Random BSOD every 5+ hours

[Irish_PXzyan]

Howdy lads,

Currently with a rig that I'm using I'm having an odd issue which I'm not used too..
After the computer has been on for over 5 hours it begins to suffer BSOD constantly with error messages like memory management being the main one but other error messages appear but it's always the same BSOD and restarts.

Thing is once it restarts the issue will arrise again randomly from 1 minute to 20 minutes every time.
It just WILL NOT stop this until the PC is off for a solid couple hours before it can be used for 5+ hours again.

This happens either gaming or just watching youtube and surfing the web. Not actually seems to trigger it no matter what I do.

Sometimes the screen just goes black and stays there until I force restart via the case.


Has anyone any ideas what might be happening here?


Windows 10
AMD Ryzen 1600 @3.8Ghz
16gb DDR4 3000mhz CL18 :/
RX 480 8gb
Gigabyte b450m ds3h Mainboard

 

[Bill_Bright]

First thought is heat. What are your temps? Is the interior clean of heat trapping dust? Do all fans spin properly?

If temps are good, case interior is clean and fans spin, I would swap in a spare PSU and see what happens.

 

[xorbe]

Well what happens if you run the CPU at stock speeds, and set the memory to baseline 2666 MHz? If you are running 4 sticks, pull 2 sticks.

 

[Solaris17]

You tried running a memory test?

 

[FreedomEclipse]

Power supply could be complete ass. I know that Antec used to be known for making quality PSUs but that was a long time ago. Googling that PSU tells me that it is around 7 years old. It might be time to consider getting a new one.

I used to have a Corsair HX1050 that did the same thing, sometimes the system would just die during gaming and other times it would just die while i was doing something in word or watching YT/browsing the net etc etc.


Im not saying that its the power supply but it shouldnt be ruled out.

 

[Zareek]

In my opinion @Bill_Bright is right, this has all the signs of a classic thermal issue. @FreedomEclipse could be right too that it is the power supply but more than likely something thermally related to the power supply. Something is getting too hot and is taking a while to cool down enough to work properly again. If I had to make a bet, I would go with something in the board's VRM.

 

[johnspack]

As much as I liked the old Antecs... I'd have to suspect the psu unfortunately.

 

[Arrakis9]

You tried running a memory test?

As much as I liked the old Antecs... I'd have to suspect the psu unfortunately.

+1 to both of these, start with running a memcheck first. Tpu has their own branded memory checker you can use for free. Another good thing to check is the PSU, the older they are the more likely you'll start running into weird issues like this. My seasonic recently kicked the bucket after 9 years of service. You can get a decent power supply now a day's for as low as $40.

 

[Mussels]

Howdy lads,

Currently with a rig that I'm using I'm having an odd issue which I'm not used too..
After the computer has been on for over 5 hours it begins to suffer BSOD constantly with error messages like memory management being the main one but other error messages appear but it's always the same BSOD and restarts.

Thing is once it restarts the issue will arrise again randomly from 1 minute to 20 minutes every time.
It just WILL NOT stop this until the PC is off for a solid couple hours before it can be used for 5+ hours again.

This happens either gaming or just watching youtube and surfing the web. Not actually seems to trigger it no matter what I do.

Sometimes the screen just goes black and stays there until I force restart via the case.


Has anyone any ideas what might be happening here?


Windows 10
AMD Ryzen 1600 @3.8Ghz
16gb DDR4 3000mhz CL18 :/
RX 480 8gb
Gigabyte b450m ds3h Mainboard

i've had similar with the crashes at idle, unfortunately they're a total pain in the ass.
i'm with everyone else here with their suggestions, but you can really only try things one by one until the problem is found.

BIOS updates, check for heat (remember its a slow build up so it a secondary component.. PSU, VRMs, PSU etc)

Good luck, when i had this on my ryzen 1400 build it turned out to be the RAM... but that was resolved with a BIOS update long ago, as a bug from first gen ryzen (initially i couldnt run above 2400Mhz, but after a few updates 3200 was stable)

 

[delshay]

With the very hot ambient temperature of the last two days my computer is crashing, but a small increase in fan speed has fixed this.

But here's something new. My TV is also playing up. Normal transmission is not effected , but the hardrive/USB is acting strange, extremely, slow access & all normal functions has more or less stopped. Tv is turning itself off then on automatically.

This is something most of us must of missed as TV's do have processors & I think this is overheating. If you have a TV that has a USB hardrive connected that is used to pause LIVE TV & is used for TV recoding, access it & see if all functions come to a crawl & stop working. It's working fine today as ambient temperature is a little cooler, but it's still very hot. The question is, do we need to take our TV'S apart to change thermal paste.

 

[Irish_PXzyan]

Thanks for all the suggestions,
I'm thinking perhaps it could be the PSU as i've certainly had this for many years... I believe it's the oldest component in my rig so it could very well be that...

I will run everything on stock for a couple days and see if that sorts the issues. My motherboards VRM looks awful so it very well could be that too.

If all else fails I'll try memtest and pull one of my 8gb ram sticks and see if that sorts it out too!
Tiz gonna be slow and probably will take a few days to figure out this mess :L

Thanks for all the suggestions and I will keep yee posted with my results.

 

[Eskimonster]

Whats most weard to me is that it takes 5 hour to begin with errors, that tells me its hardly the VRM. But more likely the PSU.

 

[Mussels]

Whats most weard to me is that it takes 5 hour to begin with errors, that tells me its hardly the VRM. But more likely the PSU.

heat can build up in dead spots in cases and slowly get worse, applies to both PSU and VRMs
i had a northbridge heatsink on an old PC (DDR2 days) that would hit 110C and crash the PC, but only after 5+ hours of the system being on due to no airflow

 

[spectatorx]

I have very similar problem, you can read about it here:

Spoiler

https://www.reddit.com/r/Amd/comments/chsu02

But i think your case may be easier and most probably it is just a problem with ram. Set it to 2133MHz and see if it helps or just borrow from someone other kit for one day and observe.

 

[Bill_Bright]

I know that Antec used to be known for making quality PSUs but that was a long time ago.

Like many PSU brands, it depends on the specific model and who the OEM provider is for that model. Some Antecs are still good, some I would avoid.

@Irish_PXzyan - you did not say if the case interior is clean and if all fans are spinning.

Certainly, if your room is not air conditioned, ambient (room) temps can play a major role in heat related problems. You might try blasting a desk fan into the open side of your PC. But understand this only blows in room temperature air. This is not living flesh so you cannot cool electronics cooler than the ambient temp. But a fan can help extract heated air from a hot heatsink faster.

And contrary to what some seem to believe, you NEVER need to replace TIM (thermal interface material) just because it is X number of years old. Even if it dries out, the solids that remain behind are still occupying the microscopic pits and valleys, preventing insulating air from getting in there. Yes, TIM does degrade in efficiency a "few" degrees over time. But if your system "needs" those "few" degrees to keep thermal protection modes from kicking in, you have bigger problems than your TIM - like inadequate case cooling, or faulty CPU fan, or excessive overclocking.

The ONLY time you need to replace the TIM is when the cured bond between the mating surfaces is broken. This happens when removing the cooler, when twisting the cooler too hard to see if loose ( ), or if the computer has been otherwise abused (roughly handled/transported/kicked).

Remember, it is the case's responsibility to provide an adequate supply of cool air flowing through the case. The CPU cooler need only toss the CPU's heat into that flow. And it is the user's responsibility to ensure case cooling is setup properly (to include proper cleaning and good cable management).

 

[Irish_PXzyan]

Hi guys I just want to update this thread.

I've tried just setting everything to default without any overclocking or anything and since then.. all week it has not once crashed :O

Now I haven't been pushing the rig very hard but I've had my system running for over 10 hours doing my usual but not heavy gaming but still... it usually would start crashing but not anymore :O

The case and fans are perfectly fine and working

I guess I was just pushing something too far??
Should I at least put my RAM back up to it's stated 3000mhz? I feel like it's a waste not using it's potential?

 

[Solaris17]

Hi guys I just want to update this thread.

I've tried just setting everything to default without any overclocking or anything and since then.. all week it has not once crashed :O

Now I haven't been pushing the rig very hard but I've had my system running for over 10 hours doing my usual but not heavy gaming but still... it usually would start crashing but not anymore :O

The case and fans are perfectly fine and working

I guess I was just pushing something too far??
Should I at least put my RAM back up to it's stated 3000mhz? I feel like it's a waste not using it's potential?

To slowly get performance back leave everything at default like you have now. And simply enable XMP (assuming that will get you to 3000) and see what happens again after time.

 

[Vayra86]

Hi guys I just want to update this thread.

I've tried just setting everything to default without any overclocking or anything and since then.. all week it has not once crashed :O

Now I haven't been pushing the rig very hard but I've had my system running for over 10 hours doing my usual but not heavy gaming but still... it usually would start crashing but not anymore :O

The case and fans are perfectly fine and working

I guess I was just pushing something too far??
Should I at least put my RAM back up to it's stated 3000mhz? I feel like it's a waste not using it's potential?

Did you OC the CPU as well? I would also go with @Solaris17 ''s suggestion right now but still, if you simply put too much volts through the CPU this could be 'fixed' more reliably.

 

[psyko12]

+1 to ^ Put RAM to XMP or manually input it's timings and then run a memory checking tool like one we have here @ TPU and see if it'll do it once more.

 

[Calmmo]

Could be you were on the limit with your VRM and its gotten past its effective lifespan and now performs much worse/cant keep up with the OC requirements. (they dont just.. blow up with nasty bubbles like they used to)

 

[xorbe]

Hi guys I just want to update this thread.

I've tried just setting everything to default without any overclocking or anything and since then.. all week it has not once crashed :O

Now I haven't been pushing the rig very hard but I've had my system running for over 10 hours doing my usual but not heavy gaming but still... it usually would start crashing but not anymore :O

The case and fans are perfectly fine and working

I guess I was just pushing something too far??
Should I at least put my RAM back up to it's stated 3000mhz? I feel like it's a waste not using it's potential?

Glad I nailed the issue. =P Sometimes, securing a stable OC becomes a tremendous pita, and takes forever. Literally, try one thing at a time for a week. One of my tips is to find what appears to be a long term stable overclock that survives various stress testing, then back up a few MHz on both CPU and RAM speed, and bump the memory by +0.03V. Sad to say but I've mostly given up on overclocking these days. I haven't achieved an overclock worth the time invested on a cpu or gpu in a while. I blew so much time trying to get a 5GHz oc out of two different 8086K chips, just for a VR rig though.

 

[Irish_PXzyan]

Heads up!
Setting RAM to XMP mode has resulted in the system crashing again after 5 hours!!

I will attempt to just overclock the CPU and leave RAM set to default and see if it can handle that!

I feel surprised that my RAM running at it's stated clocks aren't stable it seems! UNless it is my motherboard.. My motherboard is only a month old though!

 

[Mussels]

Heads up!
Setting RAM to XMP mode has resulted in the system crashing again after 5 hours!!

I will attempt to just overclock the CPU and leave RAM set to default and see if it can handle that!

I feel surprised that my RAM running at it's stated clocks aren't stable it seems! UNless it is my motherboard.. My motherboard is only a month old though!

memory controller is in the CPU
BIOS controls the automatic/XMP memory timings


both of those could be preventing you reaching full 3000Mhz speeds... if you have a lower XMP profile, give it a try
hell the CPU could be at its OC limit, drop 100-200Mhz off and see if XMP works stable

 

[Irish_PXzyan]

my oh my ladies and gents!
I was gaming with my bro " Mordhau "
BOOM crashed!!

3.7Ghz OC isn't gonna happen for me I guess.

Now I've stuck with dafault RAM 2666mh CL15.
everything else is default.
WIll run for a bit and hope for the best.

Sadly... I honestly believe that thi mobo just isn't up to it sadly...

Does anyone else think PSU is still an issue???
Surely not?

 

[xorbe]

My 2700X + 3200 ram actually run at cpu stock + 3133 because, the cpu didn't oc enough for the hassle, and the ram wouldn't hold 3200 stable. These things happen.