Question PC randomly reboots while gaming, but not during stress test ?

[HonaZak]

Hi.
I built my PC a year ago, and have been experiencing this problem ever since. After a while in game, sudden black screen appears, and reboots into windows afterwards. The only clues I’ve found are a critical error (41 - Kernel-Power) in the Event Viewer and a critical hardware error (193 - LiveKernelEvent) in the Reliability History.

Here are my specs:
MB: MSI PRO B650-s wifi
CPU: AMD Ryzen 5 7600X
GPU: NVIDIA GeForce RTX 4070
RAM: Corsair 32GB KIT DDR5 5600MT/s CL40 Vengeance RGB Grey EXPO
Cooler: Endorfy Fera 5
Storage: Apacer AS2280Q4X 2TB
PSU: GIGABYTE P750GM

Just a few additions:
PC has been to 2 different repair shops, both times returned with no reboots happening.
I have already replaced PSU. I measured a peak CPU temperature of 96°C, but before the reboot, temps are usually around 85°C, so it doesn’t seem like an overheating issue.

I would appreciate any tips or suggestions.

NOTE: this is my second thread on this topic, becuase the first one has died already. Hope thats okay.

 

[John Chesterfield]

Possibly due to transient power spikes.

But that power supply you have is disastrous, either way you should get rid of that and try using a good quality unit and see if that fixes the issue.

Edit: So you say you've replaced the power supply, you need to tell us what model both the old and new one are. If that Gigabyte unit is the new one, then you might have replaced trash with trash.

 

[HonaZak]

I had a SilverStone Strider Essential 80Plus ST60F-ES230 600W before. I accually got the gigabyte one recomended from a repair shop. Even if they are both trash, it seems very unlikely to me that they are both causing the exactly same issue. Is there any way to confirm that power spikes are the cause?

 

[Aeacus]

I accually got the gigabyte one recomended from a repair shop.

Well yeah, since no-one in the right mind is going to buy that junk, so, the shop wants to get rid of the dead weight. Hence why suggesting junk to unsuspecting customers, since they don't know better.

Keep in mind, that for repair joints or retail stores, for them, it isn't about giving you a good product, but instead to make a profit. Especially when they can get rid of the junk that no-one else is buying.

SilverStone Strider Essential 80Plus ST60F-ES230 600W

Crap quality PSU. Tier E.

PSU: GIGABYTE P750GM

Exploding junk. Tier F.

PSU tier list: https://docs.google.com/spreadsheet...JWkc/edit?pli=1&gid=1973454078#gid=1973454078

For new PSU, look any unit in Tier A. Tier A- will do too, while Tier A+ is preferred.

Even if they are both trash, it seems very unlikely to me that they are both causing the exactly same issue.

Since both are crap, both can and most likely, will produce the same result. Only when you've ruled out PSU issue, with known to be good/great quality unit (Tier A), and when issue still appears, then you can look towards other hardware, like GPU.

But do keep in mind that crap quality PSU can damage all hardware it is connected to. Not only outright kill it, but also gradually damage it, by feeding out of spec voltages or producing high ripple. It is common for crap/low quality PSU to damage e.g GPU to such an extent, where when even switching to good/great quality PSU, the issue remains, since hardware has sustained damage. And new PSU will not magically repair the damage caused by crap/low quality PSU.

So, 1st order of business: new good/great quality PSU.
2nd, CPU/GPU temps.

I measured a peak CPU temperature of 96°C, but before the reboot, temps are usually around 85°C, so it doesn’t seem like an overheating issue.

85C gaming temps for that chip? Way too high. For proper cooling, gaming temps should be ~70C,
review: https://www.techpowerup.com/review/amd-ryzen-5-7600x/25.html

So, besides PSU issue, you also have thermal issue as well.
What kind of cooling (if any) you have in there?

 

[HonaZak]

So, besides PSU issue, you also have thermal issue as well.
What kind of cooling (if any) you have in there?

I'm using enforfy fera 5. It definitly is a cheap piece compared to the rest of the build, however I am seeing very differing opinions regarding the ideal temp. I believe I once saw a saying from amd, that ryzen 7600 are designed to run at 90C+.

 

[Aeacus]

however I am seeing very differing opinions regarding the ideal temp. I believe I once saw a saying from amd, that ryzen 7600 are designed to run at 90C+.

Well, your CPU Tjmax is 95C,
specs: https://www.amd.com/en/products/processors/desktops/ryzen/7000-series/amd-ryzen-5-7600x.html

But running at it's max temp doesn't mean it is good for the CPU.
When temp goes way too high, you'd get CPU (or GPU) thermal throttle, where frequency (and performance) is lowered, just to keep the hardware to cook itself up. But if that doesn't help either, system either reboots or automatically shuts down. Else-ways, hardware will be cooked to crisp.

Btw, case fans: amount, location, orientation, size, RPM range, CFM, mmH2O is? If you even have any case fans that is.

 

[HonaZak]

My case got 3 140mm fans, 1000rpm, all facing the right direction.

 

[Aeacus]

all facing the right direction.

Right direction? All three at same direction? If so, you either have extreme positive or extreme negative pressure. Neither of them is good for cooling.

What is the fan placement?

For example, in my main build (Skylake, full specs with pics in my sig), i have:
Front - 2x 140mm Corsair ML140 Pro red LED fans as intake
Bottom - 1x 120mm Corsair ML120 Pro red LED fan as intake
Top - 3x 140mm NZXT AER140 RGB fans as exhaust
Rear - 1x 140mm Corsair ML140 Pro red LED fan as exhaust

Besides that, i have mid-tower air cooler which has 2x 120mm Corsair ML120 Pro red LED fans and those can be considered as exhaust as well.

My fans specs wise;
ML140 Pro: Airflow: 20-97 CFM, static pressure: 0.2-3.0 mmH2O, RPM range: 400-2000 RPM.
ML120 Pro: Airflow: 12-75 CFM, static pressure: 0.2-4.2 mmH2O, RPM range: 400-2400 RPM.
NZXT AER140 RGB: Airflow: 91.19 CFM, static pressure: 1.52 mmH2O, RPM range: 500-1500 RPM.

So, good fans with great airflow and high static pressure (except NZXT fans, which have mediocre static pressure). And with 7x case fans and their orientation, i have a slight negative pressure running in my system for optimal cooling.

My case got 3 140mm fans, 1000rpm

Put all your case fans at full tilt (100%) and look if CPU/GPU thermals do improve. Sure, it may be noisy but this is to see if you can get temps down with more airflow.

If you can, then you may need more fans, or if PC case only supports 3x fans (which make it one poor PC case), then much better performing fans, both in terms of airflow and static pressure.

 

[HonaZak]

I took off the side panel, added one lil fan inside and one big outside, blowing in the direrction of my airflow. Before the crash, I meassured max 83C on CPU, 70C on GPU, 43C on RAM, and 60C on SSD. I felt like it has something to do with temps, but now it seems totally random. It does it in such a wierd way. Today it crashed 5-10 min into the game 5 times in a row. The same game sometimes takes an hour, and someday doesn't crash at all.

 

[Aeacus]

High ripple from PSU causes system instability. So, unless you rule out PSU issue, PSU would remain as one of the main suspects.

Btw, when was the last time you cleaned PC internals from dust?

 

[HonaZak]

High ripple from PSU causes system instability. So, unless you rule out PSU issue, PSU would remain as one of the main suspects.

Btw, when was the last time you cleaned PC internals from dust?

About 2 months ago. But becasue I am experiencing reboots since the day I built this pc, dust probably doesn't change much.

 

[Aeacus]

Dust is to do with temps.

But since you had the issue from day 1, PSU is still a suspect.

You may get it more stable by removing RAM EXPO and running JEDEC speeds. Maybe even downclock the GPU a bit. But if it is ripple issue, only better PSU helps.

 

[HonaZak]

I just noticed an orange flashing light on mb when the pc crashes, does that change anything? Running ram at default speeds doesnt help.

 

[Aeacus]

I just noticed an orange flashing light on mb when the pc crashes

Where is that LED exactly located? Is it one of the debug LEDs? Or is it at the bottom left of the MoBo, where the audio section is?

 

[HonaZak]

Where is that LED exactly located? Is it one of the debug LEDs? Or is it at the bottom left of the MoBo, where the audio section is?

Debug LED signaling DRAM fail.

 

[Aeacus]

When PC crashes, all the protocols stop and stop responding. This can manifest in all sorts of symptoms and telemetry at the point of crash may output wrong data.

Yellow debug LED = DRAM not detected or fail.
But the LED should remain lit, not blink. Meaning that during the system freeze, system (CPU) shortly looses connection to the DRAM and debug LED blinks.

How far out the new PSU is? Also, which make/model PSU you are going 3rd time around?

R5 7600X has iGPU in it. Take your RTX 4070 out of the system, connect monitor to MoBo and look if you still get system crashes.
For one, it puts far lesser load on PSU and for two, it can show if the issue would be with GPU as well. You can game on iGPU, but don't expect the same performance as with RTX 4070 (e.g 120 FPS on 1440p, more like 30 FPS on 720p).

 

[HonaZak]

But the LED should remain lit, not blink.

My bad, I said it was flashing. It stays lit after the crash until it starts to boot again.

How far out the new PSU is? Also, which make/model PSU you are going 3rd time around?

I will be borrowing a psu in few days. Not sure which model it will be.

Do you think it is worth trying to measure the ripple using oscilloscope?

 

[geofelt]

Gaming puts added load on the psu to drive the graphics card.
Occasional power spikes can do you in.
There have been issues with your psu:
https://www.tomshardware.com/news/gigabyte-full-refund-product-exchange-explosive-psus

GIGABYTE GP-P850GM & GP-P750GM POWER SUPPLY STATEMENT | News - GIGABYTE Global

GIGABYTE is aware of certain media outfits casting doubt over the quality of Power Supply models GP-P850GM and GP-P750GM. GIGABYTE takes pride in the design and quality of our ...

www.gigabyte.com

Borrow a known good psu of sufficient power and quality to test with.

 

[Roland Of Gilead]

Even if they are both trash, it seems very unlikely to me that they are both causing the exactly same issue. Is there any way to confirm that power spikes are the cause?

This is incorrect. This is exactly what would happen with a bad PSU.

I'd second @Aeacus on the P750GM. They actually explode. Do not use it further.

You should be aiming for a Corsair RMx/Rmi, Seasonic Prime TX/GX, Leadex III. There are many good PSU's.

 

[Aeacus]

Do you think it is worth trying to measure the ripple using oscilloskop?

Well, to measure ripple, you need load unit for the PSU, so you can put different kinds of loads on the PSU and then measure the ripple on all rails.

E.g here's TH's PSU testing equipment;
article: https://www.tomshardware.com/reviews/how-we-test-psu,4042-2.html

Easy 5K cost in total.
But there is no need to further test the already confirmed exploding PSUs.

View: https://www.youtube.com/watch?v=aACtT_rzToI


Other videos in the saga: https://www.youtube.com/playlist?list=PLsuVSmND84QuX_6U7vndGfGrMewAHbkdn

 

[HonaZak]

This is incorrect. This is exactly what would happen with a bad PSU.

I'd second @Aeacus on the P750GM. They actually explode. Do not use it further.

You should be aiming for a Corsair RMx/Rmi, Seasonic Prime TX/GX, Leadex III. There are many good PSU's.

Only PSUs within specific serial number range are exploding. I am not affected by that.

 

[HonaZak]

Well, to measure ripple, you need load unit for the PSU, so you can put different kinds of loads on the PSU and then measure the ripple on all rails.

E.g here's TH's PSU testing equipment;
article: https://www.tomshardware.com/reviews/how-we-test-psu,4042-2.html

Easy 5K cost in total.
But there is no need to further test the already confirmed exploding PSUs.

View: https://www.youtube.com/watch?v=aACtT_rzToI


Other videos in the saga: https://www.youtube.com/playlist?list=PLsuVSmND84QuX_6U7vndGfGrMewAHbkdn

Only PSUs within specific serial number range are exploding. I am not affected by that.

I will be trying new psu soon, so no need to test it.

 

[HonaZak]

R5 7600X has iGPU in it. Take your RTX 4070 out of the system, connect monitor to MoBo and look if you still get system crashes.
For one, it puts far lesser load on PSU and for two, it can show if the issue would be with GPU as well. You can game on iGPU, but don't expect the same performance as with RTX 4070 (e.g 120 FPS on 1440p, more like 30 FPS on 720p).

Using only integrated gpu removed the crashes. Even though it wasnt under as heavy load as before (judging based of cpu temps), it didn't crash for 2 hours on game that it crashed after 5 mins before, with same graphic settings.

 

[Aeacus]

Even though it wasnt under as heavy load as before (judging based of cpu temps), it didn't crash for 2 hours on game that it crashed after 5 mins before, with same graphic settings.

With this, it is either of the two:
1. PSU can not sustain heavier load with dedicated GPU during gaming.
2. GPU itself (RTX 4070) has an issue to keep itself stable during gaming.

Do note that GPU instability can be caused by the PSU as well.

To fully rule out GPU issue, you need 2nd, known to work, dedicated GPU. Preferably same TDP, so that same load is put on PSU.
If it also crashes even with 2nd (same TDP) GPU, then it is confirmed PSU issue.

Trying with 2nd, known to work, good/great quality PSU and with RTX 4070, would show if it is also PSU issue (if no more crashes anymore), or GPU issue (if crashes doesn't go away).

 

[HonaZak]

I bough a 5060ti, and an ADATA XPG CORE REACTOR II VE(A tier PSU). They both didn't change anything.