PC losing display signal (DP & HDMI) and rebooting to no display signal

[GamerMoment]

THREADS MERGED, PLEASE CLICK HERE TO BE LINKED TO MY MOST RECENT UPDATE ON THIS ISSUE

[191x7]

Are you running the latest chipset drivers from AMD?

 

Have you tried disabling MultiPlane Overlay MPO in Windows?

[GamerMoment]

26 minutes ago, 191x7 said:

Are you running the latest chipset drivers from AMD?

 

Have you tried disabling MultiPlane Overlay MPO in Windows?

I just updated the chipset drivers but not holding my breath.

 

I haven't tried that. Could that really be the problem? 

[GamerMoment]

Okay, so I have posted about this before and the issue seemed to subside for awhile, but reared its head again a few months back.

 

The issue: 

PC loses display signal (both DisplayPort and HDMI) and the monitors (ROG Strix 1440p ultrawide and TUF Gaming 1440p) go black. The PC will either reboot or stay locked up like that with black screens indefinitely until I perform a hard shutdown. When this happens with audio playing, the audio clips/cuts out which makes me think it isn't a GPU issue but more likely a CPU or RAM issue. 

• This happens both under gaming load and with no applications open after bootup and login.
• Stress-testing with Unigine with the GPU pulling 440+ watts from the PSU doesn't seem to trigger this issue. Demanding gaming loads don't seem to be a component of this issue at all and the PC performs to spec when this issue is not occurring.

PC Specs:

• Corsair HX750i 750w 80+ platinum PSU
• Asus B450 TUF Gaming motherboard
• AMD Ryzen 7 5800X3D
• Asus GeForce RTX 3080 12GB ROG Strix OC
• 4*8GB G.Skill Trident Z Neo 3200Mhz CL16
• EKWB 360mm AIO

BIOS version: 3802 (need to confirm but I remember updating BIOS before installing 5800X3D).

Windows version: Windows 11 64-bit

 

Steps I have taken:

• Roll back Nvidia drivers.
• Clean uninstall and reinstall Nvidia drivers.
• Reseat RAM (this appeared to resolve issue, but issue occurred 5-6 times in a row yesterday)
• Checked connections at PSU. 
• Tried all three monitors and used HDMI and DisplayPort.
• Even tried sitting in the BIOS screen to see if it occurs within the BIOS. 

My thoughts: The PSU is 1.5 years old and was purchased new. I don't believe my system is underpowered as the 5800X3D is very power efficient. Yes the 3080 is power-hungry, but I don't seem to have issues under load. I don't think power draw from components is an issue but would I rule out the PSU dying? No, but the inconsistency of the issue I'm having points to it being something else. My G.Skill RAM's lighting is extremely erratic and sometimes I think there's something to that. Perhaps the DIMMs are faulty or maybe the G.Skill app that controls the lighting is causing an issue (I will uninstall it for a week and see what happens). My CPU performs to spec and I have no temperature issues (and that's on pre-applied EKWB thermal paste). The GPU was purchased in October 2022. Could it be faulty? Of course, I just think if it was then the issue would be more consistent. I thought maybe it was GPU sag, but I doubt that. I have been considering getting the 7900 XTX so maybe I should do that and see if a different GPU solves the problem and then I can always RMA the 3080...

 

Sorry for posting about the same thing twice. Hope to hear from you guys. 

[emosun]

14 minutes ago, GamerMoment said:

This happens both under gaming load and with no applications open after bootup and login.

it's the ram

run it with the single ram stick alone in the machine and see if its stable in that configuration

15 minutes ago, GamerMoment said:

4*8GB G.Skill Trident Z Neo 3200Mhz CL16

the fact that you have four of these makes it pretty likely one or a few are bad , or the board just hates them

[GamerMoment]

2 minutes ago, emosun said:

it's the ram

run it with the single ram stick alone in the machine and see if its stable in that configuration

the fact that you have four of these makes it pretty likely one or a few are bad , or the board just hates them

This makes too much sense. So it could be a single stick tripping my system up? I did my research so as to buy memory that was compatible with my board and CPU. Should I figure out which one is bad and replace it? Buy a whole new kit? 

[emosun]

it's either a bad stick , bad board , or the board just doesn't like those specific sticks in that configuration and will work with something else.

in any case i'd just go with two 16gb sticks so you have less of them instead. but check and make sure it's stable with 1 and then 2

[GamerMoment]

2 minutes ago, emosun said:

it's either a bad stick , bad board , or the board just doesn't like those specific sticks in that configuration and will work with something else.

in any case i'd just go with two 16gb sticks so you have less of them instead. but check and make sure it's stable with 1 and then 2

The thing that puzzles me is, I've had those memory modules in for like two years. I installed them in August 2021 or thereabouts - around the time I had a 3070 in my machine, and I never had problems like this. I'd prefer to have all DIMM slots filled for the aesthetic - is it always this hard to get 4 DIMMs to work nicely with a given board? Maybe my b450 is looking a bit long in the tooth...

[Alex Atkin UK]

1 minute ago, GamerMoment said:

The thing that puzzles me is, I've had those memory modules in for like two years. I installed them in August 2021 or thereabouts - around the time I had a 3070 in my machine, and I never had problems like this. I'd prefer to have all DIMM slots filled for the aesthetic - is it always this hard to get 4 DIMMs to work nicely with a given board? Maybe my b450 is looking a bit long in the tooth...

Its harder to drive 4 sticks than 2 so yeah, there's always a higher chance of problems.

 

Plus as mentioned it could just be one stick developed a fault.

[GamerMoment]

3 minutes ago, Alex Atkin UK said:

Its harder to drive 4 sticks than 2 so yeah, there's always a higher chance of problems.

 

Plus as mentioned it could just be one stick developed a fault.

Thank you both for your informative answers. I just had a quick search and it seems as though my board supports up to 4400Mhz memory. Do you think this set of two 16GB modules would be suitable? 

[emosun]

41 minutes ago, GamerMoment said:

Thank you both for your informative answers. I just had a quick search and it seems as though my board supports up to 4400Mhz memory. Do you think this set of two 16GB modules would be suitable? 

should be ok provided you checked and confirmed the issue first before deciding to buy hardware

[GamerMoment]

22 minutes ago, emosun said:

should be ok provided you checked and confirmed the issue first before deciding to buy hardware

DIMM is pretty inexpensive these days so ordered a kit anyway. I will try using 2 out of 4 of my current sticks later today and see how I get on. Will report back and confirm solution but that might take weeks since sometimes I can go weeks without it happening lol. 

[GamerMoment]

Okay so here's another one from me after this thread.

 

TL;DR: PC kept losing display signal and rebooting, all replies suggested memory issue. By the end of it, had a couple of solutions.

 

So just for context, my original kit of memory was 4x8GB G.Skill TridentZ Neo 3200MHz CL16. According to some, 4 sticks is often more taxing on the memory controller and board than just two (correct me if and where I am wrong). So I ordered a kit of 2x16 GB G.Skill TridentZ Neo 3600Mhz CL16. This was an oversight as a cursory Google seemed to suggest that the maximum memory speed for B450 boards is greater than 3200MHz, but I think my board in particular is limited to 3200MHz w/ D.O.C.P. profile enabled. I assume it's fine to use the 3600MHz sticks set to 3200MHz. 

 

So I installed the new memory kit, booted my PC up, entered the BIOS, and set it to 3600Mhz. After looking in, things seemed okay, so I ran with Windows Memory Diagnostic test. Unsurprisingly, the test detected memory errors. After logging in on the second restart, the system encountered a BSOD. I don't remember the error code (sorry, I know that's not helpful, but I think it's safe to assume what the cause of the BSOD was), but I restarted and set the memory to the standard D.O.C.P. profile (3200MHz) in the BIOS and booted up. 3200MHz was stable although gaming performance seemed worse. After foolishly doing a bit more tinkering trying to find a stable speed that was higher than 3200MHz. I also decided to enable Resizable BAR, but when I saved and exited, my PC encountered a seemingly fatal issue. Upon startup, the system would enter recovery mode and attempt to repair the installation which was never successful. When the recovery mode failed to repair the installation, the system would BSOD with the BAD_SYSTEM_CONFIG_INFO error. I could still enter the BIOS. I ended up doing a system restore to a restore point from a week ago and everything seems fine. I rebooted, and entered the bios to re-enable the D.O.C.P. (aka XMP profile). 

 

PC Specs:

• Corsair HX750i 750w 80+ platinum PSU
• Asus B450 TUF Gaming Plus motherboard
• AMD Ryzen 7 5800X3D
• Asus GeForce RTX 3080 12GB ROG Strix OC
• 2x16GB G.Skill Trident Z Neo 3600Mhz CL16
• EKWB 360mm AIO

BIOS version: 3802.

Windows version: Windows 11 64-bit

 

So what caused the BAD_SYSTEM_CONFIG_INFO error? Did I not let the system get used to new timings? I tried a variety of memory speeds between 3200 and 3600. 3333MHz appeared to actually be stable, and I increased it without running another Windows Diagnostic Memory test. 

What I still need to do is install the previous 4x8GB kit and run the memory test to confirm that there were memory errors that were perhaps causing my issue as outlined in my previous thread. Yes I am a huge fool for not doing this off

I will not be increasing the memory speed of the 3600 kit past 3200MHz but I was curious to know if re-enabling Resizable BAR caused an issue. I know that you have to disable CSM when enabling it. 

Hope I haven't forgotten anything - the whole ordeal was traumatising lol. 

[GamerMoment]

I don't know why 3600MHz wouldn't work if this is definitely the same board as mine

[GamerMoment]

On 5/8/2023 at 10:50 AM, emosun said:

it's either a bad stick , bad board , or the board just doesn't like those specific sticks in that configuration and will work with something else.

in any case i'd just go with two 16gb sticks so you have less of them instead. but check and make sure it's stable with 1 and then 2

 

Okay so update: 

I installed the 2x16GB kit and the issue seemed to stop for a few days...then it started again. I ran memory diagnostic tests and thought that 3200MHz with the DOCP (XMP) profile was stable. The original 4x8GB kit predates this issue, but the issue does line up with when I switched to Windows 11. Could it be an OS issue and not a hardware issue? 

[emosun]

On 5/7/2023 at 5:50 PM, emosun said:

it's either a bad stick , bad board , or the board just doesn't like those specific sticks

well again this is why you test the single stick first before buying hardware.

 

try the current config with a single stick

[GamerMoment]

4 hours ago, emosun said:

well again this is why you test the single stick first before buying hardware.

 

try the current config with a single stick

Thanks I will. I'll also try using the on-board graphics output for a few days/week and see what happens. 

[GamerMoment]

Ok so here's the last thread I posted about this issue.

 

I'll try to recap everything shortly and sharply without being vague:

 

Issue: PC will lose display signal (doesn't matter which of my three monitors I use or whether I'm using HDMI or DisplayPort and it doesn't matter which port I use) and lock up at which time it will either stay locked up indefinitely or reboot. Sometimes the other monitor will keep its picture but it will be frozen and pixelated.

PC Specs:

• Corsair HX750i 750w 80+ platinum PSU (purchased and installed in 2021)
• Asus B450 TUF Gaming motherboard (came with prebuild in 2020)
• AMD Ryzen 7 5800X3D (purchased and installed in late 2022)
• Asus GeForce RTX 3080 12GB ROG Strix OC (purchased and installed in late 2022)
• 2x16GB G.Skill TridentZ Neo CL16 3600Mhz (set to XMP @ 3200MHz as I believe this is the maximum rated / stable memory speed for my motherboard)
• EKWB 360mm AIO (purchased and installed in late 2022)

BIOS version: 3802 (need to confirm but I remember updating BIOS before installing 5800X3D).

Windows version: Windows 11 64-bit

 

Measures taken: 

• As I said, HDMI and DisplayPort cables as well as ports and monitors don't make a difference.
• Replaced previous 4x8GB memory kit (same G.Skill kit except max. speed of 3200MHz) and performed diagnostic memory tests. Errors detected when memory set @ 3600MHz.

 

This issue does not occur while under intensive gaming loads, so for that reason I find it difficult to believe that the power supply is the culprit. The GPU and the PSU seem to operate to spec and as intended even under duress. Furthermore, I want to rule out DIMM modules as well given that a completely different and fresh memory kit has not solved the issue. That leaves the CPU and the motherboard, but I am leaning toward the motherboard.

 

I do have a bit of a crackpot theory that it's the 2000rpm Noctua industrial strength fans I have in the front, but that could just be me searching everywhere for an answer. Could their power draw be spiking and tripping the power limit of the PSU? I really don't think so. I think the motherboard is the most likely culprit here, but here's my dilemma:

 

Starting with replacing the PSU is a cheaper and simpler place to start, and I do acknowledge that a faulty of failing power supply can be a ticking timebomb and it can take a whole system with it when it dies. But I don't want to replace it only for the issue to persist after the fact. Same goes for the other option which is replacing the motherboard (and maybe throw in a cheeky M.2 NVMe SSD) which will involve purchasing a physical Windows 11 USB for a clean install. 

 

This issue has been ongoing for 6-7 months now so forgive me if I've left anything out (you can always refer to my previous threads about this).

 

Thanks for everyone's help so far. 

[Whatisthis]

Go to event viewer and check the windows logs -> system section. Filter for errors and critical only. Post a screenie here.

[GamerMoment]

2 minutes ago, Whatisthis said:

Go to event viewer and check the windows logs -> system section. Filter for errors and critical only. Post a screenie here.

I forgot to say, I had a go at this and couldn't make heads nor tails of it but I will definitely definitely post a screenshot later on. Hopefuly something sticks out to someone.

[GamerMoment]

1 hour ago, Whatisthis said:

Go to event viewer and check the windows logs -> system section. Filter for errors and critical only. Post a screenie here.

Now that I think about it...

 

Could the 24-pin ATX cable have a loose connection? I think I will check this when I get home from work. 

[GamerMoment]

On 6/7/2023 at 10:56 AM, Whatisthis said:

Go to event viewer and check the windows logs -> system section. Filter for errors and critical only. Post a screenie here.

reporting back.

 

Performed clean windows reinstallation using the 'Cloud Reset' option at 3am on Saturday. PC remained powered on without issue until 2am the next morning when I shut it down. The PC has performed as intended for a whole weekend of Diablo IV, but it does have a habit (the 'problem', I mean) of working for weeks or even months and then rearing its head. My PC parts retailer only has one Aorus X470S Master left so hopefully I didn't need it after all (although the upgrade would be nice).

[GamerMoment]

Apologies in advance if people would prefer I bump existing threads, but I just feel as though the whole saga is evolving in such a way that new threads are warranted. If you aren't familiar with the story so far: read here

 

So there's a bit of copy-pasting here:

 

Issue: PC will lose display signal (doesn't matter which of my three monitors I use or whether I'm using HDMI or DisplayPort and it doesn't matter which port I use) and lock up at which time it will either stay locked up indefinitely or reboot. Sometimes the other monitor will keep its picture but it will be frozen and pixelated and the sound will become distorted as the PC decides wtf it's going to do lol. 

PC Specs:

• Corsair HX750i 750w 80+ platinum PSU (purchased and installed in 2021)
• Asus TUF B450 Gaming motherboard (came with prebuild in 2020)
• AMD Ryzen 7 5800X3D (purchased and installed in late 2022)
• Asus GeForce RTX 3080 12GB ROG Strix OC (purchased and installed in late 2022)
• 2x16GB G.Skill TridentZ Neo CL16 3600Mhz (set to XMP @ 3200MHz as I believe this is the maximum rated / stable memory speed for my motherboard)
• EKWB 360mm AIO (purchased and installed in late 2022)

BIOS version: 3802

Windows version: Windows 11 64-bit

 

Measures taken: 

• Rolling back drivers, updating drivers, wipe and clean install with DDU.
• As I said, HDMI and DisplayPort cables as well as ports and monitors don't make a difference.
• Replaced previous 4x8GB memory kit (same G.Skill kit except max. speed of 3200MHz) and performed diagnostic memory tests. Errors detected when memory set @ 3600MHz.
• Fresh Windows reinstallation

 

This issue does not occur while under intensive gaming loads, so for that reason I find it difficult to believe that the power supply is the culprit. The GPU and the PSU seem to operate to spec and as intended even under duress. Furthermore, I want to rule out DIMM modules as well given that a completely different and fresh memory kit has not solved the issue. That leaves the CPU and the motherboard, but I am leaning toward the motherboard.

 

UPDATED BIT! 

The last measure I took to resolve this issue was resetting my PC - and I do have questions about this. I opted to reset my PC and when given the option of a 'Cloud Reset' or a 'Local Reinstallation', I chose cloud reset. I'm still not quite sure which would have been more appropriate, so if someone knows, then please let me know. 

 

Next Steps:

 

Currently waiting on the following:

• Gigabyte Aorus X570S Master replacement motherboard
• Samsung 990 Pro NVMe M.2 SSD
• Windows 11 home physical USB key
• Noctua thermal paste
• Noctua thermal paste removal wipes (I read good things!)

 

I will replace motherboard and repaste CPU (temps weren't an issue). 

 

If I still have issues I will ring my local and ask if they can test my GPU. If the GPU passes their test, I will replace the PSU. 

 

 

 

[SansVarnic]

-= Topics Merged =-

OP, reposting the same topic is against the rules, please do not do this just update your topic.

[GamerMoment]

7 minutes ago, SansVarnic said:

-= Topics Merged =-

OP, reposting the same topic is against the rules, please do not do this just update your topic.

Thanks and apologies - I should have re-read the rules!