PC Crashes, Monitor Losses Signal, PC Stays On, Power and Reset Buttons Stop Working

[Lxno78]

________________________

The Problem:
________________________

I just built my computer about a month ago, I've only turned it on and used it maybe 15 times. Twice now in the same day the computer will freeze (both times at the log in screen), then the video signal stops, the PC stays on, but the power button can not be held to turn it off and the reset button does not work. I have to shut the computer off manually.

___________________________

Useful Problem Details:
___________________________

- This problem doesn't occur every time I boot up my computer, which seems it will make it hard for me to test everything.
- When the computer boots up fine, it seems to run stable and there is no slow down.
- Nothing on my computer is overclocked.
- My computer is not running hot. I have used AMD Overdrive solely to monitor CPU temperatures and they are fine. The GPU is fine as well
- The last two hardware changes I made, before the problem occurred, where putting in a graphics card and replacing the power supply unit.
- The graphics card does not require additional connection to the PSU.

___________________________

System Specs:
___________________________

- Motherboard: msi, 760GMA-P34 (FX) series [MS-7641 (v5.1)]
- CPU: AMD FX-6300 Vishera 6-Core 3.5GHz
- RAM: 8GB Klevv Neo
- PSU: Sea Sonic S12II-520Bronze Power supply - 520 Watt
- GPU: Zotac GeForce GTX 750 1 GB GDDR5
- Case: logisys 368RB
- Operating System: Windows 8.1


___________________________

What I have Researched so Far:
___________________________

- 'No Signal' from monitor. Power button won't work when turning computer off.
- No signal input, power button doesn't turn off the computer
- PC randomly loses video signal, dies
- Help, computer loses video signal then just stops responding.

So it appears I may have either:
- Faulty RAM
- Damaged or faulty motherboard
- Motherboard grounded to case (not seated right on the stand offs)
- Power supply unit not "powerful" enough or faulty power supply unit
- Some other problem

___________________________

My Questions:
___________________________

1. I would like a second opinion as to what could be wrong with my computer considering my specific set up (specs listed above).

Even if it's a rehash from what I've read. It would be very helpful for me. Maybe a specific part I have has a known issue or there's a known conflict between certain parts I am using.

2. The problem doesn't occur every time I boot up the computer. Even if I test the problem with things like, 1 stick of RAM in, or assembling the rig outside of the case, how can that help me determine the cause of my problem? How many times would I have to boot the computer up to be certain of a tests success?

3. Am I in danger of damaging parts by simply leaving it on and using the computer (as I am now) without shutting it off?

If possible, what can I do to check if I damaged anything (particularly the GPU)?

4. Could this somehow be software related? Would an automatic computer cleaner that could have removed key registry values cause something like this (although I think this is doubtful)?

Sorry for the length, I wanted to cover everything and keep it organized. Thank you very much in advance.

 

[MarkW]

Why are you using AMD overdrive with an Nvidia Graphics card? I would suggest HWinfo or SpeedFan for checking any CPU or GPU over AMD overdrive.

What video card and power supply did you replace? And why?

I need a model number on the RAM. It is a brand I have never heard of before, but I did find a review of it on LegitReviews.com, and they had no issues with it.

Now to your questions...

1) Unknown at this time. We will work our way to that by a process of elimination.

2) By removing all but one stick of memory, and testing the memory with that single stick installed, you can find out of that single stick has a problem. It is time consuming to do this, but there is no other way to be certain if a single stick of memory is causing your problem.

3) For now, I would power down any time you plan on being away from the computer for more than 30 minutes. Just to be safe.

4) Not likely. But if you want to clean things up in the registry, I have found CCleaner to be useful for that, and they do have a free version available.

http://www.piriform.com/ccleaner/download

 

[Lxno78]

Why are you using AMD overdrive with an Nvidia Graphics card? I would suggest HWinfo or SpeedFan for checking any CPU or GPU over AMD overdrive.

I wanted it more for the CPU than the GPU. I read in a few threads on the web that AMD Overdrive is more accurate to use for the newer AMD CPUs than other monitoring programs. (Here is one thread). However, I do have HWiNFO64 installed, it also says things are within a decent temperature range. (If I am remembering correctly the CPU was at about 25 C when idle and 45 C or so with load, and the GPU is about the same. Those temps are both from HWiNFO64).

(I don't run AMD Overdrive alongside HWiNFO64 nor do I run any one of those constantly.)

What video card and power supply did you replace? And why?

There was no graphics card, except the on-board card (AMD Radeon 3000). So no physical replacement there. The on board graphics card was incompatible with Windows 8 and caused the image to be stretched horizontally. Also, it's obviously not going to hold up well for gaming, when it's a low end on-board card.

The PSU was a Logisys PSU (it came with the case). I looked into Logisys and found that, not only are their PSUs extremely cheap, they over rate their output wattage, and are unreliable. It needed to be replaced and I needed more power for the graphics card.

I need a model number on the RAM. It is a brand I have never heard of before, but I did find a review of it on LegitReviews.com, and they had no issues with it.

Model KM3N4GX2C-1600-09-09-09-24-0

Thank you so much for the help! I'm definitely willing to test as needed, it may be time consuming, but I'm willing.

 

[MarkW]

OK. I like that you researched the power supply that came in the case, and replaced it before it caused problems. Most people would have just used it until something ugly happened. And then its often too late to help them much.

Until tonight, I had never heard of this memory company. And you are the second one here with it tonight. Not sure if this is a good thing or a bad thing yet. But we will find out.

So lets start out by testing the memory. I like to use MemTest. There are two sites that offer it. I prefer the free site. Go to the link below this, and download the ISO image file. Burn that to a CD. Then restart your computer. It should boot from the CD you just made. If not, go into the BIOS, and add the CD/DVD drive as the first drive in the boot order.

http://www.memtest.org/

MemTest runs at the hardware level. Windows is not loaded during this test. That is so that Memtest has absolute control of the hardware during this process. What I recommend is running this test overnight. I like to see it run for at least 10 hours. If any errors are found, you will be informed on the screen. The software is simple, looks like an old MS-DOS application, so there are no graphics. Its just barebones software, testing the memory using lots and lots of patterns. In the process, your CPU gets a decent test too.

If you run into errors, remove one stick of memory, and then test the one that is left. If that stick passes 10+ hours of testing, swap memory sticks, and test the second one.

After the above, let me know what happened.

 

[Lxno78]

I ran in to two problems.

1. http://www.memtest.org/ has the download for Memetest86+ Latest version 5.01 (Spetember 27, 2013). This ISO would not boot from my computer. The computer would state that there was no bootable device when trying to boot the iso CD from DVD drive.

However, http://www.memtest86.com/download.htm has a download for Memtest86 (no plus) V6.1.0 which did boot from and run from ISO CD on my computer. But, this leads to problem two.

2. I had only run the test for about 40 minutes when the test appeared to crash/permanently freeze. I noticed that the memtests progress bar stopped moving, and the "pattern" (it looks like hexadecimal) which was changing very frequently had completely stopped changing. As well, after it froze I could no longer access the configuration menu by pressing the "c" key. However, the power button did work to shut off the computer, unlike what happens with the current crashing problem I am having now.

Should I give the Memtest another try?

 

[MarkW]

Remove one stick of memory, and run it again. If it freezes with that stick, swap sticks, and try again.

 

[Lxno78]

Remove one stick of memory, and run it again. If it freezes with that stick, swap sticks, and try again.

Okay, I've tested one of the sticks of RAM for 16 hours (3 passes). I had zero errors. All tests were passed.



Before testing the other stick of RAM I wanted to try something. I figured I could isolate the problem by booting and logging in multiple times. These were the results:

Setup - All components in (including both sticks of RAM):

Boot Up 1 - Computer seized up and monitor went black as soon as login screen appeared. Had to turn off the PSU.

Boot Up 2 - Computer booted up and logged in normally.

Boot Up 3 - Computer seized up and monitor went black as soon as login screen appeared. Had to turn off the PSU.

Boot Up 4 - Computer booted up and logged in normally.

Boot Up 5 - Computer seized up and monitor went black as soon as login screen appeared. Had to turn off the PSU.

Boot Up 6 - Computer booted up and logged in normally.

Setup - Removed the stick of untested RAM leaving in only the stick of tested (0 errors) RAM in the slot it was tested:

Boot Up 7 - Computer seized up and monitor went black as soon as login screen appeared. Had to turn off the PSU.

Boot Up 8 - Computer booted up and logged in normally.

Boot Up 9 - Computer seized up and monitor went black as soon as login screen appeared. Had to turn off the PSU.

Setup - Left only the tested stick of RAM and Removed Screws from outside part of GPU metal cover plate (to test if the screws and metal plate was somehow grounding me to the case):

Boot Up 10 - Computer booted up and logged in normally.

Boot Up 11 - Computer seized up and monitor went black as soon as login screen appeared. Had to turn off the PSU.

Setup - One stick of RAM (the tested one) and Removed Entire GPU :

Boot Up 12 - Computer booted up and logged in normally.

Boot Up 13 - Computer booted up and logged in normally.

Boot Up 14 - Computer seized up and monitor went black as soon as login screen appeared. Had to turn off the PSU.

(Note: This time before the screen cut out I got really weird "fuzzy, green, horizontal striped" distortion that faded in over the screen before turning black.)

Boot Up 15 - Computer booted up and logged in normally.

----------

The problem seems to occur right when the log in screen comes on and only every other time.

Should I even test the other stick of RAM at this point? I've ruled out that it's not causing the problem.

Could AMD Overdrive have messed something up in the BIOS or automatically changed a setting when I had run it? (I didn't touch anything the one or two times I used it, but maybe something was changed automatically).
(I found this, it seems to be a similar problem, but I'm not sure.)

 

[MarkW]

Wow! That was a lot of testing. Now lets see if Windows can help explain whats going on.

Download a program called BlueScreenView. The download links are at the bottom of the page on this link:

http://www.nirsoft.net/utils/blue_screen_view.html

Look in C:\Windows\MiniDump\ and see if there are files with timestamps during the time you were doing all of that. If there are, run BlueScreenView. Lets get an idea of what Windows thinks is going on.

 

[Lxno78]

Look in C:\Windows\MiniDump\ ...

It looks like Windows isn't creating a crash log. When I try to go to C:\Windows\MiniDump\ I get Windows can't find 'C:\Windows\MiniDump\'. Check the spelling and try again.

Bluescreen viewer says 0 crashes and lists no dump files.

I did a search across both hard drives for minidump and nothing came up.

This isn't looking good for me is it lol.

 

[MarkW]

You can't get there?

 

[Lxno78]

I can't. Here is about where the folder should be in explorer (it's not there):

Here is me trying to access the folder directly:

Here is what BlueScreenViewer shows:

 

[MarkW]

OK, well your computer cannot have crashes then. Good Luck. . . .

Alright, I'm going to send you to a Microsoft page that shows how to enable these crash dump files.

https://msdn.microsoft.com/en-us/library/windows/hardware/ff542953(v=vs.85).aspx

 

[Lxno78]

My setting were already set exactly as the picture shown in that article. Is there anything I should change there? I found this article perhaps one of those items are happening to stop the dump from being created. How large should my paging file on the boot drive be?

I have video footage of the problem occurring that I can upload (to a site like YouTube) I can also take photos of the motherboard if that will help.

 

[MarkW]

Pagefiles should be (windows version, minimum size, maximum size):

Windows 7 and Windows Server 2008 R2 1 x RAM 3 x RAM or 4 GB, whichever is larger
Windows 8 and Windows Server 2012 Depends on crash dump setting* 3 x RAM or 4 GB, whichever is larger
Windows 8.1 and Windows Server 2012 R2 Depends on crash dump setting* 3 x RAM or 4 GB, whichever is larger

Post the video on youtube... that might help...

I doubt the motherboard pics will have any hot chicks in bikinis, so that probably won't help much... :lol:

 

[Lxno78]

Okay, page file is at 4096 MB. Set Dump file to Kernel dump (as I have read they are smaller).

EDIT: Removed broken link

I doubt the motherboard pics will have any hot chicks in bikinis, so that probably won't help much...

I can use Photoshop to fix that lol

 

[MarkW]

No you cant. Your computer will keep getting hot and bothered and crashing!

 

[Lxno78]

This link should work for the video of my computer problem.

 

[MarkW]

I've watched that a couple times already, and my head is screaming,"Video card!!!"

 

[Lxno78]

I was thinking that too, unfortunately I removed the video card (it was out when I recorded the video).
In the video the monitor is connecting to the default mobo DVI port.

My problem still occurred, with just as much frequency, with the actual GTX 750 graphics card in.
Unless that graphics card has somehow damage the on board one or the on board one somehow effected the GTX 750 card?

 

[MarkW]

Then what video card is in there... last time I checked the AMD FX-6300 did not have onboard video.

 

[Lxno78]

The ATI/AMD Radeon 3000, it's integrated into the motherboard. It took me awhile to figure that one out, it's not mentioned on the motherboard documentation nor is it advertised. However, the motherboard comes with VGA and DVI ports and I am able to get video signal somehow (when no GPU is plugged in).

 

[MarkW]

I most certainly was unaware of that.

I think this points to the motherboard then. Because its the only thing that makes sense to me right now. You have already replaced the power supply with one that I really like. And with the 750 card out, the only things that should be able to affect the video is the motherboard, the CPU and the DDR3 memory. You already tested the memory. It could be a factor. Whatever onboard video that thing comes with most certainly uses the DDR3 memory. What keeps bothering me is the every other time thing. I've never had a system work only every other time.

I think I would contact whoever you bought the parts from, and try to RMA the motherboard, CPU and memory, and see if they will exchange all 3 for you. Without spare parts, you are not going to be able to tell which part(s) are causing this. I think we have eliminated the 750 card here. And I am nearly certain that its not the power supply.

There is one more thing I might try.. Disconnect anything that does not have to be connected for the system to boot up. Printers, Ethernet, even the mouse and keyboard, anything external, except for the monitor, then boot a few times and see if anything changes. I doubt that it will, but its free to do, and relatively quick and easy to do as well. If the system continues to work properly every other time with everything else disconnected, then I think I would do the RMA thing above.

 

[Lxno78]

I figured it out. My own stupidity too lol. So, I read more than once on the forums that AMD users should use AMD Overdrive to check CPU thermal margins (the reason that AMD CPUs didn't calibrate well with third party software ). Well, maybe some AMD users should stay far away from Overdrive. I also read online that AMD Overdrive has the capability to changes BIOS setting, the voltage being sent to CPUs and GPUs, and that sort of thing (or I misread that and it doesn't). It could have been conflicting with other software I had installed as well, but I don't have many programs on my computer yet.

So, I was looking in the Startup tab of MSconfig and saw that a process called AODServices was starting up with the computer. I unchecked it and, sure enough, my system has started the last 5 or more times without any problems.

My guess is, since my CPU was incompatible with the service, AMD Overdrive was either changing voltage sent to the CPU or overriding a specific setting. I'm guessing CPUs and integrated GPUs work hand-in-hand and this probably is why it looked like a GPU issue.

Anyway, thank you so much for helping me out and my apologies for causing this problem in the first place.

 

[MarkW]

All that... and it was a software thing...I am so glad you figured that out though. I had reached the point where I was beginning to believe that system was possessed by the devil himself.

. . .

 

[Lxno78]

Lol, yeah it felt that way for sure.

How much damage do you think I may have cause any internal parts?
(Considering that I shut it down improperly about 20 times and the CPU wasn't being handled by the system correctly during that time.)

What utilities can I use to check the hard ware components (I know I can do memtest for the RAM), but what about the SSD, HDD, GPU, CPU, and motherboard?