Rebooting during games (no BSoD or dumps)

[Pill Monster]

Yes, my BIOS settings are the defaults. I just got my hands on a COOLMAX CUG-950B (950 W) PSU, and installed it. I'm posting this from the computer in question. I'm going to run the logs and play Path of Exile for a while, and see what happens. I will report back in a bit.
EDIT: Also, after cleaning the computer again earlier, I noticed that the (original) PSU fan was not spinning while the computer was running. Do those not always spin?

Coolmax? Jeez no wonder the voltage is all over the place, Coolmax are absolute junk. lol

 

[tabascosauz]

@Dark_Oppressor You mean the HX850? That's very concerning. AFAIK the HX850 has no technology that allows it to have a semi-passive fan profile.

If that's indeed the case, then we may have found our problem. The PSU may have reached 50 degrees and shut down when components are under load. Only fanless PSUs are designed to operate under load with no fan spinning.

One way to test the PSU (I'm assuming it's out of the case now) is to short the 24-pin. For me, it's very easy because EVGA provides a 24-pin jumper that plugs into the end of the 24-pin and turns the power supply on without actually plugging the PSU into anything but the wall. You can accomplish the same trick with a well-placed paper clip; just don't injure yourself.

EDIT: Never mind. PSU is designed as semi-passive. Fan is probably not the problem. But here's an idea - take your rig out and put it on a table. Then as you fire up a stress test on CPU + GPU, see if the fan starts up. If it doesn't, then maybe it is the problem.

 

[Dark_Oppressor]

Just to clarify:
The CORSAIR HX850 is my main PSU. Its fan seemed to not be spinning when the computer was running and I checked it earlier tonight. It is currently suspect #1 for my crashing problem.
The COOLMAX CUG-950B is another PSU that was in my wife's computer (which I built about 7 years ago, but I guess I didn't pick a good PSU...) I have swapped the two, and this PSU is now in my computer, and I am testing with it to see if another crash occurs.
So the logs that I posted previously are with the CORSAIR HX850. I will post some logs later tonight for the COOLMAX one.
Let me play a little longer and if it doesn't crash, I'll leave the game running and find out if the HX850 fan spins at all. I have the other computer (that I borrowed the CORSAIR PSU from) open right next to me, and I could actually easily just plug it into that one's mobo and see what it does.

 

[tabascosauz]

Just to clarify:
The CORSAIR HX850 is my main PSU. Its fan seemed to not be spinning when the computer was running and I checked it earlier tonight. It is currently suspect #1 for my crashing problem.
The COOLMAX CUG-950B is another PSU that was in my wife's computer (which I built about 7 years ago, but I guess I didn't pick a good PSU...) I have swapped the two, and this PSU is now in my computer, and I am testing with it to see if another crash occurs.
So the logs that I posted previously are with the CORSAIR HX850. I will post some logs later tonight for the COOLMAX one.
Let me play a little longer and if it doesn't crash, I'll leave the game running and find out if the HX850 fan spins at all. I have the other computer (that I borrowed the CORSAIR PSU from) open right next to me, and I could actually easily just plug it into that one's mobo and see what it does.

refer to bottom of the post above yours

On a table, you can easily see the fan. Furthermore, if the fan doesn't spin and heat builds inside the PSU, you can feel it rise up through the fan, in the same way that a fanless PSU works.

 

[Pill Monster]

Just to clarify:
The CORSAIR HX850 is my main PSU. Its fan seemed to not be spinning when the computer was running and I checked it earlier tonight. It is currently suspect #1 for my crashing problem.
The COOLMAX CUG-950B is another PSU that was in my wife's computer (which I built about 7 years ago, but I guess I didn't pick a good PSU...) I have swapped the two, and this PSU is now in my computer, and I am testing with it to see if another crash occurs.
So the logs that I posted previously are with the CORSAIR HX850. I will post some logs later tonight for the COOLMAX one.
Let me play a little longer and if it doesn't crash, I'll leave the game running and find out if the HX850 fan spins at all. I have the other computer (that I borrowed the CORSAIR PSU from) open right next to me, and I could actually easily just plug it into that one's mobo and see what it does.

Ah that makes more sense, was wondering why you had Corsair in the specs.
Even so there's a lot of ripple in the graph, not just from the PSU. and the last 20mins show the GPU and CPU are working quite a but harder. At shutdown the power related charts all have a big spike or a drop, including the 12v rail. And the network load went up too. A What were u dong in the last half hour of the game?

 

[Dark_Oppressor]

OK, I just finished playing Path of Exile for about 8.5 hours with no crashes.
I also watched Netflix on another monitor, and had several tabs open in Chrome on a third monitor while playing.
Here are the logs: http://cheeseandbacon.org/temp/log-gpuz-2.txt, http://cheeseandbacon.org/temp/log-hwinfo-2.txt

Pill Monster: During that last half hour in the first test, I was mostly sitting in a town, looking at my stats (it's an action RPG).

I haven't looked at the other PSU any more yet, as I was busy playi-- testing... It's 0430 here now, so I'm going to get some sleep. Tomorrow I will consider testing it, perhaps if you lot think it would still be helpful.
Am I right in thinking that this test seems pretty helpful? The fact that I could play 4 times as long as I ever could in the past without crashing seems to point pretty heavily towards the PSU.

 

[Vayra86]

Playing one specific game that tripped your reboot earlier may not be sufficient as a 'test'. It does give some indication, but that's about it.

I'd say, run a bench/test series of Valley or 3DMark a couple times over to really stress your system with the new PSU, especially how it performs at high load. 3DMark also provides some pretty neat data with the bench for monitoring afterwards.

 

[Pill Monster]

OK, I just finished playing Path of Exile for about 8.5 hours with no crashes.
I also watched Netflix on another monitor, and had several tabs open in Chrome on a third monitor while playing.
Here are the logs: http://cheeseandbacon.org/temp/log-gpuz-2.txt, http://cheeseandbacon.org/temp/log-hwinfo-2.txt

Pill Monster: During that last half hour in the first test, I was mostly sitting in a town, looking at my stats (it's an action RPG).

I haven't looked at the other PSU any more yet, as I was busy playi-- testing... It's 0430 here now, so I'm going to get some sleep. Tomorrow I will consider testing it, perhaps if you lot think it would still be helpful.
Am I right in thinking that this test seems pretty helpful?

Yeah for sure. Power was the main suspect anyway, which u just basically confirmed by swapping out PSU's.

After 8hrs stable you'd be a bit silly to keep fixing the issue when there's nothing to fix.....


If it was me I'd call it it a day.

 

[Vayra86]

Don't know man, I've been there too, playing for a whole week without anything going wrong, and on day 8 I saw another reboot. I would always put the system under maximum stress with a new PSU/replacement PSU, even a new store-bought one.

 

[RejZoR]

If system reboots during load with no BSOD it's PSU issue in 99% of cases. Get an entirely different PSU that can run your system and test with that one. If it works, you've confirmed the remaining 1%...

 

[Bill_Bright]

If system reboots during load with no BSOD it's PSU issue in 99% of cases.

99% ?? Ummm, not in my experience. Often, yes. But it could also be heat. Or sometimes RAM. Or leaky caps. Or any number of other things. No BSOD (or Event Log entries) simply means the crash was so sudden, Windows didn't have time to toss up a error or write a log entry.

Wait! When I said earlier the logs did not state CPU temps, we got a reply saying they never went over 50°C. But I just realized that Pill Monster said that, not Dark_Oppressor. How do you know that, PM?

I ask Dark_Oppressor, what are your CPU temps?

As far as that Coolmax being a good PSU to test with, I think it is an excellent PSU to test with. If your system remains stable with this poorly rated Coolmax PSU, then that would even further suggest the other PSU was your problem.

 

[RejZoR]

When RAM was acting up I was mostly getting weird memory BSOD's. I don't remember total restarts.

 

[Caring1]

The Corsair HX850 may be a simple fix with a replacement fan, if that is the only fault.

 

[Bill_Bright]

The link to his CPU temps is in the first post of this thread.

Now I see it. I did not realize the values were jammed together and not under the headings.

Thanks.

 

[Pill Monster]

99% ?? Ummm, not in my experience. Often, yes. But it could also be heat. Or sometimes RAM. Or leaky caps. Or any number of other things. No BSOD (or Event Log entries) simply means the crash was so sudden, Windows didn't have time to toss up a error or write a log entry.

Unexpected shutdowns are nearly always due to heat or power. No WER or .dmp file means the crash wasn't triggered by the OS, in cases where the CPU throws an MCE before Windows can write a dump, a WER log will be written on the next successful boot.

Leaky caps are pretty much a non event these days as most boards have solid caps..

As far as CPU overheating, the chances of a 4Ghz CPU reaching a shutdown temp of 90c with Turbo enabled while running a game which barely loads 4 cores, is.....extremely unlikely......

Wait! When I said earlier the logs did not state CPU temps, we got a reply saying they never went over 50°C. But I just realized that Pill Monster said that, not Dark_Oppressor. How do you know that, PM?

Because I downloaded the logsheets then converted both to CSV format so I look at them, as I said earlier.

I ask Dark_Oppressor, what are your CPU temps?

As far as that Coolmax being a good PSU to test with, I think it is an excellent PSU to test with. If your system remains stable with this poorly rated Coolmax PSU, then that would even further suggest the other PSU was your problem.

That doesn't make it an excellent PSU. And the more likely scenario would be a system which crashes with 2 different PSU's, making it even harder to diagnose the cause.
"Well hey it shuts down even after we switched PSU's - guess we can rule out the Power Supply - next:"











.

 

[Bill_Bright]

That doesn't make it an excellent PSU.

Ummm, I didn't say that.

I said, (and you quoted me!) "excellent to test with". I went on to say, and again you quoted me, "this poorly rated Coolmax PSU".

 

[Dark_Oppressor]

The reason I am testing with Path of Exile is that that's how I've been able to reproduce the crash. I tried running some benchmarkish things (occt, furmark) but they never caused any crashes. However, I think the longest I ran one was 30 minutes, so that may not mean much.

For now, I'm thinking I'm going to just replace the PSU. Now to see if I can get it replaced under warranty... I've never done that before.

 

[Pill Monster]

Ummm, I didn't say that.

I said, (and you quoted me!) "excellent to test with". I went on to say, and again you quoted me, "this poorly rated Coolmax PSU".

No, it's not, for the reason I stated above. Yes it is poorly rated which is why you don't test with it.
The purpose of swapping hardware is to eliminate a possible faulty part by testing with a KNOWN GOOD component. Not one which may or may not be stable......

 

[Bill_Bright]

Yes it is poorly rated which is why you don't test with it.

Now wait! Now you are trying to spin your original misquote with obfuscation.

We are talking about a specific test scenario on the OP's system. His old PSU was suspected of causing problems. Just because that Coolmax model was poorly "rated" under very high load scenarios IN NO WAY suggests it cannot be used to test to see if it causes the exact same problems as the OP's old PSU did.

Of course, if you just happen to have a spare PSU laying around that got a perfect 10 out of 10 you would use that instead. But if not, you use what you can. Swapping in the Coolmax to see if the exact same problem occurs is still a valid test.

 

[Dark_Oppressor]

OK, sorry for the very belated reply, but I wanted to make sure everything was OK before posting again.

I replaced my PSU about a week or two after my last reply (I was able to RMA it). Since then, everything has been great, so I feel pretty confident at this point that the PSU was the problem.
Just wanted to report my results for posterity.

Thanks again for the help, all!

 

[Bill_Bright]

Great! I am glad you got it sorted out and thanks for posting your followup!