[SOLVED] GPU Not receiving enough voltage

[Purp1e Haze]

So recently I've been crashing during games such as valorant, tarkov etc... and im constantly checking on my temps and voltages nothing seems wrong.

https://imgur.com/a/6XQB6cC

View: https://imgur.com/a/6XQB6cC

My specs are a
Z97 Gaming 5 Msi Motherboard
Intel i5-4690k CPU
GTX 980 GPU
Rosewell Photon 850 PCU
Sandisk SSD
Corsair Vengence DDR3 RAM
N/A HDD

 

[vov4ik_il]

Pls elaborate on the "crashing".
I did not find anything abnormal in your screenshot, except for 0 rpm on GPU cooling.

 

[ChumP]

Well, that +12V reading looks out of the ATX spec .... +11.4 - +12.6V are the tolerances for the 12V rail.

TMPIN5 and 8 are also crazy high, but those maybe just false readings from the sensors.

The 0 RPM on the GPU is quite normal... modern GPUs usually operate in this mode when they're not doing anything heavy.

 

[vov4ik_il]

The 0 RPM on the GPU is quite normal... modern GPUs usually operate in this mode when they're not doing anything heavy.

Not when it is at 37% and already at 57C, I would expect the fan to kick in...
And yes, I missed the 11v, that's' way too low. What does it read when under the OCCT synthetic load?

 

[Purp1e Haze]

Pls elaborate on the "crashing".
I did not find anything abnormal in your screenshot, except for 0 rpm on GPU cooling.

In game I randomly get brown screen's and then my screen freezes on its desktop.

 

[Purp1e Haze]

Not when it is at 37% and already at 57C, I would expect the fan to kick in...
And yes, I missed the 11v, that's' way too low. What does it read when under the OCCT synthetic load?

https://imgur.com/a/OMV2tJr

View: https://imgur.com/a/OMV2tJr
This is under CINEBENCH stress test.
Yea the fan doesn't start kicking until a bit hotter but this isn't of any concern since my GPU is never reaching dangerous temps.

 

[Purp1e Haze]

Well, that +12V reading looks out of the ATX spec .... +11.4 - +12.6V are the tolerances for the 12V rail.

TMPIN5 and 8 are also crazy high, but those maybe just false readings from the sensors.

The 0 RPM on the GPU is quite normal... modern GPUs usually operate in this mode when they're not doing anything heavy.

Well, that +12V reading looks out of the ATX spec .... +11.4 - +12.6V are the tolerances for the 12V rail.

TMPIN5 and 8 are also crazy high, but those maybe just false readings from the sensors.

The 0 RPM on the GPU is quite normal... modern GPUs usually operate in this mode when they're not doing anything heavy.

https://imgur.com/a/g2ONT2x

View: https://imgur.com/a/g2ONT2x

 

[vov4ik_il]

Well, the BIOS picture shows tension under 0 load. Cinebench does not load the GPU and does not load the PSU sufficiently.
OCCT does them all. If PSU tension accuracy is in question, a better way would be getting a multimeter and measuring actual tensions on any of the unused Molex connectors during a load test.

P.S The 8-pin PCIe pinout is wrong, pin 4 is normally ground.

 

[Purp1e Haze]

Well, the BIOS picture shows tension under 0 load. Cinebench does not load the GPU and does not load the PSU sufficiently.
OCCT does them all. If PSU tension accuracy is in question, a better way would be getting a multimeter and measuring actual tensions on any of the unused Molex connectors during a load test.

P.S The 8-pin PCIe pinout is wrong, pin 4 is normally ground.

This seems like a lot of work, and I do load the GPU with heaven benchmark and everything seems fine. It has to be a hardware issue right either my motherboard or PSU.

 

[vov4ik_il]

This seems like a lot of work...

So you would prefer to blindly swap components and let your billfold work instead?
I would give another PSU a try, but it might be the motherboard (aging caps) or the GPU.

 

[Purp1e Haze]

So you would prefer to blindly swap components and let your billfold work instead?
I would give another PSU a try, but it might be the motherboard (aging caps) or the GPU.

I guess I’ll give the multimeter a try in the near future so I’ll be testing the voltage output’s while running OCCT and if everything seems clear than it should be down to the motherboard or GPU.

 

[vov4ik_il]

I guess I’ll give the multimeter a try

Just measure +12v on one of the unused Molex 4-pin or PCIe plugs (probes to yellow(+) and black(-) wires) and if it is indeed 11v, swap the PSU...

 

[vov4ik_il]

So I just tested it and I ended up getting 10v and 9v on the Molex 4 Pin

Apparently, time for a new PSU... Was it that hard work it seemed to you?

 

[Purp1e Haze]

Just measure +12v on one of the unused Molex 4-pin or PCIe plugs (probes to yellow(+) and black(-) wires) and if it is indeed 11v, swap the PSU...

So I just tested it and I ended up getting 11v on the Molex 4 Pin where it’s supposed to be 12v and for the 5v I got 4v Im not sure I can set the multimeter to measure to the decimal so what’s my next course of action.

 

[Purp1e Haze]

Apparently, time for a new PSU... Was it that hard work it seemed to you?

I was measuring in the wrong voltage settings my bad can you check the latest reply.

 

[Purp1e Haze]

Apparently, time for a new PSU... Was it that hard work it seemed to you?

I was measuring in the wrong voltage settings my bad can you check the latest reply.

So I just tested it and I ended up getting 11v on the Molex 4 Pin where it’s supposed to be 12v and for the 5v I got 4v Im not sure I can set the multimeter to measure to the decimal so what’s my next course of action.

I just searched up reviews on the multimeter I’m using and I found a review saying that it’s not a very accurate multimeter and it usually shows 1v lower than the original so I’m starting to doubt the PSU could be a problem since it’s recently new and a reliable one on the other hand my motherboard has been being used for the last 8 years.

 

[vov4ik_il]

So I just tested it and I ended up getting 11v on the Molex 4 Pin where it’s supposed to be 12v and for the 5v I got 4v I'm not sure I can set the multimeter to measure to the decimal so what’s my next course of action.

Your system is supposedly designed to work within specifications (+/- 5% voltage tolerance in case of PSU, if I recall correctly). For everything out of those specifications you are an experimenting pioneer and nothing is guaranteed.
Your next action would be getting another power supply that complies with the spec and see from there...

P.S.
I assume you are using a DMM for your measurements and the scale is the lowest scale that can accommodate 15V. While you are doing that the system should be under synthetic load (OCCT or whatever another tool of your choice that loads both CPU and GPU at the same time).
11.4V is still within the spec, but 11.3 is not.

 

[vov4ik_il]

I just searched up reviews on the multimeter I’m using and I found a review saying that it’s not a very accurate multimeter and it usually shows 1v lower

That is easy to check, get any known good battery (AA, the battery in your car, any), and see what does it read.