|
Problem description: This was a computer I built for myself for the first time in September. My computer will randomly freeze after what seems like a random period of time and will either restart on it's own or I will have to hold the power button down in order to shut it down and turn it back on. It's happened to me while just surfing chrome and has also frozen on me while playing Destiny 2. My computer, for instance, was running fine from this morning up until around 5:00 PM where it froze up and restarted. Then while opening up the usual programs withing 15-30 minutes it froze again. Event Viewer is only giving me Kernel Power errors and I did not see any other errors of note. Checking out the temperatures in HWinfo it looks like the temperatures all around are fine. Nothing seems to be running too hot. Attempted fixes: I have reseated the graphics card and ram. I ran Memtest86+ over night to test the ram and it went through 3 iterations without any errors. I ordered a new PSU and switched it out but have still been getting the same issue. Recent changes: The only event I can think of was that there was a lightning storm awhile ago and I had my computer stupidly plugged into the wall for some reason instead of the surge protector and I lost power to the computer. After that I plugged it into a surge protector but I would say a bit after that was when my issues had started. -- Operating system: Windows 10 Pro 64-Bit System specs: Ryzen 5 1600 in a ASRock AB350M Pro4 motherboard. Running an Nvidia GeForce GTX 1080 along with a Samsung 820 EVO SSD that the operating system is installed on. There's a Seagate Barracuda 2TB HDD as well. PSU is the EVGA SuperNova 550 G2. Location: USA I have Googled and read the FAQ: I have
|
#
?
Nov 23, 2017 01:02
|
|
|
|
| # ? Apr 19, 2024 21:46 |
|
|
Can you describe the kernel power issues in more detail? Also can you give the specific kernel error ID? Note that when it gives this error, it is usually AFTER the event that has happened - the power issue is after the system has been rebooted. What does it say before the power issue in the logs? Have you gone into the power settings and set it to maximum performance? Are you using a specific power profile that may be causing this? How many sticks of RAM are in the system? I see this commonly on servers, where memtest says everything is fine but there's still a RAM issue (bad slot, etc). I would definitely try using only a single stick in the computer for a while (slot A1 or whatever the label says), painful as that may be, and see if this issue occurs. If it happens again, try a different RAM stick. If it happens again after that, then switch the slot that you are putting the RAM in, and see. Honestly this sounds like a memory issue to me if the PSU is good and the temperatures are fine; memtest is not always as reliable as you may think. So definitely try the RAM first
|
|
#
?
Nov 25, 2017 17:00
|
|
|
Under Event Viewer it just comes up as Event ID 41, Kernel-Power which is just indicating that it lost power suddenly and wasn't properly shut down. Looking through the administrative events for anything that would happen before the power issue I only see a DCOM error and another regarding Windows Hello for Business which I'm pretty sure are unrelated. The current power setting selected is balanced. With display and sleep settings set to 10 and 30 minutes respectively. This problem has also persisted between Windows 10 clean reinstalls. I have two sticks of RAMtotaling 16 GB installed under slots A2 and B2. I've tried running both RAM sticks by themselves under slot A2 and have gotten restarts with both. I'll try running a single RAM stick in the different slots and report back. The most annoying part about troubleshooting this so far is just the random amount of time it'll take before freezing and/or restarting.
|
|
#
?
Nov 25, 2017 22:06
|
|
|
Hello: Hmmm. You're 100% sure the power supply is good? (This is a new PSU, correct, not the one that you think might have gotten blasted by a lightning storm?) And you're also positive that the system isn't using more power than the PSU can supply? You should be using a high wattage supply with this rig. It doesn't really matter but why are you using A2 and B2 instead of A1 and B1? It should still work fine, of course, but I'm curious as to why you did that. Try sticking a stick in A1 and see if that works. The memory may not be the problem but its a good place to start here. There's literally nothing else in the event viewer than the DCOM error? Another possibility is that the motherboard is going bad - how old is it? It's unlikely that this is the culprit but that's gonna be the next place to look after the PSU and the memory. Normally my first guess here would be temperature but you said all is well on that front. How long is the average uptime before the crash? Does it crash after you do more intensive things (ie playing a game vs reading email)?
|
|
#
?
Nov 26, 2017 12:14
|
|
|
du -hast posted:Hello: Yeah this is a new power supply I bought after the lightning event because it was my initial thought on what was wrong. The supply itself is 550w which should be enough I’m pretty sure. The RAM is in A2 & B2 just because that’s what the manufacturer recommends when running two RAM sticks. Last night I tried using each stick of RAM individually in each slot and it would still freeze. In the event viewer there was nothing else. The motherboard and everything else except for the PSU I had just purchased were bought in September. Yeah the temperature when I was monitoring it on both HWmonitor and coretemp were showing the CPU at a high 30 low 40 temperature range. The average uptime at this point can vary. But if I have my computer on sleep or shutdown for a few hours than during use it will last a few hours usually before a freeze/restart. After the first one the subsequent restarts are usually quicker more in the range of about 30 minutes. Earlier today it froze on me and I was just reading stuff on Chrome. Sometimes it’ll restart and just hang on a black screen with the fans still running as well.
|
|
#
?
Nov 26, 2017 20:48
|
|
|
Ok we're in more confusing territory then. To clarify: Have you had this problem the ENTIRE time you've had this setup? ie from day 1? Because if so, another thing to note - you're cutting it pretty close on the power supply. I don't know why you got a 550 watt for a rig like this  Generally I'd recommend 750 (or if you wanna be  like me, a 1000W). So if you have had this issue for the entire duration, that could very well be the cause. AMD is generally pretty power hungry and so is this video card.A great idea: remove the video card, and plug into the one on the motherboard (according to the spec sheet it has an onboard video connector). You won't be able to play any cool games with it, but this will be a good way to diagnose if the PSU is the issue; the video card generally uses a large amount of power and you will definitely be under 550W with that. The PSU is the one thing I can think of right now. I think that may solve the issue. If it works fine / no freezes after you remove the card then that's probably the issue: power underrun. If that doesn't work, our next trip down the rabbit hole is going to be checking to HDD (CHKDISK to see if theres some corruption). I would W10 would allow a user to have a corrupt drive without notifying them but at the same time its the botnet OS that demands you allow it to update, so it may very well be the issue. But currently, since we've eliminated the RAM, and the motherboard is probably good (difficult to test its integrity as well), the power is my next guess.
|
|
#
?
Nov 27, 2017 12:01
|
|
|
Hey, sorry for the delay, work has been busy. Also just wanted to say I appreciate the advice and ideas. It helps me from going crazy trying to diagnose this on my own. This problem did not start occurring until recently. Maybe the beginning of November, the computer I had built the beginning of September. While the motherboard itself has an onboard video connector the Ryzen chip doesn't have integrated graphics and the only video card I have is the Nvidia. Agh. Just the standard CHKDSK command for both the C: and D: Drives returned no filesystem errors.
|
|
#
?
Nov 29, 2017 05:33
|
|
|
|
| # ? Apr 19, 2024 21:46 |
|
|
Did you try a re-install anyway on the SATA drive? Sorry it's taking so long to figure out the issue here. Please clarify what you mean when you say that the Ryzen doesn't have a video controller? According to the spec sheet it should have an integrated graphics chip (that kinda sucks). You may need to set the video preference in the BIOS to ONBOARD instead of OFFBOARD if the card is not working. The purpose of removing the video card was to see if the issue was power-related (550W PSU being possibly not enough wattage to handle the setup you've got). So first and foremost remove the video card, and try running on the onboard one for a while and see if it works. I mean specifically pulling the card from its slot and setting it on the desk, not just leaving it in there. That's my first guess - it should probably work. The other idea was that the OS is somehow hosed, which is why I recommended doing an install to the SATA (non SSD) drive and see if that helps. CHKDISK is handy but unreliable in diagnosing OS-related issues. I have also been off work for a few days but will be around for the next week. I recommend trying these two ideas and seeing if it fixes it - I really think you may just have a PSU that is not quite high wattage enough to handle the video card and system when its running at full blast. Anyway any other ideas you have, let me know. We've diagnosed that it's not the RAM, is unlikely that the video card itself is hosed, that the motherboard appears to be good to go - now the two things that we have left are the OS itself that may be hosed somehow, or a lack of adequate wattage. If you have any other ideas by all means - this is a tricky one. Anyway give these two ideas a shot and let me know. You're not getting any beeping, etc, anything like that - correct?
|
|
#
?
Dec 1, 2017 12:36
|
|