu3000gt]

The OS drive is a Crucial M4 from 2012, maybe it's just finally dead. Ran crystal disk on it recently through and it got a clean bill of health.

Thing is though if I can get into the recovery options, the PC will stay on indefinitely, and ONCE I got into safe mode, at which point it also stayed on indefinitely. Makes me think it's not hardware, but who knows.

My copy of W10 is an upgrade from a legit purchased copy of W7. If I go buy a new SSD and do a fresh install, will I have to buy a new copy of W10 too due to the shit they pulled back in the summer? Or will it let me use my W7 CD key?

The only thing I haven't tried yet is a system restore via recovery options which blows away all my programs. Might as well try that though if I can get back into that screen somehow before a reboot.

EDIT: Breaking news, I let my computer sit for 10 min turned off and it started up just fine :dunno: :dunno: What a waste of 3 hours, fuck. Now I think it's hardware again...

Edit#2 coming to you from my phone...no sooner did I write that and my computer powered off again into the loop. Probably my PSU or SSD, and my PSU is new but maybe a lemon I guess. Can a failing SSD even cause a full power down or reboot loop?

[spikerS]

wonder if it could be a heat issue then?

[sl888]

I had this exact problem with my ASUS G73 laptop as soon as I installed W10. I had to do a clean restore for W10 then update the ASUS BIOS and graphic driver. Somewhat solved this reboot loop problem. Now I must unplug all external devices before starting the laptop or else it just freezes at sign in. FUCK W10!

[Mitsu3000gt]

What's interesting to me is it doesn't freeze, and it doesn't do a soft reset with continus power. It does a HARD power off as if someone pulled the plug out of the wall, then powers back on and restarts on its own.

I don't think it's the PSU because if I can get past certain startup points (there is a pattern of about 10 progressively longer reboots before it lets me into recovery mode) it will run indefinitely. If my PSU was the issue it would be indifferent to what the rest of my computer was doing.

The pattern if I just let it do its thing is about 5 immediate failures before the bios screen (all hard automatic resets), followed by 2 failures at the windows loading screen, followed by one failure at the windows recovery screen, and then it finally let's me into Windows recovery. It goes progressively further.

Starting on safe mode does a HARD reset either immediately when I select the type of safe mode, or when I enter my login password and hit Enter. Once I'm in recovery mode though it will let me stay in there indefinitely and navigate around until I either reboot or go into safe mode. All other recovery tools will run (and fail) but don't cause reboots.

I'm thinking it's my SSD or a badly corrupt windows update. I might go buy a new SSD today and clean install windows. Everything has been working flawlessly with W10 and this hardware for months, so this was very sudden. At least according to my bios, temps are all nice and cool. I don't have the ability to check temps beyond that point.

At the end of last night when I miraculously got into Windows normally after letting my computer sit off for 10 mins, it was like a ticking time bomb. I had about 20 seconds before the hard power off and beginning of the reboot loop.

[Seth1968]

Holy shit Mitsu, that's all over the place.

A shut down like you describe is almost certainly a hardware failure, but to be sure, try running a Linux distro.

[Mitsu3000gt]

^^ Yeah it's super weird. Every time I had an idea, a different event contradicted it.

Well after sitting over night, I just fired it up and so far so good. That again makes me think hardware. HW monitor telling me temps are good.

Last windows update that wasn't windows defender definition updates was November 20, so I don't think it's a windows update issue.

Any checks I should run while I have it up and running?

The fact that it sat overnight and now works (for now) makes me think it's my SSD. The Crucial M4 can be finnicky with power cycles, but what it doesn't explain is the first hard reboot from windows. I've got the latest FW on it.

What I might do is get a new SSD anyway, this one is ~5 years old, and gets used pretty hard. It's the unpacking drive for my usenet and also my scratchdisk for Photoshop as well as of course the OS drive. At least with it running I can use Samsung's slick drive mirror tool.

I tried to run League of Legends just to try give it something to do, and it won't even launch. Super weird. Photoshop will run though. The PC in general seems a little slower than usual.

Crystal Disk says my SSD is 96% healthy.

[revelations]

Yea next step might be hardware test tools (RAM, CPU, board) , in a bootable environment like Linux or WinPE.

I used to have one of those M4s- make sure you get the latest firmware update to it as there was some issues with the 2012/2013 era firmware.

[Seth1968]

Chkdsk since W8 is really good. Run it twice to be sure, as it does have a quirk similar to System Restore.

Don't bother with the command prompt chkdsk, do it from the c drive properties.

[Mitsu3000gt]

Originally posted by Seth1968
Chkdsk since W8 is really good. Run it twice to be sure, as it does have a quirk similar to System Restore.

Don't bother with the command prompt chkdsk, do it from the c drive properties.

Went to run it, it said I didn't need to because it was OK. Ran it twice anyways, no errors.

My gut tells me this has something to do with the Crucial M4's sensitivity to power, even though it's a different event than what typically happens when it's unhappy.

Just sitting overnight shouldn't really fix a problem haha, but I've been up and running for record time this morning.

A am using a EVGA Supernova P2 power supply and I have a pure sine wave UPS battery backup (which hasn't kicked in) so it should be getting the best possible power.

[Seth1968]

Don't want to burst your bubble, but I suspect the issue will continue.

My guess is the SSD. Either physically, or a firmware update if available.

[Mitsu3000gt]

Originally posted by Seth1968
Don't want to burst your bubble, but I suspect the issue will continue.

My guess is the SSD. Either physically, or a firmware update if available.

My gut tells me it's the SSD as well, especially with the M4's known history with power sensitivity. I'm happy its running now but by no means do I think I should just leave it alone.

Brand new Samsung 850 250GB is $119, that is a no-brainer to swap it out at this point.

[Mitsu3000gt]

Well it died again, curiously immediately after I used the Windows search bar again. Off to memory express.

[Seth1968]

Originally posted by Seth1968
Chkdsk since W8 is really good. Run it twice to be sure, as it does have a quirk similar to System Restore.

Just thought I'd elaborate:

The first attempt at a chkdsk may show no need to run it as Windows found no errors, but if you force the chkdsk, Windows will often find errors and attempt to fix them.

System Restore has a glitch which shows SR has failed, but if you restart, you'll often find that it didn't fail, and the problem you were addressing has been resolved. You'll even get the message that SR completed successfully.

[Mitsu3000gt]

Put a new SSD in and it and it hard power loops after about 2 seconds of being on....fuck. I cant even get into the bios. Looks like the SSD was fine. Probably mobo or PSU, might have to pay memex to look at it, I don't have the time to start swapping out parts one by one to see what works.

[blitz]

Should have bought a Mac.

[Xtrema]

Originally posted by blitz
Should have bought a Mac.

So you can just throw the whole thing away and just buy a new one?

:rofl:

You can do that with PC too.

[Seth1968]

Originally posted by Mitsu3000gt
Put a new SSD in and it and it hard power loops after about 2 seconds of being on....fuck. I cant even get into the bios. Looks like the SSD was fine. Probably mobo or PSU, might have to pay memex to look at it, I don't have the time to start swapping out parts one by one to see what works.

Damn it Mitsu, that's why I suggested trying a linux distro to confirm.

Now you're most likely looking at a mobo issue. It's a long shot, but you could try one ram stick at a time, and also resetting the bios. I don't know if you've reset a bios before, but unplug it, and pull the battery for a few seconds.

EDIT: I misread your first post and somehow thought you already replaced the PSU. If you take it to a shop, that's the first they'll check out at this point.

[Mitsu3000gt]

Originally posted by Seth1968


Damn it Mitsu, that's why I suggested trying a linux distro to confirm.

Now you're most likely looking at a mobo issue. It's a long shot, but you could try one ram stick at a time, and also resetting the bios. I don't know if you've reset a bios before, but unplug it, and pull the battery for a few seconds.

I got it running again (sort of), it's currently sitting in the BIOS screen, having recognized both my boot USB and new Samsung SSD. I could pretty easily try to put windows on it and see if that works still - not sure its worth it though.

Probably mobo, or some other component.

My PSU is new, but I haven't replaced it as part of this trouble shooting. It's possible I got a lemon there but I bought a very good one.

[Seth1968]

I edited my last post while you were posting.

Intermittent problems are the worst.

[Mitsu3000gt]

Yeah the problem is I've got it to work normally for up to an hour or so. I could think it's fixed when it's not which makes it super annoying.

Well I tried to boot from the USB drive with W10 on it, since I had the bios up and running but as soon as I hit "enter" it did a hard shutdown. I think I will just be bringing it in. I'm 99.9% sure it isn't the SSD at this point.

Reset BIOS and pulled all 3 other hard drives as well and still same issue. Just going to pay memex to look at it at this point, I don't have the time right now to start replacing things one by one.

[Mitsu3000gt]

Got a call from Memory Express and they are classifying it as "no fault found". All my hardware passes with flying colors including stress test.

He did find a couple windows errors and is going to update my bios, but that's it. I'm still going to try out the new SSD and a fresh windows install just in case but beyond that if it happens again there isn't much more I can try.

[revelations]

Couple of questions:

a) did they use a different PSU or do you have another one to use?
b) how much did they charge for the test?

Also, (this happens all the time) did you perform a complete system de-power? This means disconnecting the power cable and hitting the power on button (to drain the capacitors, etc.).
Many times, weird electrical problems get solved this way.

[Mitsu3000gt]

Originally posted by revelations
Couple of questions:

a) did they use a different PSU or do you have another one to use?
b) how much did they charge for the test?

Also, (this happens all the time) did you perform a complete system de-power? This means disconnecting the power cable and hitting the power on button (to drain the capacitors, etc.).
Many times, weird electrical problems get solved this way.

I don't think they changed out the PSU. It would be easy because it's modular and the cords could stay in place. The tech couldn't replicate the issue I don't think, so there was no reason to start swapping things out. My computer did display the issue for the person who checked me in at the service desk though, so they witnessed it.

I disconnected my PC many times and flicked the "off" switch on the PSU but I never held down the power button. The LED's eventually go out which I assume is the capacitors draining. At this point, "weird electrical problem" might just be the most likely culprit based on all the info I have at this point.

They charge $50 non-refunable for the first hour of diagnostics and $50/hr thereafter. They call you after the first $50 to tell you what's wrong and it's your call how to proceed. I talked to the actual tech on the phone today and he seemed like a good guy, knowledgeable, and pretty loose with hours/rates. He said he would only charge me $50 to reinstall the new SSD, test it, and install windows on it (something I could easily do myself) so I told him to go ahead because I've already spent enough time on it. Hopefully when I take it home today it's good to go.

I still wonder if my PSU isn't a lemon, but if the system passes stress tests and whatnot, the PSU is probably working just fine.

[Seth1968]

I've never used a PSU tester, as they don't really address intermittent issues, or under load issues. It's far more reliable to quickly hook up a new PSU for testing. It's not like the old one even has to be taken out.

Ditto for software ram testers.

The sucky thing is that the problem is intermittent.

[Mitsu3000gt]

Originally posted by Seth1968
I've never used a PSU tester, as they don't really address intermittent issues, or under load issues. It's far more reliable to quickly hook up a new PSU for testing. It's not like the old one even has to be taken out.

Ditto for software ram testers.

He tested my ram and mobo which both were fine.

If my PSU holds up to full system stress tests and such I would think it's fine. If the issues are intermittent, swapping out the PSU might not reveal anything either which would be very frustrating.

It's the HARD power-downs that make me still suspect the PSU, but who knows. It does start back up by itself. Maybe the PSU is entering protection mode or something which kills the PC temporarily. Or maybe it's not the PSU at all....haha.

[revelations]

I've seem RAM fail tests many times, resulting in an uncommanded reboot - but a SHUTDOWN usually implies a power issue.


OP, if the problems continue - go back to mem ex and pickup a different PSU - although if you have a 6 month old 1000watt Platinum-grade PSU, its highly unlikely to be the problem (but not impossible).

If you end up trying a different PSU, take the whole system apart and check the pins on the board - sometimes I've seen the pins pushed up and away from the connector resulting in some random weirdness like this.

You have 7 days to return the PSU for full refund - you should know then if the problem is PSU related or not.

[Mitsu3000gt]

Originally posted by revelations
I've seem RAM fail tests many times, resulting in an uncommanded reboot - but a SHUTDOWN usually implies a power issue.


OP, if the problems continue - go back to mem ex and pickup a different PSU.
You have 7 days to return it for full refund - you should know then if the problem is PSU related or not.

Yeah swapping out the PSU was my next plan if all else fails. I was going to do that on the weekend actually but I had already spent too much time on it. If the PSU is in fact the issue, it's definitely strange and intermittent. It's a few month old EVGA Supernova P2 (650W) which has been reviewed as pretty much a flawless PSU. Possible I got a lemon though of course.

The hard shutdowns as if you yanked the power cord out of the wall (and then it stays off for 2-3 sec before turning back on) is what made me suspect the PSU at the start of all this. The "looping" it does though is more curious, especially since it's stable at certain points (Like bios screen or windows recovery screen).

Tonight we will see I guess. It will have a new SSD, fresh W10 install, and all firmware/bios updates.

I thought maybe my PSU might be underpowered, but everyone told me it was fine when I bought it. PC Part picker said I was way under 650W too.

i7 3770K modest overclock on air
3 fans
3 HDD's / 1 SSD
nvidia GTX TITAN X (Maxwell)
16 GB RAM

The most demanding game I play is Leage of Legends haha. It gets a workout with Photoshop though and it's always been fine with good temperatures.

[Swank]

Until you get this resolved I'd disconnect the HDDs if possible, unless newer hard drives are OK with sudden power loss, I remember hearing the heads can be easily damaged when they power back up.

[Mitsu3000gt]

Originally posted by Swank
Until you get this resolved I'd disconnect the HDDs if possible, unless newer hard drives are OK with sudden power loss, I remember hearing the heads can be easily damaged when they power back up.

They are strictly storage drives, so I think 99% of the time they aren't spinning. If it can happen regardless though, maybe I'll leave them out for a bit.

I am hoping that it is somehow resolved already with the updates & fresh SSD/Windows but we will see. I'm skeptical due to the hard nature of the power-downs that it's truly resolved. In my mind, when it did it with the new SSD, that pretty much ruled out windows or the SSD as the issue.

[cam_wmh]

Originally posted by Mitsu3000gt
Got a call from Memory Express and they are classifying it as "no fault found". All my hardware passes with flying colors including stress test.

He did find a couple windows errors and is going to update my bios, but that's it. I'm still going to try out the new SSD and a fresh windows install just in case but beyond that if it happens again there isn't much more I can try.

Similar issue on a friend's Lenovo that he picked up from MemEX about 16 months ago. Just kept looping to that off green pre-login screen, then would reboot.

Managed to get it into Safe Mode to get his data off, but haven't been able to since. BIOS wasn't much help either. Dropped it off at MemEx.

Will let you know what they say.

[Mitsu3000gt]

Originally posted by cam_wmh


Similar issue on a friend's Lenovo that he picked up from MemEX about 16 months ago. Just kept looping to that off green pre-login screen, then would reboot.

Managed to get it into Safe Mode to get his data off, but haven't been able to since. BIOS wasn't much help either. Dropped it off at MemEx.

Will let you know what they say.

Yeah I'd be curious.

I keep circling back to power supply just because of the nature of the hard power cuts, but then enough other stuff happens that makes me think it can't be the PSU :nut: Mine would also crash (hard power-off) if you tried to select safe mode, but would stay in the safe mode options screen indefinitely, which would suggest it is not the PSU haha.

My PC has been good to me for 5 years, I can't ask much more of it. If it turns out I need a new one I won't be upset, but I remain optimistic for the time being. Maybe it just needed a different environment for a while at MemEx haha who knows. It's probably something silly still.

[spikerS]

I still say it was a W10 update. I have been hearing these stories a lot. I sent mine back the first time it happened. Got it back, it did an update, and started doing it again. I was able to system restore it, and it has been fine now for the last month or so.

[Seth1968]

^ Given that Mitsu heard a pop, it often halts at the BIOS, and he's already tried a new SSD, then it's most certainly not a software issue.

If I had it on my bench, I would loosely connect a psu, do other stuff, and attempt a restart every 10 minutes or so. The psu was always the main culprit, but I had a read fail on Mitsu's original post and thought he already tried another psu.

[spikerS]

Originally posted by Seth1968
^ Given that Mitsu heard a pop, it often halts at the BIOS, and he's already tried a new SSD, then it's most certainly not a software issue.

If I had it on my bench, I would loosely connect a psu, do other stuff, and attempt a restart every 10 minutes or so. The psu was always the main culprit, but I had a read fail on Mitsu's original post and thought he already tried another psu.

I know. It is exhibiting the identical symptoms mine did. When I sent my laptop in under warranty, they replaced the motherboard and HDD, and then a fresh copy of W10.

Worked for a few hours, till it did an automatic update, and then started doing it again. I was able to get it to do a recovery from the fresh install date, and mysteriously, no more hard boots and struggling to get past the bios check.

[Mitsu3000gt]

I talked to the Memex Tech again and he said he tested my PSU and found no issues. Unfortunately the issue *did not* happen for the tech at all, but did happen when I dropped it off so the person who checked me in witnessed it. My concern is that the tech was simply on borrowed time, since I've got about an hour out of it before as well before it died. He stress tested it and everything though so we'll see.

I didn't have time to play around with it last night, but tonight I will be setting it all back up again with my programs since the fresh W10 install. If the issue happens again, I will be swapping out the PSU as I can't imagine there is much else that could cause a hard power down.

My last windows update was November 20th, so it worked for a long time on the latest updates without issue, but who knows. When it was exhibiting the issue on a fresh SSD without even having windows installed though, it's hard to believe it's much other than the PSU or other faulty system hardware.

I appreciate everyone's input, it's always annoying diagnosing stuff like this. This one stumped me, first time in my entire life I've had to bring a computer in for diagnostics and they couldn't find anything either haha.

[Seth1968]

If the issues continues (it probably will), and the mobo has a video port, pull the GPU and try the stock graphics.

Following that, the next step is a PSU swap, as it doesn't sound like the shop did that.

Also, there is no practical way to "test" a motherboard. A failed mobo is determined by a process of elimination.

[mazdavirgin]

Originally posted by Mitsu3000gt
I talked to the Memex Tech again and he said he tested my PSU and found no issues. Unfortunately the issue *did not* happen for the tech at all, but did happen when I dropped it off so the person who checked me in witnessed it. My concern is that the tech was simply on borrowed time, since I've got about an hour out of it before as well before it died. He stress tested it and everything though so we'll see.

You may want to monitor your onboard temps. If the problem takes a while to occur temperature is one of those things that's real easy to test. Watch your temps and if needed get some cold spray. If spraying the component with cold spray causes it to go back to being stable voila problem solved.

[Mitsu3000gt]

Originally posted by mazdavirgin


You may want to monitor your onboard temps. If the problem takes a while to occur temperature is one of those things that's real easy to test. Watch your temps and if needed get some cold spray. If spraying the component with cold spray causes it to go back to being stable voila problem solved.

When I had it up for an hour, I had HWMonitor running the entire time, and the basement I'm in has an ambient temperature of 15-16C usually. All CPU cores stayed around 25C according to the program. You're right though that what it's doing would seemingly be consistent with a thermal shutdown, or some kind of protection circuit, but nothing I can tell with monitoring programs, and apparently nothing that happens during a system stress test.

Fingers crossed that it is somehow resolved now but we will see over the next little while I guess.

[Mitsu3000gt]

Got my computer home, fired it up and within 2 minutes it hard powered down...I kind of expected it to be honest.

However....I switched the power cable from one of the battery backup slots on my Cyberpower UPS into one of the non-battery backup (but still protected) slots, and so far so good. Too early to tell, but I wonder if all along it was a stupid power issue from my UPS or something, triggering something.

Of course, to contradict that logic, it did display the issue when I first dropped it off at Memex.

At the very moment it's working, but I am back to square one on my diagnosis.

EDIT 1: It died again. Almost every time it's happened, I clicked in the windows/cortna search box, but not all.

EDIT 2: I have a space heater in the room running on what is probably the same circuit, however I have had it on for HOURS before with no issues. Turning the heater off now and my computer started back up...interesting, but not consistent. You'd think my UPS and PSU would deal with any power fluctuations.

EDIT 3: Nevermind, it still does it. Literally every imaginable possibility is strongly contracted by something else. I might try a new PSU and also plug it straight into the wall.

EDIT 4: I swapped out my power cable and it is currently working again...stay tuned for the update where it dies again. If after all this it was a power cable I am going to feel like a bit of an idiot but I'll be happy.

EDIT 5: Died again. Curiously every time it dies I'm trying to access something from the windows start menu, but it isn't repeatable so probably coincidence. Tried new power cable, tried plugging it straight into the wall - no go. I would be convinced it is my PSU or Mobo at this point if it didn't pass stress testing at memex.

[Seth1968]

Originally posted by Mitsu3000gt
I would be convinced it is my PSU or Mobo at this point if it didn't pass stress testing

Again, there is no such thing.

[Mitsu3000gt]

Originally posted by Seth1968


Again, there is no such thing.

Well then it's probably one of those two things.

EVGA said they would send me a new PSU, so I'll try that I guess.

I got it running again, and as long as I'm not doing much, it stays running seemingly indefinitely. As soon as I start using the start menu / windows search functions, that seems to always be when it shuts down but could be purely coincidence. Also I think we've ruled out windows being the culprit.

[Seth1968]

Originally posted by Mitsu3000gt
EVGA said they would send me a new PSU, so I'll try that I guess.

Uh huh :dunno:

[Mitsu3000gt]

I talked to them some more and it's going to take too long, and I have to pay to ship it back, so I'm just going to buy one locally instead and try it out. I'll have 7 days to see if it works or not from Memex.

Things are running perfectly right now which is why this is so weird. I know though as soon as I click in the windows search menu it's going to power down again.

[Seth1968]

And this is on the new ssd?

[Mitsu3000gt]

Originally posted by Seth1968
And this is on the new ssd?

Yes sir.

What makes this so strange is there is strong evidence of a software issue, but then the issue still happened when windows wasn't even on my PC at all. Then, when I'm in windows, it seems I can do certain things to trigger it. I just installed League of Legends and it went fine, but I bet if I start playing around with the windows start button it will power off. So weird.

[Seth1968]

No, all evidence points to a hardware problem.

Did you read my post about the GPU?

[revelations]

If not PSU, my next guess is a bad RAM stick or channel.

[mobius]

I'm going to throw this out there, but what if it is not a PC hardware problem, but rather a problem with the AC power? Can you try to plug it in a run it on a different plug, preferably on a different circuit breaker?

[Mitsu3000gt]

Memex tried with my GPU in & out, and memory in & out. I don't think it's those (just trusting them on that of course).

My computer has been on since my last post. Everything seems fine since I haven't ventured into the Windows start menu or search box.



Originally posted by mobius
I'm going to throw this out there, but what if it is not a PC hardware problem, but rather a problem with the AC power? Can you try to plug it in a run it on a different plug, preferably on a different circuit breaker?

It's possible. I've tried two separate outlets, tried different power cords, and everything is running through a higher end Cyberpower UPS/power conditioner. No other electronics, including my laptop, have issues on the same outlet either.

[BigMass]

Originally posted by Seth1968

Also, there is no practical way to "test" a motherboard. A failed mobo is determined by a process of elimination.

Common in terms of hardware issues and the most annoying piece of hardware to replace lol Don't get me started on bad motherboards :nut:

[Mitsu3000gt]

Well I double clicked the league of legends icon to test it and it immediately powered down. Stuck in the loop again. So weird. Every time I think there's a pattern there isn't one haha. And if I let it sit for a while it will work again for a bit.

I'm going to try a new PSU as soon as I have time. A PSU and a HDD are the only two computer related items I've ever had fail on me in my life so I won't be too surprised its that.

[Sal0]

Like most people said, it's probably the PSU. I had Memory Express "test" a bad PSU once, and they said it was fine so you can only trust them so much.

You should also check the event log to see if you find any interesting errors around the time it shuts down.

[Mitsu3000gt]

Here's a thought (that I'm sure will be soon disproved via hard power down). My Cyberpower UPS has a USB cable that connects to my computer so I can monitor things, change settings etc. I unplugged that and so far so good. I'm sure soon I will be editing this to tell you it still failed, but any chance it could be receiving false shutdown signals from that? I did experience the problem after bypassing the UPS, but I did leave the USB cable plugged in.

Edit: yeah it wasn't that lol. Hard shutdown while trying to install Photoshop with the UPS USB cable unplugged.

[mobius]

Edit: NVM, just read your last post.

In any case, I had a service call recently that had a similar situation. The company replaced the PC 3 times, and bought a lot of expensive UPS and conditioning equipment to no avail. Turned out to be a loose connection in an electrical box was causing a low voltage situation under heavier load, which would cause the PC to power cycle (but wouldn't switch the UPS over to battery).

Might be worth your time if you have a multi meter to check the voltage at your outlet while you have a heavy load on it (ie, plug that portable heater in, then measure the voltage from the other half of the outlet).

[Seth1968]

^

Originally posted by Mitsu3000gt
I talked to the Memex Tech again and he said he tested my PSU and found no issues. Unfortunately the issue *did not* happen for the tech at all, but did happen when I dropped it off so the person who checked me in witnessed it.

[Mitsu3000gt]

If I can get out early enough today I'm going to pick up the identical PSU at Memory Express and try that. Should be a quick swap since it's modular and I can leave all the cables in place.

If that doesn't fix it, I guess it's the mobo or some environmental factor I haven't discovered.

My PSU Does have the following:

Heavy Duty Protection Circuitry:
OVP (Over Voltage Protection)
UVP (Under Voltage Protection)
OCP (Over Current Protection)
OPP (Over Power Protection)
SCP (Short Circuit Protection)

I wonder if one of those is bring tripped (either as a malfunction or not) which cuts the power. The "click" I hear does sound like I would expect protection circuitry to.

That doesn't explain why the issue repeated itself when dropped off at Memex though, or how I've been running for a month previously without issues.


The only other thing I can think of that is possibly relevant is that each shutdown from a windows state seems to happen when I initiate a process after it sits for a while. For example, I have it running idle in windows for 1 hr, then i double click the League of Legends icon --> boom shutdown. I have it running idle for a bit, then click on Photoshop --> shutdown. I have it running for a bit then I click in the search box which brings up some options and--> shutdown. It never shuts down from windows just sitting idle, only after I initiate a process. The first time this ever happened was after 2 hours of plex streaming in an otherwise idle state, and I clicked the Firefox icon --> Shutdown

[Mitsu3000gt]

I don't want to jinx anything but I have a possibly interesting development. I've been using my computer for about 2 hours now without issues. Installed Photoshop and set up a bunch of other stuff on the new SSD. Been doing lots of things I couldn't do before (at least not for long). I also transferred 128GB of files to a USB drive successfully.

I turned off the "ECO MODE" switch on my PSU. I plugged in my USB devices to a different row of USB ports than I have been using for the past ~5 years. I also reinstalled my 3 additional storage HDDs.

That's it...so far so good. Any thoughts as to why either of those two things would make any difference at all? Perhaps my PSU still isn't functioning perfectly, but disabling ECO Mode has bypassed some sort of fault? Or maybe nothing is fixed and this is just lucky.

[Seth1968]

Eco mode controls the PSU fan. Although I don't know if it does anything else, or whether it's controlled by load or temperature.

Perhaps in eco mode a fault protection relay is being tripped, and that's what you heard?

[revelations]

I've seen issues on many devices with "eco" mode turned on. From switches, powerbars, to HDDs etc. it seems like a good failure point.

If you now, cannot replicate the problem, I would say you have managed to isolate the issue.
If the issue persists, the next step is to disconnect all peripherals that you can do without, to replicate the issue.

[Seth1968]

Originally posted by Mitsu3000gt
It says "Automatic repair could not repair your PC"

That's reg / file corruption caused by the abrupt shut downs. In such an instance start up repair often fails, but if the drive is slaved, or you can get to a command prompt, then a manual chkdsk usually succeeds.

BTW- When the system is running fine, create a new restore point, as the previous ones may be pooched.

[Mitsu3000gt]

Yeah as far as I can tell, ECO mode just keeps the fan off on the PSU until it hits a certain temperature, I think 55C which it would likely never even hit in my cold basement. It is a physical switch on the PSU and you need to open the PC case to access it. It does nothing else as far as I know or as far as the manual says.

I don't think using different USB ports mattered because it sill died at memex without any peripherals other than the monitor connected.

[Seth1968]

Is it still running ok with eco mode off?

[Mitsu3000gt]

I still don't trust it, but yes so far so good with ECO mode off still (knock on wood).

[Mitsu3000gt]

It died again under the precise circumstances as the original time the issue occurred. Computer was on for hours without issues, I woke it up, and clicked on Firefox (I think i could have clicked on anything) and "click" hard shutdown and it begins the hard power on/off loop. Not sure anymore, I thought I was on to something after I got 2 full days of usage after turning off eco mode.

The only thing I've confirmed is that it will be virtually impossible to diagnose going forward because I've now had as long as 2 days of perfect operation, so any success I might get in the future means nothing until a long time has passed.

Also, the dozen or so programs I installed on the weekend are now gone. I think it did some sort of automatic restore point. They are totally gone, as if I had never installed them.

[revelations]

Grab that PSU.

[UndrgroundRider]

Originally posted by Mitsu3000gt
I don't think it's the PSU because if I can get past certain startup points (there is a pattern of about 10 progressively longer reboots before it lets me into recovery mode) it will run indefinitely. If my PSU was the issue it would be indifferent to what the rest of my computer was doing.

IMHO, stop trying to overthink the failure modes here. The interactions between different components are too complicated to logic out without an intimate low level knowledge of the entire system. For example, you may be hitting over current protection, and the 1/10th time you make it past the loop is because a HDD is delayed spinning up or something of that nature. Or maybe you have a bad solder joint on your mobo that separates after enough thermal expansion and it takes ~10 reboots to cool off and make a solid connection again. Literally there are tens of thousands of ways the system can fail like this.

Your best bet is to swap 1 part at a time until you eliminate the problem.

[Mitsu3000gt]

Originally posted by revelations
Grab that PSU.

One step ahead of you :rofl: Drove in today so I can get one on the way home. Should actually be a very easy swap. Then, *if* it works, I need to let it run for many days through various tasks.

If the PSU doesn't fix it, I believe at this point it has to be mobo or environmental.

Two more kind of strange issues happened

1) My computer won't recognize my eSATA portable/backup hard drive anymore.

2) My MS Office 2007 is now corrupted and won't uninstall or repair itself, so I will have to find some way to delete it or uninstall it without leaving any junk behind.

[Seth1968]

If you don't have the key for Office, then the first thing to do is run Belarc advisor to get the key, then run Revo to uninstall Office. If you don't have the disk for a re-install, you can get Office from here:

https://www.microsoft.com/en-us/software-download/office?culture=en-US&ref=o12

As far as the external drive goes, try another port and / or see if it shows up in the BIOS and disk management.

I'm still leaning toward the PSU as the main culprit, but you've got some really varied symptoms.

[Mitsu3000gt]

Thanks Seth, I'll try Revo uninstall.

I *think* what happened is that it may or may not be the most genuine copy of Office 2007. Windows did a fairly significant update, and then when I try to run Excel or Word, it says the files are corrupted because Windows can't access certain files. I ran this copy for years on my previous SSD without issues, but it's possible the latest windows update did something. Only problem is now I can't uninstall it, so I will try Revo. I've tried the windows uninstaller and also CCLeaner's uninstaller with no luck. Your link has Office 2007 though, so I will try that out first with the Key I have.

Can you even buy MS office anymore without getting the monthly Office 360 plan? I use it like twice a year so I'm not going to pay monthly for it. Maybe I'll just use google docs and forget about it altogether.

As for the external drive, I tried both of my eSATA ports (only have 2), but I haven't checked the BIOS yet so I can try that when I get back up and running.

After one of the hard shutdowns, windows said it was repairing the C drive (my OS drive) and that's when it seemed to revert back to 3 days ago, undoing all of the things I installed over the weekend EXCEPT the Excel 2007 - it kept that for some strange reason, and now I can't remove it.

And yeah, if it's not my PSU I guess it's my mobo. I did some looking and I don't think I can even buy my mobo anymore, or any LGA1155 mobo's, so I will probably have to buy a new mobo/cpu just to see if it's the mobo. I won't be too upset if I need to do that, but I was hoping to wait until 2H 2017 to buy a full new system.

[Seth1968]

Yes, you can buy older versions of Office from sites like Amazon. BTW- I still use Office XP even on my W8. Installed the office Compatibility pack to open DocX and it's good to go :)

One other thing you can try other than swapping the PSU. Disable the auto restart on system failure. It should then show a BSOD instead of restarting.

PC Properties>Advanced System Settings>Settings, and you'll see the restart the option.

[Mitsu3000gt]

I did actually try disabling auto-restart on system failure pretty early on in my troubleshooting process and unfortunately it did absolutely nothing. The hard power-downs and power-ups seemed to ignore that setting regardless. I feel like this also points toward the PSU, since it doesn't matter what PC settings I have selected if power is being physically cut.

My biggest fear is that the PSU is working just fine, but something to do with the house/wall power is tripping a protection circuit. Nothing else plugged into those outlets though have any issues whatsoever.

I'll install an identical PSU this evening and we will see I guess. I have 7 days to test it before it has to go back to Memex.

[Seth1968]

Yep, that's why I said, "should", but a better term would have been "might".

Windows ignoring the option typically means Windows doesn't even know it due to component failure.

[Mitsu3000gt]

Yeah I am 99.999% sure it's hardware ever since the new SSD didn't fix it.

Even if the wall power isn't great, it should be getting conditioned through my UPS which has these features:

Line Interactive Topology
Exists when a line interactive UPS has an autotransformer that regulates low voltages (e.g., brownouts) and over voltages (e.g., spikes) without having to switch to battery.

Automatic Voltage Regulation (AVR)
Provides clean, consistent AC power by automatically increasing (boosting) low voltage without using battery power when incoming utility power experiences minor fluctuations.

Fingers crossed on the PSU swap.

[Seth1968]

You tried it without the UPS right? Although it could still be a line issue.

The other factor is it happened once at Memex (that we know of). I say that because like any repair tech, we shudder at intermittent issues.

[Mitsu3000gt]

Originally posted by Seth1968
You tried it without the UPS right? Although it could still be a line issue.

The other factor is it happened once at Memex (that we know of). I say that because like any repair tech, we shudder at intermittent issues.

Yeah I tried two different PSU power cords, with the UPS but in a non-battery backup slot, without the UPS straight into the wall, and without the UPS straight into the wall using an extension cord to an outlet I am pretty sure is on a different circuit. Also, as you brought up, the issue happened at the service desk at Memory Express when I dropped the computer off for diagnostics, which theoretically would rule out my house power being the issue. Unfortunately the issue did not happen on the tech's bench a day or two later when he ran all the tests.

Only thing I haven't done is move it to a different house, and see if it works. Problem there is if it works, I won't know for many days if it's just behaving temporarily or not. Also, it worked for over a month at my house without issue prior to the shutdowns.

[Mitsu3000gt]

Running on a new, identical PSU now (EVGA Supernova 650 P2).

The more I think about it and exactly how it happens, the more I think it's my mobo. Fingers crossed the PSU works but I'm not too optimistic...stay tuned...

[Swank]

When I had flakey power issues it was the mobo too. Every once in a while the power button would do nothing, and it wasn't the power button itself. I'm really curious about the outcome here.

[Mitsu3000gt]

We'll see what happens throughout the rest of the week. I will do my best to re-create scenarios where it failed (coming back to it after a long period of non-use, and clicking any application).

My UPS Software actually tells me the voltage, power quality, load/wattage, etc. and that all looks great so I'm pretty confident it has nothing to do with the wall power either.

[cam_wmh]

Originally posted by Mitsu3000gt


Yeah I'd be curious.

I keep circling back to power supply just because of the nature of the hard power cuts, but then enough other stuff happens that makes me think it can't be the PSU :nut: Mine would also crash (hard power-off) if you tried to select safe mode, but would stay in the safe mode options screen indefinitely, which would suggest it is not the PSU haha.

My PC has been good to me for 5 years, I can't ask much more of it. If it turns out I need a new one I won't be upset, but I remain optimistic for the time being. Maybe it just needed a different environment for a while at MemEx haha who knows. It's probably something silly still.

MemEX rooted it to Win10 update.

[Mitsu3000gt]

Originally posted by cam_wmh


MemEX rooted it to Win10 update.

Interesting so his was software then I guess. Mine is 100% for sure hardware unfortunately.

[Mitsu3000gt]

Power is at the moment working fine but I have a new problem. Windows is completely fucked. It worked fine yesterday and I did a normal shutdown last night. Started it up after work today and I get this error:

"Recovery" a required device isn't connected or can't be accessed. Error code 0xc0000000e. You will need USB recovery tools blah blah blah"

When it restarts it "collects some data" and then restarts into the above message in a loop. "Stop Code"
0xc000021a

I can't get past those screens even in safe mode.

My feeling is something got corrupted in the last power cut before I replaced my PSU because when it did an auto repair, a bunch of programs were fucked. They wouldn't uninstall, windows thought they were there when they weren't and vise versa and I had to revo-uninstall a ton of stuff and reinstall it. After that all seemed ok and I shut down normally. At best it's looking like I need to do a fresh windows install again.

Only other thought is that there's a bad windows update floating around.

EDIT: I disabled "Driver signature enforcement" and it worked....now to find out why...

Windows is also hung on this update:

Cumulative Update for Windows 10 Version 1607 for x64-based Systems (KB3206632) - Error 0x800f081f

EDIT 2: So it appears my computer works as normal except for the following 4 issues:

1) It will not boot without disabling driver signatures. I get the "stop code" error twice while it "collects data" and then the third restart gives me the boot options to disable driver signatures.

2) Windows update KB3206632 won't install. I've tried forcing windows to re-download it by deleting the software distribution Windows file contents but no improvement.

3) One of my spinner HDD's wont write data any faster than ~12.0 MB/s anymore. Confirmed with Crystal Disk. This is a brand new problem. Curiously, Samsung Drive Magician says this hard drive "SATA 3.0 Available" under a little caution exclamation mark which does not appear on any of my other 3 drives. Maybe one of the hard power cuts damaged it.

4) Nothing I do will make my eSATA ports work again (SOLVED: Rebooted with external drive running and windows "repaired" it, now it is discoverable)

5) Every single time the computer starts up, it gives me a prompt about Windows OOBE.

Tried updating mobo chipset drivers, didn't do anything. BIOS is the latest available version. Probably going to have to try do a full system restore again. Or maybe I should just put my old SSD in and see if that puts me back to normal...

[bookem]

Sounds like the mobo might be f'ed.

[Swank]

Originally posted by Mitsu3000gt
Maybe one of the hard power cuts damaged it. I wouldn't doubt it. Storage drive or not, everything spins up when your PC powers up, then you set up power profiles to shut down certain discs after x amount of idle time (or let the default power settings do it). I'm pretty sure the lowest you can set is 1 minute (like my storage drives) and your reboot loops are almost certainly shorter than that. Back up that drive that is breaking ASAP and keep all HDDs disconnected as much as possible until you get to the bottom of this gremlin.

And your new windows problem, holy hell, the plot thickens :banghead:

[Mitsu3000gt]

Originally posted by Swank
I wouldn't doubt it. Storage drive or not, everything spins up when your PC powers up, then you set up power profiles to shut down certain discs after x amount of idle time (or let the default power settings do it). I'm pretty sure the lowest you can set is 1 minute (like my storage drives) and your reboot loops are almost certainly shorter than that. Back up that drive that is breaking ASAP and keep all HDDs disconnected as much as possible until you get to the bottom of this gremlin.

And your new windows problem, holy hell, the plot thickens :banghead:

Yeah I *think* my power issue is gone (I've been using it enough that I would expect it to have happened by now) but I won't be sure until it works perfectly for a week or so. I have already started backing up that HDD, just in case it's physically damaged and not just a driver/mobo/SATA issue.

Yeah I'm pretty tired of this. I've probably done a lot more than many would have patience for at this point. I'm going to just put in my old SSD and see if that works now that (hopefully) the power issue is solved. If that doesn't work I'm going to fresh install windows again on the new SSD. If that doesn't work I am going to buy a new computer lol.

Even more annoying is that I have until Monday to return the new power supply to Memory Express, but I can't initiate a RMA with EVGA on the old PSU until I am sure the PSU is the issue, because I have to pay ~$40 to ship it back to EVGA. So there will be a period of time where I am completely without a PSU, probably for a week or so.

[Mitsu3000gt]

Put the old SSD back in and things seem to be OK, assuming no more hard power-downs. I can test it again tonight in the same scenario that it likes to reboot in the most.

Hopefully, after all this, I just have to get a RMA PSU. If that's actually what failed I will be a little bit surprised.

Dec 20 EDIT: 5 days later still no issues on the new PSU.

EVGA customer service is still amazing. They said they wouldn't get anything processed or shipped until the new year because they closed on the 23rd, and that the replacement would come via UPS ground shipping. Turns out they did process my RMA, and shipped it to me via overnight shipping :thumbsup: Glad to see they are still keeping up their legendary customer service. I'm really surprised my PSU failed though, given that it was new and they are some of the best ones available but I guess everything has lemons.

I'm just happy it wasn't my mobo, or so I think (knock on wood).