Multiple errors, all started simultaneously

DataSnake

Distinguished
Jul 19, 2011
36
0
18,530
I just built a new budget gaming rig. For about a week, it worked fine, but now it's having several different problems, all of which started at about the same time. I don't see how they could all be connected, but given their timing, I thought I should post them all in the same thread just in case.

1. Sometimes it just freezes. Can't move the mouse, Ctrl+Alt+Del does nothing, have to reset. Transferring lots of data to/from my USB hard drive is almost a surefire way to trigger this, but it happens other times when the drive isn't even connected.
2. Sometimes it partially powers down. The fans are still on, including the GPU fan, but the monitor is in orange-LED power saving mode, and it won't respond unless I reset it. This usually happens either right after boot or if I leave it unattended for more than a few minutes, but has happened a couple of times while I was using it.
3. Sometimes it resets itself without warning. Other times it starts to reboot, gets as far as the GPU BIOS ("GeForce GTX 465 version blah blah blah"), then goes into the half-off state from problem 2.
4. After the above problems, there is about a 50-50 chance of the reset button only taking it to the half-off state from 2, sometimes without even showing the video BIOS screen, and the power button doing nothing, even when held down for 10 seconds.
5. I usually hibernate it when not in use, but after it reset itself while entering hibernation, it has been unable to resume properly. I thought that might be because of a corrupted file, so I got it booted, selected hibernate, and let it power off (which should have overwritten the bad hibernate data), but when I went to resume, I got as far as the Windows logo and "Resuming Windows" message, then it cut to a black, text-mode screen with a flashing cursor in the top left corner.

Relevant info:
*I had been running with the multiplier increased to 20 and vcore to 1.45, which was stable all last week, but as soon as the problems started I dropped it to stock multiplier and voltage
*I had been using MSI Afterburner to overclock the GPU to 750/1500/1700 with fans at 100%. I tried with fans set to scale up on their own once, and it froze. All the problems started after that, but I've dropped the clocks to stock (607/1215/1603) and set the fans to 100%, and it still isn't stable. I don't think the problem is the GPU, though.
*Another thing I changed before it all went pear-shaped was that I installed RivaTuner, although I only ever used the "d3doverrider" tool. I uninstalled it as soon as the problems started, just in case.
*The errors are essentially random. Sometimes I can play Crysis for hours on end, other times it freezes solid just browsing the web.
*I have reset all BIOS values to their defaults, no luck
*I enabled speedfan's logging feature, and the core and GPU temps right before a freeze are never above 40 and 50, respectively
 

DataSnake

Distinguished
Jul 19, 2011
36
0
18,530
One other thing that might be relevant, I've been using ASRock's "app charger" utility to charge my iPod touch while the computer was suspended. I started doing that about as soon as I got the computer, though, and the problems only started this weekend.
 

DataSnake

Distinguished
Jul 19, 2011
36
0
18,530
Specs are the same as in my sig:
*Rosewill Challenger case
*Phenom II x4 965 C3
*Cooler Master Hyper 212+
*Galaxy GeForce GTX 465
*8GB (2x4GB) G.Skill Ripjaws DDR3-1333 CL9
*1TB Samsung Spinpoint F3 SATA II
*Rosewill Green Series RG630
*Windows 7 Ultimate x64 SP1
 

AdrianPerry

Distinguished
The PSU actually has some good reviews and the 80PLUS rating has been tested to be fairly efficient so your PC shouldn't be crashing out due to not enough wattage.

As much as this seems like it would be a PSU failure from the problems you listed at the start, it could be due to some overclock issues. Have you tried running you PC with everything back at stock specifications and voltages?

Have you kept an eye on temperatures and CPU usage when the system crashes? Does the CPU suddenly hit 100% and lock up then restart?
 

DataSnake

Distinguished
Jul 19, 2011
36
0
18,530
Currently running everything at stock (except my GPU fans, which I have set to 100% in Afterburner; GPU clocks are stock, though), still crashes. A couple of the freezes happened when I had SpeedFan open in graph mode, CPU and core temps never broke 40c. The other times, and the resets/shutdowns, I've checked speedfan's log and found the same thing. Like I said, there doesn't seem to be any correlation between how much load it's under and when it crashes; it can run any game I own for half the day, or it can freeze or crash from just opening a new window, or even not doing anything.
 

DataSnake

Distinguished
Jul 19, 2011
36
0
18,530
Also, just because the PSU is generally good doesn't necessarily rule out this particular one being a lemon, but I'm a little surprised PSU failure could be the culprit for two reasons:
1. Sometimes it freezes, but what was on screen stays on.
2. Even when it powers itself partway off, all the fans (including the ones with LEDs) stay on.
If those are things PSU failure can cause, my best bet may be to RMA it (I only got it 10 days ago, so newegg should replace it without much hassle).
 

AdrianPerry

Distinguished
There's a few things it could be then really; RAM, Motherboard, CPU or PSU.

Since the fans keep spinning even when your PC shuts down, this can probably eliminate PSU.

Your best option is to try running MemTest and see if it brings up any errors.
http://www.memtest.org/
 

DataSnake

Distinguished
Jul 19, 2011
36
0
18,530
Alright, I'll do that. Also, I looked at my last post and it seems a bit poorly worded. To clarify, the fans DO turn off when I shut down the computer, but they don't when it just decides to shut itself down.
 

AdrianPerry

Distinguished


It's ok, I understood that :)
 

DataSnake

Distinguished
Jul 19, 2011
36
0
18,530
Gonna run memtest86+ when I get home, if it'll stay on long enough. Could RAM be responsible for both the freezes and the shutdowns, or will I have to look for a second culprit even if memtest finds a problem?
 

DataSnake

Distinguished
Jul 19, 2011
36
0
18,530
I don't have a spare power supply. I guess I'll just use the one I've got and hope for the best. If the test doesn't run, I think I'll just take the shotgun approach and RMA both parts. I just hope it doesn't turn out to be the CPU or motherboard, installing those wasn't exactly fun. Any way to rule them out?
 

DataSnake

Distinguished
Jul 19, 2011
36
0
18,530
One other thing I forgot to mention: the vcore I read under load in speedfan is usually about .05v higher under full load than I set it to (Set it to 1.45, it gets up to 1.5; set it to 1.4, it gets to 1.45). Is that relevant?
 

DataSnake

Distinguished
Jul 19, 2011
36
0
18,530
It was doing that with the voltages before I started overclocking, too, so if it IS a CPU problem I don't see how it can be my fault. Temperatures don't seem too high (never seen them break 50 stock, and they only hit 56 at 4 ghz under prime95's "maximum heat" large FFT test) , unless speedfan is underreporting them. Pretty sure that only happens on core 2 processors, though.
 

AdrianPerry

Distinguished
IF it is the CPU, i think a replacement would be required. Or maybe just a proper re-config.

Replacing the heatsink would only be necessary if temps were unusually high.

Please wait for some more feedback though from someone more experienced, dont take my word for it the CPU is at fault.