Question Have Been Having Lots of BSODs and Hard Freezes on a New Computer, Finally Narrowed It Down, Need Help Solving It

Page 4 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Feb 12, 2021
60
0
30
0
One other thing that has been questioned, the ability of the 120mm AIO to keep CPU cool.

Do you use any software that tracks temperatures?
something like hwinfo? https://www.hwinfo.com/download/

how is case set up? where is the 120mm radiator mounted?

this isn't yours but is the case - https://www.cyberpowerpc.com/system/Gaming-Instant-Ship-GXL-99101
pictures don't match system as it shows air cooling and mentions the AIO. So curious where its mounted
I am assuming the apevia are 2 on front and 1 on rear?
2 in front, and I think that's it for case fans. The radiator/fan is mounted to the back of the case, about where motherboard I/O panel is. I was keeping track of temperatures for a while, and while they are definitely higher than a lot of people report online, I live in Florida and my A/C in the apartment doesn't dehumidify well. It is always within safe ranges, even when running prime95 for an extended time. It seems to be sitting at around 35-40C while not much is going on, and I think it got up to around the 70s, maybe 80s C during Prime95
 
Last edited:

Colif

Win 10 Master
Moderator
It seems to be sitting at around 35-40C while not much is going on, and I think it got up to around the 70s, maybe 80s C during Prime95
those temps are about what I get in winter, summer... i don't know yet as last year wasn't normal for here (Australia) and cool... nice, wish all were like it, but anyway.

when do most crashes occur? when at idle or when playing games?
Just curious what temps of CPU are like when GPU is also in the picture, Prime only stresses CPU but I am fairly sure a 3090 added to mix might increase temps. And if the AIO is on back of case as exhaust, it might get a bit warm. I see top of case is mesh at least.

I guess 2 in, 1 out is standard fan layout
 
Feb 12, 2021
60
0
30
0
when do most crashes occur? when at idle or when playing games?

A lot of crashes occur when nothing is happening. Actually, it feels like the most crashes occur while doing next to nothing, after closing something I was doing, or while nothing is happening at all (like I got up and stepped away for a bit).

Also, very consistently, after it freezes or crashes once, it crashes a few more times, within a few minutes of being turned on each time. Also, once those repeat crashes are done, it might actually be able to do stuff for a while after, like the 12 hour prime95 test.

As for games, I don't think it's crashed a single time while playing DotA 2, though it often crashes once I close DotA 2 and step away for a few minutes. I can play a heavily modded Minecraft game with shaders running for hours just fine, yet if I open the game and walk away while it's loading, it often crashes if it sits at the menu a while (but not always). Lastly, Total War: Warhammer 2 (a demanding game for the cpu and gpu) will ALWAYS crash the pc; sometimes it takes an hour, sometimes a few hours, but it had always crashed. Last time I tried to play it, the pc hard froze while in a loading screen within minutes of starting up the game.

I truly think the most crashes occur while I'm idling, or doing little tasks, like having a YouTube video play or streaming Plex to another room. There's also a consistency of it crashing after I close a demanding task, like a game, and let it idle for a few minutes after.

This is all conjecture and small patterns I've noticed, but none of them occur 100% of the time (except Total War Warhammer 2, that crashes every time eventually).
 

Colif

Win 10 Master
Moderator
crashes at idle are common problems with bad PSU, but not with Seasonic Primes. Its not PSU as it happened before you replaced it.
its not mb as its new and it happened before
AIO must be able to cope then. I expect all the 3090 heat goes out top of case anyway.

You need to do the stripped down build on weekend and see if it still occurs. Shame you can't get a CPU sent out but thats wishful thinking
 
Feb 12, 2021
60
0
30
0
crashes at idle are common problems with bad PSU, but not with Seasonic Primes. Its not PSU as it happened before you replaced it.
its not mb as its new and it happened before
AIO must be able to cope then. I expect all the 3090 heat goes out top of case anyway.

You need to do the stripped down build on weekend and see if it still occurs. Shame you can't get a CPU sent out but thats wishful thinking

Is the CPU still a likely culprit, even with prime95 and the processor tool passing? If so, is it the most likely? I still plan on testing with a safe mode+fresh install with as few things plugged in externally as possible, and I still need to check the wifi card by disabling it.
 

Colif

Win 10 Master
Moderator
The only reason CPU is questioned is because its one of the few parts that wasn't replaced. There is no rock solid proof it is CPU that I saw.**
Its possible swapping CPU isn't answer but then we left looking at what wasn't replaced.
Only way to know is to use another CPU in PC and see if it happens. Its why I suggested repair store as could hope they have a known working CPU they can swap in and test the games that won't play now.

** I am not perfect, there are people out there who can read these dumps better than I can but they just don't stick around here long enough.
Did I suggest you try asking on other sites yet?
Not giving up but no reason you can't try sites like:
Sysnative
Tenforums (they seem to always blame CPU for the people I send there)
Bleeping Computer

you can link to this thread to save time if you want.
 

Vic 40

Titan
Ambassador
Just a thought, see people point at power delivery and psu is replaced, but if the power delivery in the appartment complex you live in isn't good might that maybe give problems. So have you heard about others having problems with that? Read about the airco that doesn't work well so that maybe something else that interferes.
 
Reactions: Colif
Feb 12, 2021
60
0
30
0
Just a thought, see people point at power delivery and psu is replaced, but if the power delivery in the appartment complex you live in isn't good might that maybe give problems. So have you heard about others having problems with that? Read about the airco that doesn't work well so that maybe something else that interferes.
I haven't heard of anybody else having issues, and I think the guy in the apartment above me also has a custom built computer. I've also used my old computer before at the same location with no issues for about a year, that one was a custom built with a 4790k and a 1080. That said, blaming my apartment wouldn't surprise me; despite spending over 1400$ a month on rent, it's had more issues than it should, and receives shoddy maintenance at that. I'll see if I can't borrow a UPS from work and see if that helps anything, but it'll be a while before that can happen (months) due to everything being closed down at the moment.


Edit: Following some advice @ Bleeping Computer, I changed the RAM speed manually (off of auto or XMP) to 2933 MHz, the rated speed for the processor, and it seemed to last longer than usual. I managed to play Total War for almost 6 hours, and I even left it on and streamed to Plex for almost 2 hours before that, and hadn't had a single crash yet today -- until now.

Dump file here: https://www.dropbox.com/s/92eaig0nvtp53pu/090521-7796-01.dmp?dl=0

Strangely enough, CPU-Z says this about my RAM:
This: View: https://i.imgur.com/ITQwNTC.png

This: View: https://i.imgur.com/UCBY3iX.png


Is it normal for the unit to be listed as 3200, which the XMP says it is, but the actual Max Bandwidth is 2400, which lines up with the JEDEC. Of note, the current DRAM frequency is about 1465-1466, which is 2933/2, so lines up with what I set manually.
 
Last edited:

Vic 40

Titan
Ambassador
Might be ram problem then,maybe incompatibility. Can you test with ram at 2666mhz? How does it run that? Can you test with another set? Maybe from a friend or ....

To really know would i rather see the next list so i can see what ram is in there (partnr) and what speeds it can run.

download hwinfo,
install and open it=click run,
close the top window which is the system summary,
in the main window at the left top click "save report",
at the bottom of the next window check "Summary for Clipboard",
after that you'll see what's in the pc,
copy by clicking "copy to clipboard" and rightclick+paste in your next respons
The cpu-z tool does list jedec speeds under max bandwith so 2400 seems not odd.
 
Reactions: Colif
Feb 12, 2021
60
0
30
0
Might be ram problem then,maybe incompatibility. Can you test with ram at 2666mhz? How does it run that? Can you test with another set? Maybe from a friend or ....

To really know would i rather see the next list so i can see what ram is in there (partnr) and what speeds it can run.



The cpu-z tool does list jedec speeds under max bandwith so 2400 seems not odd.

I've set it to 2666 MHz for now, will report back how it goes.

Here is the requested info from hwinfo:

Computer: ASUS System Product Name
CPU: Intel Core i9-10850K (Comet Lake-S, Q0)
3600 MHz (36.00x100.0) @ 4796 MHz (48.00x99.9)
Motherboard: ASUS PRIME Z490-V
BIOS: 1602, 01/14/2021
Chipset: Intel Z490 (Comet Lake PCH-H)
Memory: 16384 MBytes @ 1332 MHz, 15-17-17-36
- 8192 MB PC25600 DDR4 SDRAM - Team Group TEAMGROUP-UD4-3200
- 8192 MB PC25600 DDR4 SDRAM - Team Group TEAMGROUP-UD4-3200
Graphics: EVGA RTX 3090 FTW3 ULTRA GAMING (24G-P5-3982)
NVIDIA GeForce RTX 3090, 24576 MB GDDR6X SDRAM
Drive: ST3000DM007-1WY10G, 2930.3 GB, Serial ATA 6Gb/s @ 6Gb/s
Drive: WDC WDS100T2B0C-00PXH0, 976.8 GB, NVMe
Sound: Intel Comet Lake PCH-H - cAVS (Audio, Voice, Speech) [A0]
Sound: NVIDIA GA102 - High Definition Audio Controller
Network: RealTek Semiconductor, Device ID: C821
Network: RealTek Semiconductor RTL8168/8111 PCI-E Gigabit Ethernet NIC
OS: Microsoft Windows 10 Home (x64) Build 19043.1165 (21H1)
 

gardenman

Distinguished
Moderator
I ran the dump files through the debugger and got the following information: https://jsfiddle.net/1nL2dv30/show This link is for anyone wanting to help. You do not have to view it. It is safe to "run the fiddle" as the page asks.
File information:090521-7796-01.dmp (Sep 5 2021 - 00:10:14)
Bugcheck:KMODE_EXCEPTION_NOT_HANDLED (1E)
Probably caused by:ntkrnlmp.exe (Process: System)
Uptime:0 Day(s), 11 Hour(s), 22 Min(s), and 32 Sec(s)

File information:090421-7500-01.dmp (Sep 4 2021 - 01:50:23)
Bugcheck:UNEXPECTED_KERNEL_MODE_TRAP_M (1000007F)
Probably caused by:ntkrnlmp.exe (Process: Spotify.exe)
Uptime:0 Day(s), 1 Hour(s), 55 Min(s), and 41 Sec(s)

File information:090321-8812-01.dmp (Sep 3 2021 - 23:54:24)
Bugcheck:ATTEMPTED_EXECUTE_OF_NOEXECUTE_MEMORY (FC)
Probably caused by:memory_corruption (Process: OriginWebHelpe)
Uptime:0 Day(s), 1 Hour(s), 52 Min(s), and 15 Sec(s)

File information:090121-7250-01.dmp (Sep 1 2021 - 11:49:19)
Bugcheck:CLOCK_WATCHDOG_TIMEOUT (101)
Driver warnings:*** WARNING: Unable to verify timestamp for RTKVHD64.sys
Probably caused by:Unknown_Image (Process: Registry)
Uptime:0 Day(s), 0 Hour(s), 55 Min(s), and 26 Sec(s)

File information:090121-7171-01.dmp (Sep 1 2021 - 14:35:13)
Bugcheck:SYSTEM_SERVICE_EXCEPTION (3B)
Probably caused by:ntkrnlmp.exe (Process: tv_x64.exe)
Uptime:0 Day(s), 2 Hour(s), 40 Min(s), and 06 Sec(s)
This information can be used by others to help you. Someone else will post with more information. Please wait for additional answers. Good luck.
 

Colif

Win 10 Master
Moderator
BIOS update can't hurt.

clock watchdog times out report they are caused by realtek audio drivers.


I don't know if its just picking on them now because of ram or if they really are cause. You had a crash on 17th Feb (page 1) that could have been either realtek or Sound blaster and this was your answer

As for audio, I almost exclusively use the Sound Blaster for audio output (digital headphones connected to it), and for input, I use a usb dongle to connect a modmic wireless. As far as I know, I am not particularly using the realtek stuff.
I know you removed SB to see if BSOD would stop... seems not. If you didn't use the backpanel audio for anything its shouldn't have been realtek.

You still have same realtek Audio drivers as you did in February, is it worth updating them or will bsod just pick another driver next? It doesn't feel likely all 5 are same driver.

had any since changing ram speed?
 
Feb 12, 2021
60
0
30
0
clock watchdog times out report they are caused by realtek audio drivers.


I don't know if its just picking on them now because of ram or if they really are cause. You had a crash on 17th Feb (page 1) that could have been either realtek or Sound blaster and this was your answer



I know you removed SB to see if BSOD would stop... seems not. If you didn't use the backpanel audio for anything its shouldn't have been realtek.

You still have same realtek Audio drivers as you did in February, is it worth updating them or will bsod just pick another driver next? It doesn't feel likely all 5 are same driver.
I'll look into the realtek drivers and see if I can't find an update, though if I remember correctly I have the latest version -- but to be fair, I'm not great at finding proper drivers. Its the one cybersleuthing skill I never picked up.

There are two newer bios revisions so might want to try the latest as well.
Have not done this yet, because...


had any since changing ram speed?
I have had no crashes as of now, after lowering it even further, to 2666 MHz. I have not ran Total War for any amount of time yet, and I turn my computer off at night, but even so, I was still using the computer for a good chunk of the day yesterday.

I'll look into finding Realtek drivers, and a BIOS update won't hurt, so I'll look into that, but for now, I'll keep trying to crash it to see if the problem is.. well, I wouldn't say fixed, as opposed to it happens little enough that it isn't really a big problem, which at this point, is a win in my book.
 

Colif

Win 10 Master
Moderator
sound drivers suck - they used to be easy but realtek confused things by having 2 types and I am never sure which ones to suggest now.

you have
Dec 22 2020RTKVHD64.sysRealtek Audio System driver https://www.realtek.com/en/

realtek have 2 types of drivers,
the ones that come with HD Audio Manager
the ones where you get drivers from Realtek and the app from Windows store

yours are the 2nd type
these are drivers for ROG series, I don;'t know what the difference might be for yours, they should be okay - https://rog.asus.com/forum/showthread.php?117596-DRIVERS-Realtek-Audio-(MB-Intel-2xx-3xx-4xx-amp-AMD-3xx-4xx-5xx)
 
Last edited:

gardenman

Distinguished
Moderator
I ran the dump file through the debugger and got the following information: https://jsfiddle.net/e5rjxm24/show This link is for anyone wanting to help. You do not have to view it. It is safe to "run the fiddle" as the page asks.

File information:090721-7984-01.dmp (Sep 7 2021 - 21:31:11)
Bugcheck:MEMORY_MANAGEMENT (1A)
Probably caused by:memory_corruption (Process: System)
Uptime:0 Day(s), 12 Hour(s), 09 Min(s), and 19 Sec(s)

This information can be used by others to help you. Someone else will post with more information. Please wait for additional answers. Good luck.
 
Feb 12, 2021
60
0
30
0
As it happens, I was busy this weekend with work from home stuff, so I didn't have time to do any BIOS updates or research. Today, on a whim, I decided to try and do some Total War as a crash-test, since I hadn't crashed since the 7th, the last time I tried TWW2, and sure enough, within 20 minutes of booting into it, it crashed. With the memory speed set at 2666 MHz, it seems like it doesn't crash EXCEPT for when in Total War: Warhammer 2. This time, I ran it without any mods, to verify that wasn't an issue. TWW2 is certainly a demanding game, that taxes the CPU (Total War games are massive army vs. army battles with lots of AI processing), GPU (lots of models), and RAM (lots of stuff to load), as well as the SSD (loads between battle maps and campaign map quite often), but even so..

Anyways, here's the dump file: https://www.dropbox.com/s/x8om9sddg2v7unt/091321-7875-01.dmp?dl=0
 

Colif

Win 10 Master
Moderator
You don't need to apologize for not doing something, its your problem. We probably be here somewhere anyway.
I seen total war games, I can only imagine they worse than before, since everything looks better now.

last error (Wednesday) occurred in the page file. C drive. I need to look through this thread and see what we done before I suggest anything based on that.

on reflection, it didn't touch the drive. it accessed ram, then CPU tried to access ram it didn't control, then it had a Userfault, and I am not sure what that is. Then it had the action I thought was C drive but it might be ram or CPU.

the part where it tried to access ram it doesn't control makes me think its a driver
 
Last edited:

gardenman

Distinguished
Moderator
I ran the dump file through the debugger and got the following information: https://jsfiddle.net/n4px1v2d/show This link is for anyone wanting to help. You do not have to view it. It is safe to "run the fiddle" as the page asks.

File information:091321-7875-01.dmp (Sep 13 2021 - 15:53:54)
Bugcheck:IRQL_NOT_LESS_OR_EQUAL (A)
Probably caused by:hardware (Process: System)
Uptime:0 Day(s), 6 Hour(s), 53 Min(s), and 17 Sec(s)

Comment: It's rare for a minidump to directly say "hardware" in the text.

This information can be used by others to help you. Someone else will post with more information. Please wait for additional answers. Good luck.
 
Feb 12, 2021
60
0
30
0
I was just trying out a new game that came out, Pathfinder: Wrath of the Righteous, and my computer.... hard froze/crashed? My screen went faded/black, but some colors/pixels remained and got glitchy looking, I don't know how to describe it, but I've seen it before but can't recall where. Anywho, my guess, based on what I saw, was it overheated. It restarted itself after a few moments, and when it loaded up, I immediately opened HWMonitor and saw the CPU (I think) was at/slightly over 100C, which is definitely bad. HWMonitor even painted it red.

So now I have a question -- this is my first time using liquid cooling, and since I didn't build it, I don't have experience with it or how it should work. What my research tells me, is the pump should be running at 100% all the time, and the fan on the back of the case that the cables run to should be running dynamically with PWM like any modern fan does. How can I tell if my pump is even working/running, or if the fan is doing things properly? When I look at HWMonitor, there are only 2 entries:

View: https://i.imgur.com/Vx9illO.png


Neither seem to be running at a constant, high RPM.. does the pump normally show up here? I have no idea how this works, but the PC shouldn't be overheating if everything is working properly. Maybe this issue ties in with everything else, though no guarantees there.

A note about the setup: There seem to be two fans at the front of the case, I assume running off the same cable?, and then the AIO, which the fan is mounted to the back of the case.

Edit: Got the "Armoury Crate" software from ASUS, this is what it shows.. : View: https://i.imgur.com/cYciOIA.png


So... the AIO pump isn't running?

Edit 2: Found these images in the manual for my motherboard: View: https://i.imgur.com/0CaQix3.png


and View: https://i.imgur.com/WnamVEj.png


At a cursory glance, I don't see a cable running from the pump that sits on the CPU like in image one anywhere. I'd have to unplug it all from the wall and open it up to really see, which I won't do right now as I'm going to bed, but, assuming they should all basically have some cable running from the pump to the motherboard, mine... at least seems like it doesn't. Certainly not a good thing, and I'm surprised, if this is the case, that its been running so many games as is. It could very well be that all the blue screens, at least now, that only seem to happen when playing Total War, a very CPU intensive game, is.. well, due to it stressing it. That said, Prime95 ran fine over night so who knows..


Edit 3: Even though the documentation is for my motherboard, it doesn't seem like it fully matches what I have, or at least, from the side panel, I don't see the AIO connection port.
And now I have gone back into the manual and found a more accurate depiction:
View: https://i.imgur.com/p9aao9S.png


But it still does not look like there is a cable in that area at all. After looking up my specific cooler, there is a single cable running out the top right in the images, which I can see, but I can't verify where its going to easily.

Based on the picture here: View: https://i.imgur.com/Us6UDGy.png


I can tell there definitely should be a cable for both the pump and cpu fan.
 
Last edited:

Colif

Win 10 Master
Moderator
So at least cyberpower don't make their own AIO

according to photos of AIO on link above, the 2 cables run to pump from fan/radiator and there should be a USB cable from pump to a header on mb. So that it can regulate fan speeds. the cable for the fans must be inside the cables for the liquids. All the power must be drawn through the USB cable. I don't think you missing anything unless you don't have the USB cable.

2.47 into video shows what cables are attached to pump

sux you can't even download a manual or quick start guide to compare to.

my fans always spin, its a CPU cooler, it shouldn't just stop. If it were a 360 or something i could understand but its a 120. it should always spin at a consistent rate. My 240 fans only stop if I choose zero rpm mode, otherwise minimum is 300rpm
I leave my pump speed alone, i can change it but i haven't seen the need to yet.

open case and take photos of what you see.

your fan speeds are slightly more believable than mine

but CPU fan is really my pump and I suspect its not even right there, as Icue sees different scores

2 of my case fans not linked to mb as far as I know. I need to look at that.
 
Last edited:

ASK THE COMMUNITY

TRENDING THREADS