- Jan 4, 2013
- 40
- 0
- 10,540
I've used other posts on this forum for assistance with various technical issues before and the community has always been knowledgeable before but this issue is beyond me so I bring it to you. Fair warning, this is going to get very very long and very very confusing so I don't blame you if you skip parts or ignore it altogether. Here goes!
System Specs:
intel Core i7 3770k @ 3.5GHz (Arctic Cooling Hyper TX3 cooler, using MX2 Thermal Paste)
Asus P8Z77-V
16GB (4x4GB) Corsair Vengeance 9-9-9-24 1600MHz 1.5B (Model CMZ16GX3M4A1600C9)
Evo Labs 1200W Modular PSU
Gainward nVidia GTX 570 Phantom 3 1280MB GDDR5
Generic Samsung Blu-Ray Drive (S-ATA)
Seagate Barracuda 160GB 7200RPM (S-ATA)
Seagate Barracuda 500GB 7200RPM (S-ATA)
Samsung Spinpoint F1 2TB 7200RPM (S-ATA)
Windows 7 Ultimate Edition x64 (Fully updated)
Corsair K90 Keyboard
Corsair M90 Mouse
Xbox 360 Pad
The Problem:
My machine restarts while playing games but not during general use nor in stress testing or benchmarks. The restart is a hard power-off as though the plug has been pulled with the PC restarting after 2-3 seconds of appearing to be idle. Motherboard power indicator LED never goes off or dims (manual and my own testing says it flashes 5 times when power is lost, finally going out when residual power from the PSU has been spent)
Tests Undertaken:
Firstly I went to the BIOS and reset all clock speeds to their default, disabled Intel SpeedStep and Q-Fan and made sure everything was running at default values for the pieces installed.
My first thought was to blame drivers for the issue so I downloaded a full set of updated drivers for everything (motherboard, GPU, Keyboard/Mouse, the whole lot basically) and decided to be thorough and reinstall Windows at the same time. So I reinstalled Windows, reinstalled all the drivers one at once (with a single system restart between each install).
I did NOT plug in the 360 pad at this point, preferring to leave that particular driver off my system to eliminate it from testing).
I also did NOT install any antivirus or firewall software, to avoid conflicts, instead using already downloaded files to avoid a net connection), additionally I disabled Data Execution Prevention and User Account Control.
I reinstalled Steam and re-installed two games I've been using for testing (simply because they cause the restarts the fastest, all games eventually restart but these two are just best for testing - Magic The Gathering Duels Of The Plainswalkers 2013 and XCOM Enemy Unknown). Of course it wouldn't be that simple, the restarts continued.
With software aside I decided I'd give the hardware a test. I downloaded Furmark, Unigine Heaven and Catzilla benchmarks and allowed each to run for 2 hours each. The PC never restarted during this period so I allowed Furmark to run all night (approximately 8 and a half hours) and the PC had not restarted during this time.
I then moved on to Prime 95 for a bit of CPU torture testing. I ran Prime 95 on all 8 cores (4 physical, 4 logical) overnight and had no restart. I then disabled the extra 4 logical threads on my CPU and ran Prime 95 overnight the following night. Again, no restart.
Next culprit is RAM, so I grabbed Memtest and set it up on a bootable USB stick and let that run all day (a few minutes shy of 14 hours) and it discovered NO errors. I wanted to be thorough though so I started yanking sticks of RAM (using the B2 socket as advised by my motherboard manual for testing single sticks of RAM) and booting up. The PC boots fine with each of the 4 sticks and the restarts continue to occur (with every single stick). I tried dual-channel (using the B2 and A2 sockets as advised by the motherboard manual) for a repeat of the single-stick tests and got the same results.
Next, I blame the PSU and download OCCT. I immediately notice that it's mis-reporting the 12V as 6.6V and pop to the BIOS to verify, indeed it's running at 11.94V (well within the +-0.4V range that I understand to be acceptable). Nontheless I proceed to let OCCT torture my PSU for three hours and it doesn't trigger a restart. Going through OCCT's graphs reveal absolutely no spikes, dips or other inconsistencies in the power flow other than the 12V graph occasionally rising to 7V (but we already know that's misreporting in OCCT).
Time to blame the HDDs. The 160GB drive has Windows on, so I decide to yank all but that and run the games again. Restarts persist. I've followed the clean windows install on both other drives to test those individually, all the while listening carefully for any clicks or thrashing around the time of the restart that might indicate a hardware failure in the HDD itself, no such sounds and the restarts persist. So I format all but the 160GB drive and move on.
Time to blame the case itself (an old NZXT Tempest). I spent an afternoon moving my machine over to an Antec Skeleton I had spare (re-doing the thermal paste and all component seating to be sure) and try again. Naturally, it's still restarting while gaming. At this point I took a break and played Skyrim for almost 2 hours before it restarted and I left it alone for the rest of the day (fully powered off at the wall socket).
Starting a-fresh I replace all the cabling inside with spares (apart from the CPU cooler). What I couldn't replace was the case front-panel cabling so I yanked all but the necessary power/reset switch cables and resumed gaming. Sure enough, it continues to restart. Now thoroughly hacked off with it I left it alone for the rest of the day again.
I decide to start all over again and start with the RAM (numbered 1-4 for simplicity here). Testing one stick at once in the primary RAM slot (B2) I get a clean boot and subsequent gaming restart from sticks 1 and 2. Stick 3 allows the machine to power on but produces hard drive thrashing after several seconds and no picture on screen. Stick 4 again allows the machine to power on but has no picture on the screen and no thrashing so I cycle back to sticks 1 and 2 together (in slots B2 and A2 - gotta follow that motherboard manual) and it still won't boot (despite literally 5 minutes ago it worked).
Out of curiosity I switch over to the intel graphics (simply moving the HDMI cable over to the motherboard HDMI socket) and it boots (still with sticks 1 and 2 in the slots). Intrigued I allowed the machine an hour to cool off without power and removed the graphics card and the associated power cables and attempted to boot off the intel graphics again. No boot. No biggie right? The motherboard probably checks to make sure a GPU is plugged in, so I stick it back in.
For some reason it dawned on me that the machine might boot with the discreet graphics card if I put all the RAM back in, so I did. I have no reasoning behind this and I don't understand it but sure enough, the graphics card results in a boot again with all 4 sticks of RAM in.
I confirmed the reboots were still happening and then ran Furmark and Prime 95 together to see what would happen. The result - nothing, nothing at all happened, they ran for several hours in tandem, pushing GPU and CPU temps up to 50-ish degrees (CPU clocked 54C highest).
In an act of desperation I have at last come to blame my keyboard and mouse. I uninstalled their drivers an control software, replaced them with a generic USB keyboard and mouse and ran CCleaner to sort out the registry rubbish before restarting.
After the restart I tried one more time and I still get the restarts while gaming.
Clarifications:
All Windows installs were fully updated and NOT cloned off the first clean install.
When I refer to motherboard drivers I mean JUST the drivers, not the Asus AI Suite that comes bundled with the motherboard.
In Closing:
I'm stumped. I only get a picture out of the GPU with all 4 sticks of RAM in (despite it booting twice off a single stick) and I can't seem to find any errors through torturing my system yet it restarts ONLY when gaming.
I almost forgot, I've played a few other games just now for the sake of messing about with it aimlessly, none of which triggered the restarts after about an hour playing each. The games include; Wing Commander 1 and 2, Anachronox, X-Com UFO Defence (all from GOG), Awesomenauts and Dungeons Of Dredmore (both on Steam).
Any suggestions or assistance would be very much welcomed. I've exhausted everything I can think of and I seem to have a nonsensical situation on my hands that defies my troubleshooting logic.
As a note, I have NO money whatsoever and cannot afford to go purchase any alternate components for testing purposes and if it is a hardware issue I'll simply have to live with it.
Thankyou very much for taking the time to look at this.
***NOT-SO-STEALTH-EDIT*** Forgot to mention previously that it doesn't restart when playing movies, audio or using game design toolsets like Unity or Game Maker Studio.
System Specs:
intel Core i7 3770k @ 3.5GHz (Arctic Cooling Hyper TX3 cooler, using MX2 Thermal Paste)
Asus P8Z77-V
16GB (4x4GB) Corsair Vengeance 9-9-9-24 1600MHz 1.5B (Model CMZ16GX3M4A1600C9)
Evo Labs 1200W Modular PSU
Gainward nVidia GTX 570 Phantom 3 1280MB GDDR5
Generic Samsung Blu-Ray Drive (S-ATA)
Seagate Barracuda 160GB 7200RPM (S-ATA)
Seagate Barracuda 500GB 7200RPM (S-ATA)
Samsung Spinpoint F1 2TB 7200RPM (S-ATA)
Windows 7 Ultimate Edition x64 (Fully updated)
Corsair K90 Keyboard
Corsair M90 Mouse
Xbox 360 Pad
The Problem:
My machine restarts while playing games but not during general use nor in stress testing or benchmarks. The restart is a hard power-off as though the plug has been pulled with the PC restarting after 2-3 seconds of appearing to be idle. Motherboard power indicator LED never goes off or dims (manual and my own testing says it flashes 5 times when power is lost, finally going out when residual power from the PSU has been spent)
Tests Undertaken:
Firstly I went to the BIOS and reset all clock speeds to their default, disabled Intel SpeedStep and Q-Fan and made sure everything was running at default values for the pieces installed.
My first thought was to blame drivers for the issue so I downloaded a full set of updated drivers for everything (motherboard, GPU, Keyboard/Mouse, the whole lot basically) and decided to be thorough and reinstall Windows at the same time. So I reinstalled Windows, reinstalled all the drivers one at once (with a single system restart between each install).
I did NOT plug in the 360 pad at this point, preferring to leave that particular driver off my system to eliminate it from testing).
I also did NOT install any antivirus or firewall software, to avoid conflicts, instead using already downloaded files to avoid a net connection), additionally I disabled Data Execution Prevention and User Account Control.
I reinstalled Steam and re-installed two games I've been using for testing (simply because they cause the restarts the fastest, all games eventually restart but these two are just best for testing - Magic The Gathering Duels Of The Plainswalkers 2013 and XCOM Enemy Unknown). Of course it wouldn't be that simple, the restarts continued.
With software aside I decided I'd give the hardware a test. I downloaded Furmark, Unigine Heaven and Catzilla benchmarks and allowed each to run for 2 hours each. The PC never restarted during this period so I allowed Furmark to run all night (approximately 8 and a half hours) and the PC had not restarted during this time.
I then moved on to Prime 95 for a bit of CPU torture testing. I ran Prime 95 on all 8 cores (4 physical, 4 logical) overnight and had no restart. I then disabled the extra 4 logical threads on my CPU and ran Prime 95 overnight the following night. Again, no restart.
Next culprit is RAM, so I grabbed Memtest and set it up on a bootable USB stick and let that run all day (a few minutes shy of 14 hours) and it discovered NO errors. I wanted to be thorough though so I started yanking sticks of RAM (using the B2 socket as advised by my motherboard manual for testing single sticks of RAM) and booting up. The PC boots fine with each of the 4 sticks and the restarts continue to occur (with every single stick). I tried dual-channel (using the B2 and A2 sockets as advised by the motherboard manual) for a repeat of the single-stick tests and got the same results.
Next, I blame the PSU and download OCCT. I immediately notice that it's mis-reporting the 12V as 6.6V and pop to the BIOS to verify, indeed it's running at 11.94V (well within the +-0.4V range that I understand to be acceptable). Nontheless I proceed to let OCCT torture my PSU for three hours and it doesn't trigger a restart. Going through OCCT's graphs reveal absolutely no spikes, dips or other inconsistencies in the power flow other than the 12V graph occasionally rising to 7V (but we already know that's misreporting in OCCT).
Time to blame the HDDs. The 160GB drive has Windows on, so I decide to yank all but that and run the games again. Restarts persist. I've followed the clean windows install on both other drives to test those individually, all the while listening carefully for any clicks or thrashing around the time of the restart that might indicate a hardware failure in the HDD itself, no such sounds and the restarts persist. So I format all but the 160GB drive and move on.
Time to blame the case itself (an old NZXT Tempest). I spent an afternoon moving my machine over to an Antec Skeleton I had spare (re-doing the thermal paste and all component seating to be sure) and try again. Naturally, it's still restarting while gaming. At this point I took a break and played Skyrim for almost 2 hours before it restarted and I left it alone for the rest of the day (fully powered off at the wall socket).
Starting a-fresh I replace all the cabling inside with spares (apart from the CPU cooler). What I couldn't replace was the case front-panel cabling so I yanked all but the necessary power/reset switch cables and resumed gaming. Sure enough, it continues to restart. Now thoroughly hacked off with it I left it alone for the rest of the day again.
I decide to start all over again and start with the RAM (numbered 1-4 for simplicity here). Testing one stick at once in the primary RAM slot (B2) I get a clean boot and subsequent gaming restart from sticks 1 and 2. Stick 3 allows the machine to power on but produces hard drive thrashing after several seconds and no picture on screen. Stick 4 again allows the machine to power on but has no picture on the screen and no thrashing so I cycle back to sticks 1 and 2 together (in slots B2 and A2 - gotta follow that motherboard manual) and it still won't boot (despite literally 5 minutes ago it worked).
Out of curiosity I switch over to the intel graphics (simply moving the HDMI cable over to the motherboard HDMI socket) and it boots (still with sticks 1 and 2 in the slots). Intrigued I allowed the machine an hour to cool off without power and removed the graphics card and the associated power cables and attempted to boot off the intel graphics again. No boot. No biggie right? The motherboard probably checks to make sure a GPU is plugged in, so I stick it back in.
For some reason it dawned on me that the machine might boot with the discreet graphics card if I put all the RAM back in, so I did. I have no reasoning behind this and I don't understand it but sure enough, the graphics card results in a boot again with all 4 sticks of RAM in.
I confirmed the reboots were still happening and then ran Furmark and Prime 95 together to see what would happen. The result - nothing, nothing at all happened, they ran for several hours in tandem, pushing GPU and CPU temps up to 50-ish degrees (CPU clocked 54C highest).
In an act of desperation I have at last come to blame my keyboard and mouse. I uninstalled their drivers an control software, replaced them with a generic USB keyboard and mouse and ran CCleaner to sort out the registry rubbish before restarting.
After the restart I tried one more time and I still get the restarts while gaming.
Clarifications:
All Windows installs were fully updated and NOT cloned off the first clean install.
When I refer to motherboard drivers I mean JUST the drivers, not the Asus AI Suite that comes bundled with the motherboard.
In Closing:
I'm stumped. I only get a picture out of the GPU with all 4 sticks of RAM in (despite it booting twice off a single stick) and I can't seem to find any errors through torturing my system yet it restarts ONLY when gaming.
I almost forgot, I've played a few other games just now for the sake of messing about with it aimlessly, none of which triggered the restarts after about an hour playing each. The games include; Wing Commander 1 and 2, Anachronox, X-Com UFO Defence (all from GOG), Awesomenauts and Dungeons Of Dredmore (both on Steam).
Any suggestions or assistance would be very much welcomed. I've exhausted everything I can think of and I seem to have a nonsensical situation on my hands that defies my troubleshooting logic.
As a note, I have NO money whatsoever and cannot afford to go purchase any alternate components for testing purposes and if it is a hardware issue I'll simply have to live with it.
Thankyou very much for taking the time to look at this.
***NOT-SO-STEALTH-EDIT*** Forgot to mention previously that it doesn't restart when playing movies, audio or using game design toolsets like Unity or Game Maker Studio.
Facebook
Twitter
Instagram