An odd issue

[SuperDuperRon]

Hello all, I’m having a bit of an odd problem that seems to go away for a week and then return. Last month I begun looking into overclocking my Phenom II. The guide I found most helpful and was the one I followed was Dolk’s guide (http://www.overclockers.com/forums/showthread.php/596023-Dolk-s-Guide-to-the-Phenom-II) . Since this was my first time overclocking I didn’t want to push it too hard so I only overclocked it from 3.2 to 3.8 Ghz by making my CPU ratio 19 and bumping up the CPU-NB and HTT to 2200. I ran tests with prime and everything seemed okay, my temps didn’t go above 40 degrees and no crashes. I ended up sticking with those setting, they seemed to work just fine, my games ran better and I didn’t have to change any voltages.

Okay so about two weeks later my power supply (Diablotek 500W) died, a thermistor inside blew. My power supply was about 4 years old so I wasn’t sure if it was just old or had something to do with my overclocking. I bought a new PSU (Thermaltake 750W 80+ Gold) and installed it Jan 5th. I also ordered a new thermistor to try and repair my old PSU as a backup, which I now have and works just fine. Well I went to use my computer (which was on over night like usual) on the 16th and the mouse or keyboard weren’t working, also there was a “windows recovered from a critical error” message. I held in the power button till it powered down, a waited a few seconds and turned it back on.

The fans all came on and all the LED’s were lit but there was no “beep” and nothing came on the display. Plus I couldn’t turn it off, not even by holding in the power button. I had to flip the switch on the PSU to get it to shut off. I messed around with it for about 20 mins removing all sorts of things, eventually it started working again but I’m not sure why. Well even though I got it to boot it kept freezing (in windows) after around 10 mins.

I went back into the BIOS and set everything back to factory and that seemed to fix everything. I figured I’d wait a week or two before trying to overclock again just to see if the problem was fixed. Ofcourse it wasn’t though lol. It worked fine all week, I was playing mass effect all week with no problems. Then last night I passed out watching netflix, when I woke up this morning I had the same keyboard/mouse won’t work problem again, minus the “critical” windows message. Had to hold in the power to get it to shut off again, then when I turned it back on the same problem came up, no beep, no display, holding power won’t shut it off but all the fans and lights come on. I flipped the switch on the PSU and left if off for about 10 mins, then tried again but still had the problem.

So I unplugged everything from it and cleaned a space to my table to start taking it apart. I came back to it to move it to the table but decided to give one more go before I start taking things apart. I pluged in the power cord and the HDMI cable and it came on, no problems. I’m guessing unplugging the power cord let all the power drain out of it where just flipping the switch to “off” didn’t quite do the trick. Still doesn’t answer why this is happening in the first place though.

Hoping that this is not an issue with my motherboard due to my overclocking I swapped my new Thermaltake 750W PSU for my old (and now repaired) Diablotek 500W PSU. The PSU is the only hardware change I’ve had since September.

I’ve been running the computer with the 500W PSU for a couple of hours now with no issues. Now I’m just going to leave my pc on and see if the issue comes up again. What do you think? Is it possible I screwed up something when I overclocked, or when my first PSU died it screwed something up? Or maybe it’s an issue with the new PSU? Any help is greatly appreciated.

AMD Phenom II x6 1090T Black Edition
Asus M4A89GTD-Pro Motherboard
Nvidia GeForce GTX 660
PNY Dual Channel 4096MB (2x2048MB) PC12800 DDR3 Memory - 1600MHz, Non-ECC, Unbuffered
Thermaltake Water 2.0 Extreme cpu cooler
Windows 7
PSU 1 - Diablotek PSDA500 DA Series ATX Power Supply - 500W
PSU 2- Thermaltake Toughpower 750W Power Supply Unit - ATX, 80+ GOLD, Modular

120GB Intell SSD
500GB SATA Hard drive
80GB IDE hard drive
sata dvd drive

Thermaltake Chasar A71 case
Case Fans: one 120mm rear fan
One 200mm Front fan
One 200mm side fan

 

[Mandrake4565]

I would bet the Overclock isn't the issue. When a Psu goes sometimes it wreaks havoc on the system. It is possible when it blew it hurt some other component. It is tough to trouble shoot exactly which one unless you have another system that you can swap parts into to see if anything causes the issue you're having. I can tell you I had a similar issue happen with a Psu on a AMD AM2 motherboard. It started randomly crashing shortly after replacing the Psu and not being able to start. Eventually the motherboard bit the dust all together. Your best bet is to leave it a default settings for the time being too see if the problem reappears with the rebuilt Psu in it.

 

[Lvcoyote]

If you find you get the same problem with the repaired PSU, maybe run a memory tester just to make sure the memory wasn't damaged during the first episode of PSU failure.

 

[Itachiii]

What the people above said. I would actually run a blend test on P95 for 2 hours. That will tell you if the CPU and RAM are in a good shape. If you get an error start testing each module seperately

 

[SuperDuperRon]

Okay cool, thanks for all the advice. It's been on for 2 days now with no trouble, though the issue seems to take a week to surface but so far so good. In the mean time i'll test the memory and run P95 too. Keep ya posted either way, thanks again.

 

[SuperDuperRon]

Hey again, so this morning before i walked out the door for work i checked to see if my pc was froze up and sure enough it was. I realized the other day that i should look at the system clock when it freezes next. The system clock was froze at 7:46, real world time was 8:05. I didn't have time to mess with it though so i left it on and figured I'd hit the problem when i got home. When i did get home i went to check it again and it wasn't froze anymore, and the system clock was at the correct time again. I really wanted to see if the freezing was more of a software issue and the "no beep, no display" when I'd try to restart the system had more to do with the new psu, but since its working i can't test it now lol. I did however shut it off anyway just to see what would happen but it booted up just fine. Weird.

Also i ran P95 for two hours as well as running a memory test (mdsched.exe) and didn't have any issues.

 

[Itachiii]

Well that's good news I guess. Can you download HWmonitor? Open it, use ur pc normally and just before you shut it down for the day, take a screenshot of it maximized. Maybe the power delivery is wrong from time to time

 

[SuperDuperRon]

Yeah sure i have that, i'm not sure what the power delivery should be, but under "Vbat" it's 3.456v. Also when i turned it on today if froze up after 3 mins of being on. So i held in the power to shut it off and when i turned it back on it booted up just fine. I do still have the rebuilt power supply in so idk if that had anything to do with it. Now I'm curious if software is cause the crash. I bought an ssd back in September and cloned my existing drive on to it. I'm gonna throw in that old drive so i don't have reinstall anything but can still check to see if I'm having some sort of software conflict.

 

[Dlaw]

It very well could be software, but it sounds to me more like your PSU. I had similar issues with a failing PSU, swapping it out took right care of it.

 

[bob4933]

Have you done anything with your video card drivers lately?

Your freezing sounds exactly like what happened when I got my 290 in last week. I wound up getting a higher psu and installing more stable older drivers; that seemed to have solved my problems.

 

[SuperDuperRon]

According to device manager my drivers are from 12/13/14. but i do remember there was some kind of geforce up date not that long ago, like around two weeks ago. i don't think it was for my video cards drivers, i believe it was for the "geforce game ready drivers". There is an option to update them again, should I just up date everything and see what happens?

 

[Itachiii]

According to device manager my drivers are from 12/13/14. but i do remember there was some kind of geforce up date not that long ago, like around two weeks ago. i don't think it was for my video cards drivers, i believe it was for the "geforce game ready drivers". There is an option to update them again, should I just up date everything and see what happens?

If I were you I would format my drive. That would at least take out the software problem.

 

[SuperDuperRon]

Okay, yeah it blue screened on me this morning, something like "paging error in a non paging area". Then it froze on me a little while ago and restarted on it's own. Maybe it's the ssd. In any case i'm going to format it and reinstall windows, crossing my fingers that fixes it.

 

[jasonlylevene]

Let's get an unaddressed issue dealt with here.

The Diablotek is very likely toast. Without a scope to see you can't really tell. The units are generally built with cheap capacitors. 4 years is beyond their age. There could be some units I'm not familiar with built but better sources, but I doubt it.

You might be able to tell if you examine the capacitors in the PSU. If the tops aren't flat, they're cooked.

The unit is likely generating electronic noise on the rails. That will cause unpredictable behavior, overwork the caps and VRMs on your motherboard and other devices, potentially causing damage if left in use long term.

Just because a PSU generates power doesn't mean it's working.

A failing PSU can cause unpredictable disk operation, which in turn can corrupt an OS installation. At that point you have layers of issues which can't be isolated. Your approach to diagnosis assumes only one part, a PSU, is at fault when at this point several issues could be overlapped, and the PSU subsistution generates false impressions as to what the problem is.

Use the new PSU (it's a much better unit to begin with). Then, isolate issues from each other. I'd suggest you consider a new OS installation on a spare drive or spare partition to eliminate that as a potential source of problems.

Then, approach the rest of the machine with the knowledge that any one of the components may have been damaged by the previuos PSU, and isolate them for testing one at a time.

Only lightning strikes are more puzzling than investigating a machine that's been powered by a failing PSU over time. You might be tempted to think the PSU only failed when the thermistor blew, but what I'm telling you is that power supplies degrade gradually over time, without obvious symptoms. It may have been driving your system nuts for quite a while before the thermistor blew.

 

[SuperDuperRon]

Hey all just checking back in, i clean installed windows on the 29th and so far it seems to have fixed everything. I just wanted to thank everyone for their help.

 

[Itachiii]

Hey all just checking back in, i clean installed windows on the 29th and so far it seems to have fixed everything. I just wanted to thank everyone for their help.

Np mate! Have fun