Question New CPU + AIO cooler upgrade, now causing black screen + fans on full blast

kunnome

Honorable
Nov 23, 2016
21
1
10,515
0
Hello,

I recently purchased a CPU upgrade (Ryzen 9 5900x) as well as an AIO cooler (NZXT Kraken X53 RGB) and installed them yesterday.
The computer seemed to be working as intended after boot for awhile and suddenly both my screens went black and my fans went on full blast. Nothing was responsive and I wasn't able to hold my power button to force shut down the computer, so I resorted to flipping the I/O switch on my psu. Turned it back on and it worked normally for awhile, only to have it happen again.

I did some googling into similar issues and tried my hands at troubleshooting the problem. The problem persisted and I ended up making one last move that "seems" to have fixed it but I am not sure if this is the correct and safe method. So I wanted to ask here to get some opinions.

I'll briefly list the steps I took so you can track what I've done.

Step 1) I initially tried to rewire the AIO pump + radiator fans setup I had since I read somewhere that the CPU_FAN mobo header could be having issues detecting CPU temp and caused the computer to shut down to protect itself from excessive heating.
I plugged the pump 3-pin directly to CPU_FAN and I used a splitter to connect the radiator fans to a separate SYS_FAN header.
Step 1 Result) System booted fine, same issue happened some minutes into using the computer.

Step 2) Installed a few different system monitoring software to track the temperature, load, fan rpm of both CPU and GPU. I installed NZXT CAM, HWINFO64, and OCCT.
Step 2 Result) Both NZXT CAM and HWINFO64 showed relatively stable temperature for both CPU and GPU, ranging between 36 - 41 Celsius. A 12 min stress test on OCCT showed some "errors found on physical cores", but when I googled about this the responses online were too complicated for me to understand as they were talking about adding/sending voltages(??). Temperature of the CPU for the duration of the stress test was around 60 Celsius.

Step 3) I read somewhere that the mobo could be mistaking the AIO pump's signals and could be forcefully shutting down the protect itself. I went into BIOS and disabled the "over temperature protection" feature.
Step 3 Result) This step seems to be giving me the most promising result. It has been over an hour since I disabled the protection feature and the computer has not had the issue since then.

Conclusion / Question : Is it okay to leave the protection feature disabled? The CPU and GPU temps are being monitored and they are holding high 30s to low 40s when I'm browsing the internet and as I am writing this post. Also, should I be concerned about the errors found on physical cores messages during the OCCT stress test? I don't plan to OC at all, but the tool gave error messages, so now I am wondering if this is all happening due to a defective CPU.

Sorry for the long post.
Thank you!

Build:
CPU (old): Ryzen 5 1600
CPU (new): Ryzen 9 5900x
AIO: NZXT Kraken X53 RGB
GPU: RTX 2070
Mobo: ASRock AB350M Pro4 (BIOs updated fully)
PSU: 650W (about 2 year old)
 
Hello,

I recently purchased a CPU upgrade (Ryzen 9 5900x) as well as an AIO cooler (NZXT Kraken X53 RGB) and installed them yesterday.
The computer seemed to be working as intended after boot for awhile and suddenly both my screens went black and my fans went on full blast. Nothing was responsive and I wasn't able to hold my power button to force shut down the computer, so I resorted to flipping the I/O switch on my psu. Turned it back on and it worked normally for awhile, only to have it happen again.

I did some googling into similar issues and tried my hands at troubleshooting the problem. The problem persisted and I ended up making one last move that "seems" to have fixed it but I am not sure if this is the correct and safe method. So I wanted to ask here to get some opinions.

I'll briefly list the steps I took so you can track what I've done.

Step 1) I initially tried to rewire the AIO pump + radiator fans setup I had since I read somewhere that the CPU_FAN mobo header could be having issues detecting CPU temp and caused the computer to shut down to protect itself from excessive heating.
I plugged the pump 3-pin directly to CPU_FAN and I used a splitter to connect the radiator fans to a separate SYS_FAN header.
Step 1 Result) System booted fine, same issue happened some minutes into using the computer.

Step 2) Installed a few different system monitoring software to track the temperature, load, fan rpm of both CPU and GPU. I installed NZXT CAM, HWINFO64, and OCCT.
Step 2 Result) Both NZXT CAM and HWINFO64 showed relatively stable temperature for both CPU and GPU, ranging between 36 - 41 Celsius. A 12 min stress test on OCCT showed some "errors found on physical cores", but when I googled about this the responses online were too complicated for me to understand as they were talking about adding/sending voltages(??). Temperature of the CPU for the duration of the stress test was around 60 Celsius.

Step 3) I read somewhere that the mobo could be mistaking the AIO pump's signals and could be forcefully shutting down the protect itself. I went into BIOS and disabled the "over temperature protection" feature.
Step 3 Result) This step seems to be giving me the most promising result. It has been over an hour since I disabled the protection feature and the computer has not had the issue since then.

Conclusion / Question : Is it okay to leave the protection feature disabled? The CPU and GPU temps are being monitored and they are holding high 30s to low 40s when I'm browsing the internet and as I am writing this post. Also, should I be concerned about the errors found on physical cores messages during the OCCT stress test? I don't plan to OC at all, but the tool gave error messages, so now I am wondering if this is all happening due to a defective CPU.

Sorry for the long post.
Thank you!

Build:
CPU (old): Ryzen 5 1600
CPU (new): Ryzen 9 5900x
AIO: NZXT Kraken X53 RGB
GPU: RTX 2070
Mobo: ASRock AB350M Pro4 (BIOs updated fully)
PSU: 650W (about 2 year old)
if you try making it loose, and it worked (albeit the temp), then your mounting might be too tight. but it also could be a failing cpu, maybet better get the rma if the loosening the mount didnt work.
 
Try this. Pomp to a header where you can make it run at full speed, all the time. Preferably to AIO_Pump header. Radiator fans to CPU_FAN header. Make sure that 3pin header is set to Voltage regulation a 4pin to PWM in BIOS.
Top of radiator should be higher than pump and if possible, hoses at bottom of radiator for best effect.
 

kunnome

Honorable
Nov 23, 2016
21
1
10,515
0
if you try making it loose, and it worked (albeit the temp), then your mounting might be too tight. but it also could be a failing cpu, maybet better get the rma if the loosening the mount didnt work.
Thank you for the reply!
When you say "loosen the mount," do you mean the AIO pump that is sitting on top of the CPU?
I went back and unscrewed the pump and only tightened it moderately using my fingers (they are thumb screws to begin with).

I will have to test it out a bit more to see if I need to RMA the CPU since it is a hassle and all. I'll try gaming on it today a bit to put it to the test.
 

kunnome

Honorable
Nov 23, 2016
21
1
10,515
0
Try this. Pomp to a header where you can make it run at full speed, all the time. Preferably to AIO_Pump header. Radiator fans to CPU_FAN header. Make sure that 3pin header is set to Voltage regulation a 4pin to PWM in BIOS.
Top of radiator should be higher than pump and if possible, hoses at bottom of radiator for best effect.
Thank you for the reply!
Unfortunately I do not have an AIO_Pump header on my motherboard as it is rather old.

The only 3 headers I have available are:
one 4-pin CPU_FAN1,
one 4-pin CHA_FAN1, and
one 3-pin CHA_FAN2.

The current setup is:
AIO pump 3pin is plugged into the 4-pin CPU_FAN1
2 radiator fans + 1 rear case fan via fan splitter is connected to the 4-pin CHA_FAN1
1 top case fan is plugged into the 3-pin CHA_FAN2

What I tried doing for now is:
I went into BIOS and changed CPU Fan 1 Setting from 'Standard Mode' to 'Full Speed'. Left "Over Temperature Protection" on 'Disabled.

I will have to test it out a bit more to see if I need to RMA the CPU since it is a hassle and all. I'll try gaming on it today a bit to put it to the test.

**edited after posting: Confirmed that the top point of the radiator is indeed higher than the pump, and the hoses are at the bottom of the radiator! Also, I'm not sure where to set voltage regulation to the 4-pin CPU_FAN1. My BIOS has a setting labeled "CPU Fan 1 Temp Source" that is apparently set to "Monitor CPU." Is this maybe it?
 
Last edited:
Reactions: Koekieezz

kunnome

Honorable
Nov 23, 2016
21
1
10,515
0
So, a bit of an update + some follow up questions:

I tried to game a bit on the system yesterday, and the longest I got through my gaming sessions without the screen going black and fans going full blast was about 30 min. max. Sometimes the thing would happen right when I launch the game.

I read that completely uninstalling the GPU drivers with the DDU software and reinstalling the latest drivers might help, but no luck there too.

I finally got fed up with it and tried putting my old CPU back in to see if the issue came from the AIO set up and not the CPU.
The computer's been holding stable since then, with hours of gaming and no problem.

This tells me that it's the CPU that's causing this.
Also, something to note: With the new CPU, my computer would boot up with the fans on full blast for a few seconds, which would then slow down after. With the old CPU back in, the fans don't do this, even with the AIO still installed.

Would this point to a faulty, defective CPU, or can it be a compatibility issue with my motherboard & ram? The reason why I'm asking is I'm now contemplating whether to hold onto this CPU until I get the chance to upgrade my motherboard and ram together, or to try and refund/exchange this CPU.
 
So, a bit of an update + some follow up questions:

I tried to game a bit on the system yesterday, and the longest I got through my gaming sessions without the screen going black and fans going full blast was about 30 min. max. Sometimes the thing would happen right when I launch the game.

I read that completely uninstalling the GPU drivers with the DDU software and reinstalling the latest drivers might help, but no luck there too.

I finally got fed up with it and tried putting my old CPU back in to see if the issue came from the AIO set up and not the CPU.
The computer's been holding stable since then, with hours of gaming and no problem.

This tells me that it's the CPU that's causing this.
Also, something to note: With the new CPU, my computer would boot up with the fans on full blast for a few seconds, which would then slow down after. With the old CPU back in, the fans don't do this, even with the AIO still installed.

Would this point to a faulty, defective CPU, or can it be a compatibility issue with my motherboard & ram? The reason why I'm asking is I'm now contemplating whether to hold onto this CPU until I get the chance to upgrade my motherboard and ram together, or to try and refund/exchange this CPU.
Well. R9 5900x is much hotter processor than R5 1600, Twice the cores and allowed up to 90c without dropping the boost.
 
Reactions: Koekieezz

ASK THE COMMUNITY