Apologies for the length of this post before getting to the point: feel free to skip to the bottom. However I wanted to narrate the entire sequence of events (just in case I may have overlooked a cause), and to allow other users to search for similar problems.
Here's my tale of woe:
Last night my Windows 10 froze solid. The mouse cursor would not move. I rebooted, and it worked fine for about 2 minutes - then froze again. Next time I rebooted, it crashed at the logon screen.
I booted from a recovery drive and tried to run System Restore. However none of the 3 restore points worked: after apparently restoring, it gave an error message that System Restore could not be completed.
Exploring the other options from the recovery drive's control panel, it was at this point that I found out that the other option to "reset Windows" would set the installation back to an out-of-the-box state: very annoying! I had thought that the recovery drive would be the same thing as a System Image, but no: I am faced with losing all my settings. Fortunately I store all personal documents on other drives, including a network drive (NEVER save documents to C:\). Still, it will take ages to re-install all programs, drivers and settings. Anyway, I was resigned to it.
When I tried the "Reset Windows" feature, it went through the motions - but then after rebooting, it got stuck at 16% (Installing drivers, or similar). Multiple attempts produced the same result.
I then got a new SSD which I had spare, and tried to re-install Windows from scratch (although I upgraded from 8.1, I created an installation DVD). I disconnected all other drives and PCI cards, leaving only the graphics card, memory and system drive connected to the motherboard. However the computer similarly froze during the installation after rebooting.
Before I went to sleep last night I proceeded on the theory that perhaps I was not being patient enough, and that the installation had not froze, but it was just taking a long time. I therefore left the "frozen" installation running. During the night, however, the computer powered itself down. I did not know why this would have happened until later (see below)
I figured it must be a hardware problem, as a fresh installation on a fresh SSD should have no remaining malaise from whatever went before. I therefore ran Memtest86 from a boot CD, with the intention of systematically removing my memory sticks to identify errors. No errors were found - however even the Memtest86 UI froze after a few minutes! A longer attempt at letting it run resulted in the computer shutting itself down.
It was ONLY AT THIS POINT that I finally noticed that my CPU cooler's fan had stopped spinning - oops! If I had noticed that before, I could have saved myself a lot of bother. Unfortunately I could not diagnose the cause of the stoppage (it was not blocked by dust or anything).
Anyway, at least I now had an identifiable problem. I replaced my Zalman cooler with the previously-unused stock cooler which came with the CPU - still in its box, and with a pre-applied layer of thermal paste. I made sure to wipe clean the old thermal paste from the CPU before installing the new cooler.
I then turned on the computer again to try Memtest86. It proceeded for a good time without freezing (and without RAM errors). Good news, right?
I then proceeded to retry the installation of Windows 10. This time it actually installed to completion. Faced with a lot of reinstallation of stuff, the first thing I did was download the Nvidia drivers in order to be able to display a proper resolution. When running the driver installation program, however, I noticed that the noise from the CPU cooler was getting louder... and louder. Within about a minute, it was whining away at top speed.
Touching the heatsink, it was obvious to me that it was a lot hotter than it should be, considering the amount of time the computer had been on. And surely the simple matter of installing a graphics driver should not overwork the processor, no?
More alarmingly, however, I then noticed that the motherboard's own chipset heatsink (fanless) was VERY hot. Too hot to leave my finger on comfortably for more than a couple of seconds.
Not wanting to cause further problems, I turned off the computer (before the graphics driver installation had completed). I left the computer to cool, then tried again. This time upon loading Windows I immediately tried to install the graphics drivers before doing anything else. Again, at the same stage of the process, the CPU fan got faster and faster, working overtime to cool it.
So... I'm now thinking that the fact that the CPU's old fan had failed (and I hadn't noticed it) means that it overheated to such a level as to cause permanent damage
Could someone please let me know if they think this may be the case? Note that the CPU's own failsafe appears to have worked, given that the computer did shut itself off a couple of times.
If it is indeed the case that the CPU has been permanently damaged, what could explain the motherboard chipset also getting so hot? Is it possible that a damaged CPU could in itself cause the chipset to overheat? Or is it also possible that the motherboard has also suffered damage? If so, how could this be?
Basically, the central problem I have is that I want to know if I need a new CPU; new motherboard; or both. I don't want to waste money on replacing them both (or replacing the wrong one, if it's the other one at fault).
Any help greatly appreciated. Sorry for the length - and thanks!
Here's my tale of woe:
Last night my Windows 10 froze solid. The mouse cursor would not move. I rebooted, and it worked fine for about 2 minutes - then froze again. Next time I rebooted, it crashed at the logon screen.
I booted from a recovery drive and tried to run System Restore. However none of the 3 restore points worked: after apparently restoring, it gave an error message that System Restore could not be completed.
Exploring the other options from the recovery drive's control panel, it was at this point that I found out that the other option to "reset Windows" would set the installation back to an out-of-the-box state: very annoying! I had thought that the recovery drive would be the same thing as a System Image, but no: I am faced with losing all my settings. Fortunately I store all personal documents on other drives, including a network drive (NEVER save documents to C:\). Still, it will take ages to re-install all programs, drivers and settings. Anyway, I was resigned to it.
When I tried the "Reset Windows" feature, it went through the motions - but then after rebooting, it got stuck at 16% (Installing drivers, or similar). Multiple attempts produced the same result.
I then got a new SSD which I had spare, and tried to re-install Windows from scratch (although I upgraded from 8.1, I created an installation DVD). I disconnected all other drives and PCI cards, leaving only the graphics card, memory and system drive connected to the motherboard. However the computer similarly froze during the installation after rebooting.
Before I went to sleep last night I proceeded on the theory that perhaps I was not being patient enough, and that the installation had not froze, but it was just taking a long time. I therefore left the "frozen" installation running. During the night, however, the computer powered itself down. I did not know why this would have happened until later (see below)
I figured it must be a hardware problem, as a fresh installation on a fresh SSD should have no remaining malaise from whatever went before. I therefore ran Memtest86 from a boot CD, with the intention of systematically removing my memory sticks to identify errors. No errors were found - however even the Memtest86 UI froze after a few minutes! A longer attempt at letting it run resulted in the computer shutting itself down.
It was ONLY AT THIS POINT that I finally noticed that my CPU cooler's fan had stopped spinning - oops! If I had noticed that before, I could have saved myself a lot of bother. Unfortunately I could not diagnose the cause of the stoppage (it was not blocked by dust or anything).
Anyway, at least I now had an identifiable problem. I replaced my Zalman cooler with the previously-unused stock cooler which came with the CPU - still in its box, and with a pre-applied layer of thermal paste. I made sure to wipe clean the old thermal paste from the CPU before installing the new cooler.
I then turned on the computer again to try Memtest86. It proceeded for a good time without freezing (and without RAM errors). Good news, right?
I then proceeded to retry the installation of Windows 10. This time it actually installed to completion. Faced with a lot of reinstallation of stuff, the first thing I did was download the Nvidia drivers in order to be able to display a proper resolution. When running the driver installation program, however, I noticed that the noise from the CPU cooler was getting louder... and louder. Within about a minute, it was whining away at top speed.
Touching the heatsink, it was obvious to me that it was a lot hotter than it should be, considering the amount of time the computer had been on. And surely the simple matter of installing a graphics driver should not overwork the processor, no?
More alarmingly, however, I then noticed that the motherboard's own chipset heatsink (fanless) was VERY hot. Too hot to leave my finger on comfortably for more than a couple of seconds.
Not wanting to cause further problems, I turned off the computer (before the graphics driver installation had completed). I left the computer to cool, then tried again. This time upon loading Windows I immediately tried to install the graphics drivers before doing anything else. Again, at the same stage of the process, the CPU fan got faster and faster, working overtime to cool it.
So... I'm now thinking that the fact that the CPU's old fan had failed (and I hadn't noticed it) means that it overheated to such a level as to cause permanent damage
Could someone please let me know if they think this may be the case? Note that the CPU's own failsafe appears to have worked, given that the computer did shut itself off a couple of times.If it is indeed the case that the CPU has been permanently damaged, what could explain the motherboard chipset also getting so hot? Is it possible that a damaged CPU could in itself cause the chipset to overheat? Or is it also possible that the motherboard has also suffered damage? If so, how could this be?
Basically, the central problem I have is that I want to know if I need a new CPU; new motherboard; or both. I don't want to waste money on replacing them both (or replacing the wrong one, if it's the other one at fault).
Any help greatly appreciated. Sorry for the length - and thanks!
Facebook
Twitter
Instagram