F5F Stay Refreshed Hardware Desktop PC display flickers with green and black, then restarts immediately.

PC display flickers with green and black, then restarts immediately.

PC display flickers with green and black, then restarts immediately.

G
GC_Lewk
Member
219
12-13-2021, 04:15 AM
#1
Hi everyone. I’ve been dealing with a really frustrating PC crashing issue for a long time now and I’m honestly running out of ideas, so I’d really appreciate some opinions.
My PC crashes completely at random times. The screen turns green or black, the audio buzzes for about half a second (sometimes it doesnt buzz) , and then the system instantly restarts on its own. There’s no BSOD and no error message. And What's confusing me is that it does not happen under load specifically, it may happen while gaming (in game), or when im sitting in a lobby looking for a match, or sometimes even when the GPU usage is low, in the menu.
I've been diagnosing this issue for a few months now, I tried like 14 different AMD drivers, I tried 23.11.1, 23.12.1, 24.5.2, 24.12.1 25.10.2, 25.11.1, 25.12.1, 26.1.1 but it always keeps crashing, I’ve tried GPU stock settings, I tried undervolting both the CPU and GPU, forced PCIe x16 to Gen 3 in the BIOS, tested multiple BIOS versions, disabled XMP, and adjusted the power limit in AMD Adrenalin, but none of these made a difference. Crashes still happen regardless. I would like to note that the PC never crashes when im on desktop
So in the beginning I suspected that it I had a faulty PSU, so I downloaded HWINFO64, i ran furmark for an hour, heaven benchmark 4.0, OCCT and CINEBENCH R24 and surprisingly my PC never restarted during the stress test, so I kept checking HWINFO to monitor temps, my GPU never goes above 80 (The GPU hotspot stays at 98 but when i undervolt it stays at 87), my CPU stays at 77 and HWInfo also shows the 12V rail staying around ~12.2V (so no sudden power dips)
Now The PSU is a Connect 80 Plus Bronze 650W, which I know isn’t a high-end or well-known unit. I also know that the RX 5700 XT is known for transient power spikes, which makes me wonder if the PSU could be shutting down briefly even if monitoring software doesn’t catch it. But at the same time, the crash pattern (green/black screen when gaming, buzzing audio, instant reboot) also seems like something people report with unstable or failing 5700 XT cards.
My question is: Based on these symptoms and everything I’ve tested so far, does this sound more like a failing RX 5700 XT or a faulty PSU?
PC Specs:​
GPU:
PowerColor RX 5700 XT 8GB (dual fan)
CPU:
Ryzen 5 3600
Motherboard:
ASUS B450M Pro R2.0 (Bios is up to date to the latest NON beta version, chipset drivers are also updated.)
RAM:
16GB (2x8GB) DDR4 3200MHz T-FORCE (Tried turning XMP OFF so im currenrtly running 2400)
PSU:
Connect 80 Plus Bronze 650W
OS:
Windows 11 (Updated it to the latest)
G
GC_Lewk
12-13-2021, 04:15 AM #1

Hi everyone. I’ve been dealing with a really frustrating PC crashing issue for a long time now and I’m honestly running out of ideas, so I’d really appreciate some opinions.
My PC crashes completely at random times. The screen turns green or black, the audio buzzes for about half a second (sometimes it doesnt buzz) , and then the system instantly restarts on its own. There’s no BSOD and no error message. And What's confusing me is that it does not happen under load specifically, it may happen while gaming (in game), or when im sitting in a lobby looking for a match, or sometimes even when the GPU usage is low, in the menu.
I've been diagnosing this issue for a few months now, I tried like 14 different AMD drivers, I tried 23.11.1, 23.12.1, 24.5.2, 24.12.1 25.10.2, 25.11.1, 25.12.1, 26.1.1 but it always keeps crashing, I’ve tried GPU stock settings, I tried undervolting both the CPU and GPU, forced PCIe x16 to Gen 3 in the BIOS, tested multiple BIOS versions, disabled XMP, and adjusted the power limit in AMD Adrenalin, but none of these made a difference. Crashes still happen regardless. I would like to note that the PC never crashes when im on desktop
So in the beginning I suspected that it I had a faulty PSU, so I downloaded HWINFO64, i ran furmark for an hour, heaven benchmark 4.0, OCCT and CINEBENCH R24 and surprisingly my PC never restarted during the stress test, so I kept checking HWINFO to monitor temps, my GPU never goes above 80 (The GPU hotspot stays at 98 but when i undervolt it stays at 87), my CPU stays at 77 and HWInfo also shows the 12V rail staying around ~12.2V (so no sudden power dips)
Now The PSU is a Connect 80 Plus Bronze 650W, which I know isn’t a high-end or well-known unit. I also know that the RX 5700 XT is known for transient power spikes, which makes me wonder if the PSU could be shutting down briefly even if monitoring software doesn’t catch it. But at the same time, the crash pattern (green/black screen when gaming, buzzing audio, instant reboot) also seems like something people report with unstable or failing 5700 XT cards.
My question is: Based on these symptoms and everything I’ve tested so far, does this sound more like a failing RX 5700 XT or a faulty PSU?
PC Specs:​
GPU:
PowerColor RX 5700 XT 8GB (dual fan)
CPU:
Ryzen 5 3600
Motherboard:
ASUS B450M Pro R2.0 (Bios is up to date to the latest NON beta version, chipset drivers are also updated.)
RAM:
16GB (2x8GB) DDR4 3200MHz T-FORCE (Tried turning XMP OFF so im currenrtly running 2400)
PSU:
Connect 80 Plus Bronze 650W
OS:
Windows 11 (Updated it to the latest)

_
_Lolikc_
Member
184
12-17-2021, 04:03 AM
#2
Welcome to the forums, newcomer!
Now The PSU is a Connect 80 Plus Bronze 650W, which I know isn’t a high-end or well-known model.
+
PSU:
Connect 80 Plus Bronze 650W
Is this the one you’re using? If yes, I’m unfamiliar with this unit until today. If I were you, I’d choose a solid, reliable power supply with 750W capacity and check if the problem continues. To be clear, it’s not about premium specs—it’s about avoiding an unreliable device, as things seem uncertain at the moment.
Motherboard:
ASUS B450M Pro R2.0 (Bios is up to date to the latest non-beta version, and chipset drivers are also current.)
To prevent confusion, could you confirm whether you’re using an ASUS or an ASRock board? They differ even if you combine both brands;
https://www.asrock.com/mb/AMD/B450M Pro4 R2.0/index.asp
https://www.asus.com/motherboards-compon...0m-pro-ii/
So, which one is it? Please also mention the BIOS version instead of just saying “latest.” We have several threads suggesting “latest” to keep everyone informed about pending updates.
Add the make and model of your case, along with its airflow and storage details. Event Viewer logs and .dmp files would be helpful if you’ve seen BSoD errors.
I’ve been troubleshooting this for a few months now. I’ve tested 14 different AMD drivers—23.11.1, 23.12.1, 24.5.2, 24.12.1, 25.10.2, 25.11.1, 25.12.1, 26.1.1.
Can you explain how you switched from one driver version to another?
_
_Lolikc_
12-17-2021, 04:03 AM #2

Welcome to the forums, newcomer!
Now The PSU is a Connect 80 Plus Bronze 650W, which I know isn’t a high-end or well-known model.
+
PSU:
Connect 80 Plus Bronze 650W
Is this the one you’re using? If yes, I’m unfamiliar with this unit until today. If I were you, I’d choose a solid, reliable power supply with 750W capacity and check if the problem continues. To be clear, it’s not about premium specs—it’s about avoiding an unreliable device, as things seem uncertain at the moment.
Motherboard:
ASUS B450M Pro R2.0 (Bios is up to date to the latest non-beta version, and chipset drivers are also current.)
To prevent confusion, could you confirm whether you’re using an ASUS or an ASRock board? They differ even if you combine both brands;
https://www.asrock.com/mb/AMD/B450M Pro4 R2.0/index.asp
https://www.asus.com/motherboards-compon...0m-pro-ii/
So, which one is it? Please also mention the BIOS version instead of just saying “latest.” We have several threads suggesting “latest” to keep everyone informed about pending updates.
Add the make and model of your case, along with its airflow and storage details. Event Viewer logs and .dmp files would be helpful if you’ve seen BSoD errors.
I’ve been troubleshooting this for a few months now. I’ve tested 14 different AMD drivers—23.11.1, 23.12.1, 24.5.2, 24.12.1, 25.10.2, 25.11.1, 25.12.1, 26.1.1.
Can you explain how you switched from one driver version to another?

B
Buns_of_Steel
Member
217
12-17-2021, 09:35 PM
#3
The situation with your case is quite unusual. When I encountered similar problems on my PC, I removed everything except a basic setup: mainboard, CPU, GPU, mouse, and keyboard. No extra components. I cleaned the system thoroughly, dusted the fans, and checked temperatures. As long as temperatures stayed normal until a crash occurred, it suggested the overheating wasn’t the issue. I replaced the thermal paste on the CPU and GPU if possible, and stopped worrying about temperature if they were within acceptable ranges. I gradually replaced cables one by one, testing after each change. If crashes persisted even with a new SATA cable, it pointed to a faulty cable rather than the hardware itself. I also considered whether the CPU had an integrated GPU and tested removing it. Running an overnight Mem86 test or at least several hours could help detect memory problems. What you’re doing is methodically narrowing down potential causes—removing heat-related issues, eliminating hardware failures, and checking cables one at a time. You’ve successfully identified the root cause: a faulty 2 dollar front panel power cable with a defective switch.
B
Buns_of_Steel
12-17-2021, 09:35 PM #3

The situation with your case is quite unusual. When I encountered similar problems on my PC, I removed everything except a basic setup: mainboard, CPU, GPU, mouse, and keyboard. No extra components. I cleaned the system thoroughly, dusted the fans, and checked temperatures. As long as temperatures stayed normal until a crash occurred, it suggested the overheating wasn’t the issue. I replaced the thermal paste on the CPU and GPU if possible, and stopped worrying about temperature if they were within acceptable ranges. I gradually replaced cables one by one, testing after each change. If crashes persisted even with a new SATA cable, it pointed to a faulty cable rather than the hardware itself. I also considered whether the CPU had an integrated GPU and tested removing it. Running an overnight Mem86 test or at least several hours could help detect memory problems. What you’re doing is methodically narrowing down potential causes—removing heat-related issues, eliminating hardware failures, and checking cables one at a time. You’ve successfully identified the root cause: a faulty 2 dollar front panel power cable with a defective switch.

C
Candy_737
Senior Member
254
12-18-2021, 04:40 AM
#4
Thank you for your response and for taking the time to review this.
The PSU I’m using is the Connect 80 Plus Bronze 650W from UltraPC. I understand it’s not a widely recognized or trusted model, so I chose it because my budget was limited when assembling the system. I plan to swap it for a 750W bronze Cooler Master MWE (https://www.ultrapc.ma/entre-700-w-et-79...0w-v3.html) if needed, but I was unsure whether the issue lies with my GPU or the quality of the PSU. Would that be a suitable alternative?

For the motherboard, please accept my apologies for any confusion — it is an ASRock B450M Pro R2.0, not an ASUS model. The BIOS version I’m running is 10.20, and all chipset drivers are up to date. My system runs the Hybrok Shadow ARGB (https://www.hybrok.com/?product=hybrok-rs6-argb-2) with three intake fans and one exhaust fan. The airflow appears adequate, and temperatures remain within normal limits even before any crashes occur.

In terms of storage, I’m using a 256GB SSD (model: micron 1100 mtfddav256tbn) paired with a 1TB hard drive. After reviewing Event Viewer, I noticed several errors around the moments when the system crashed.

First Error
A critical hardware error occurred. Reported by component: Processor CoreError Source: Machine Check Exception
Error Type: Cache Hierarchy Error
Processor APIC ID: 4

Second Error
The AMDRyzenMasterDriverV31 service failed to start because the system couldn’t locate the specified file.

Third Error
The AMDRyzenMasterDriverV22 service also failed to start, with a similar error message.

Fourth Error
The system restarted without a clean shutdown. This could result from the system failing to respond, crashing, or losing power unexpectedly.

Regarding drivers, I used DDU in Safe Mode during driver updates and performed clean installs each time. I also carried out a full Windows reinstall. However, I haven’t upgraded to newer drivers; instead, I always perform a complete clean install.

Edit
Thank you for your reply.
I’ve been keeping an eye on temperatures for more than a week now, and everything looks normal during gaming or when running stress tests like FurMark.
Unfortunately, my CPU lacks an integrated GPU (Ryzen 5 3600), so I can’t verify performance without the GPU. I’ll proceed with a longer MemTest86 run and inspect/replace any suspicious cables if possible.
Appreciate your support!
I’ll clean out my PC and test each component individually.
C
Candy_737
12-18-2021, 04:40 AM #4

Thank you for your response and for taking the time to review this.
The PSU I’m using is the Connect 80 Plus Bronze 650W from UltraPC. I understand it’s not a widely recognized or trusted model, so I chose it because my budget was limited when assembling the system. I plan to swap it for a 750W bronze Cooler Master MWE (https://www.ultrapc.ma/entre-700-w-et-79...0w-v3.html) if needed, but I was unsure whether the issue lies with my GPU or the quality of the PSU. Would that be a suitable alternative?

For the motherboard, please accept my apologies for any confusion — it is an ASRock B450M Pro R2.0, not an ASUS model. The BIOS version I’m running is 10.20, and all chipset drivers are up to date. My system runs the Hybrok Shadow ARGB (https://www.hybrok.com/?product=hybrok-rs6-argb-2) with three intake fans and one exhaust fan. The airflow appears adequate, and temperatures remain within normal limits even before any crashes occur.

In terms of storage, I’m using a 256GB SSD (model: micron 1100 mtfddav256tbn) paired with a 1TB hard drive. After reviewing Event Viewer, I noticed several errors around the moments when the system crashed.

First Error
A critical hardware error occurred. Reported by component: Processor CoreError Source: Machine Check Exception
Error Type: Cache Hierarchy Error
Processor APIC ID: 4

Second Error
The AMDRyzenMasterDriverV31 service failed to start because the system couldn’t locate the specified file.

Third Error
The AMDRyzenMasterDriverV22 service also failed to start, with a similar error message.

Fourth Error
The system restarted without a clean shutdown. This could result from the system failing to respond, crashing, or losing power unexpectedly.

Regarding drivers, I used DDU in Safe Mode during driver updates and performed clean installs each time. I also carried out a full Windows reinstall. However, I haven’t upgraded to newer drivers; instead, I always perform a complete clean install.

Edit
Thank you for your reply.
I’ve been keeping an eye on temperatures for more than a week now, and everything looks normal during gaming or when running stress tests like FurMark.
Unfortunately, my CPU lacks an integrated GPU (Ryzen 5 3600), so I can’t verify performance without the GPU. I’ll proceed with a longer MemTest86 run and inspect/replace any suspicious cables if possible.
Appreciate your support!
I’ll clean out my PC and test each component individually.

W
wolfpup118
Member
229
12-18-2021, 06:21 AM
#5
My case is named Hybrok Shadow ARGB. It might be useful to add more fans to the case, but it’s clear that the front fascia has limited ventilation and mainly cools the rest of the components. You can test this by removing the front fascia—your temperatures should drop. Regarding the PSU, when I mentioned a source, I meant borrowing rather than purchasing. If you bought a replacement part for troubleshooting, you’d end up with two identical parts, but the second set would be brand new and your expenses would be reduced. Similarly, you could run your GPU on a donor system with a stable PSU that provides ample power to check if the problem continues.
W
wolfpup118
12-18-2021, 06:21 AM #5

My case is named Hybrok Shadow ARGB. It might be useful to add more fans to the case, but it’s clear that the front fascia has limited ventilation and mainly cools the rest of the components. You can test this by removing the front fascia—your temperatures should drop. Regarding the PSU, when I mentioned a source, I meant borrowing rather than purchasing. If you bought a replacement part for troubleshooting, you’d end up with two identical parts, but the second set would be brand new and your expenses would be reduced. Similarly, you could run your GPU on a donor system with a stable PSU that provides ample power to check if the problem continues.