Need help diagnosing a power cycle problem

00Dan

Weaksauce
Joined
Dec 10, 2012
Messages
78
As of late I have been having an issue with my PC shutting down hard and then powering back on a few seconds later. This normally occurred during gaming but sometimes during desktop use. After running memtest overnight and disk check on all drives, I presumed it was either an issue with my PSU or my GPU. To experiment, I turned my GPU back to stock clocks (I had been running a slight overclock) and tried gaming for a few hours. I was able to do this without any power issues.

This leads to my conundrum:
Is my GPU not stable with it's OC? (It had been running fine for several months prior to this).
Is my PSU going bad? (When gaming I was usually able to run for about 10-15 minutes before the shutdown)
Is my PSU incapable of supplying the extra power needed for the OC on my GPU?

I'm not entirely sure where to go from here on diagnosing this. I would think that if it was my GPU I would maybe get BSOD instead of a straight power loss but I'm not certain.

For reference, my GPU is a 295x2, and my PSU is a Corsair HX850.
 
There is indeed a good chance that your GPU has degraded in its ability to handle overclocks. But, if your HX850 is approaching 5-6 years of use, then it is indeed possible that the PSU is the culprit since it's been shown that PSU can degrade in output capability over time:
SilverStone Olympia 1000W Power Supply 7 Year Later - SilverStone Olympia 1000W Power Supply 7 Year Redux

If you really need that GPU overclocked, then your next course of actions are:
1) Try a different overclocked R9 295X2.
2) Try a different 850W PSU

Otherwise, if the system is stable as is with default clocks and you don't really need the GPU OC'd, then I really wouldn't worry about it.
 
So, full hard shutdown, fans stop, monitor goes off, and then it just starts up again? after how many seconds?

There's a setting in most BIOS setups that instructs the mobo what to do in the event of AC power re-appearing. Check that setting and let us know.
Also, check the setting relating to what to do when the power button is pressed - usually the choices are 'shutdown immediately' and 'shutdown after 4 seconds of holding it down'. Let us know.

For a quick check if it's the power button or the wires leading from it to the mobo pins, just open the PC up and disconnect your case power button from the corresponding motherboard pins. Obviously - remove the PSU power cable first so it won't spontaneously turn on as you're digging inside.

When you're ready, turn the computer on by shorting those motherboard pins for a split second using anything from a paperclip to a flathead screwdriver.
I don't think a HX would degrade so badly after 5 years, that's why I'm latching to the power button first.
 
I'm rolling with PSU issues. Started problem in a loaded scenario, and then became more frequent with light use.
 
Sounds like PSU to me.

295 uses 500w all by itself, and because it breaks ATX spec on how much power it can draw from those 8-pin connectors. Perhaps this con tributes to early PSU death?

AMD Radeon R9 295X2 8 GB Review

For certain, if your video card draws 500w, if your PSU starts to slip even a little, you're going to notice. Most other builds have several hundred watts breathing room.

If the rest of your system is overclocked, you may be pushing beyond what your PSU is capable of. What is your CPU model?
 
Last edited:
Sounds like PSU to me.

295 uses 500w all by itself, and because it breaks ATX spec on how much power it can draw from those 8-pin connectors. Perhaps this con tributes to early PSU death?

AMD Radeon R9 295X2 8 GB Review

For certain, if your video card draws 500w, if your PSU starts to slip even a little, you're going to notice. Most other builds have several hundred watts breathing room.

If the rest of your system is overclocked, you may be pushing beyond what your PSU is capable of. What is your CPU model?


that link says that the card will us up to 650W under full load. that only leaves 200W left for the rest of the system. so if the psu is degrading that would be the most likely problem.
 
that link says that the card will us up to 650W under full load. that only leaves 200W left for the rest of the system. so if the psu is degrading that would be the most likely problem.

Only if the game you are playing is called Furmark :D

But 500w is still insane for video cards. Luckily his PSU has a single 12v rail, so that's not the problem (lots of people complained about stability issues with this card on multi-rail PSUs).
 
Bringing this thread back from the dead as the topic is still relevant. Today I finally got around to replacing my power supply. I upgraded from my Corsair HX850 to an EVGA 1200 P2. After having everything assembled and browsing the internet some, I again encountered the power cycle.
This power cycle occurred when I attempted to load a new tab in Google Chrome. This behavior was also observed under the old power supply. I have not yet tried gaming. Event viewer logs it as just Event 41: Kernel power.

At this point, moving past the PSU as the culprit, I'm somewhat at a loss of what could be causing a full power loss and restart. My RAM all checked fine with memtest. The computer always boots and POSTs without issue. Could a bad system SSD be the culprit (Wouldn't this cause a BSOD instead)? or is it perhaps the motherboard and/or CPU (How is the system even remotely stable if this is the case)?
 
Bringing this thread back from the dead as the topic is still relevant. Today I finally got around to replacing my power supply. I upgraded from my Corsair HX850 to an EVGA 1200 P2. After having everything assembled and browsing the internet some, I again encountered the power cycle.
This power cycle occurred when I attempted to load a new tab in Google Chrome. This behavior was also observed under the old power supply. I have not yet tried gaming. Event viewer logs it as just Event 41: Kernel power.

At this point, moving past the PSU as the culprit, I'm somewhat at a loss of what could be causing a full power loss and restart. My RAM all checked fine with memtest. The computer always boots and POSTs without issue. Could a bad system SSD be the culprit (Wouldn't this cause a BSOD instead)? or is it perhaps the motherboard and/or CPU (How is the system even remotely stable if this is the case)?

Giving this thread a bump. The issue has been reoccurring and I am at a loss for what may be causing it. My mobo is on the latest BIOS, I updated my SSD firmware and it is passing a checkdisk. Short of taking a plunge and replacing the entire motherboard I'm not sure what else to troubleshoot.

Event viewer seems to always show an error right before the power loss, and it is event 10, EnhancedStorage-EhStorTcgDrv. Google is throwing me all over the place with that event so maybe someone here can decipher it.
 
i think the error is pointing towards the sata controller. if you can try changing the sata cable and maybe if you have another drive to use try a new install on that to see if the issue goes away.
 
i think the error is pointing towards the sata controller. if you can try changing the sata cable and maybe if you have another drive to use try a new install on that to see if the issue goes away.

As a quick followup, I have noticed that the settings for programs I have installed on my system SSD get cleared when one of these power cycles occurs. If that is indicative of the SSD health or just a side effect of the power cycle, I do not know.
 
Back
Top