Jump to content
Search In
  • More options...
Find results that contain...
Find results in...
Lasne

Narrowing down my problem to the GPU or PSU

Recommended Posts

Posted · Original PosterOP

Hello,

 

Some Background:

In the last 1-2 weeks, my computer has been crashing in one of two ways. 1. the computer loses power and restarts itself right afterwards. there is no error messages, warning, or bsod. 2. The screen goes black, programs still operate in the background, but the display will not return unless the computer is restarted. These crashes always occur under load.

 

What I've done so far:

I have taken out the GPU and performed a prolonged CPU stress test with no issues. I have stress tested the graphics card with Unigine Heaven Benchmark and the computer crashes within 10 seconds of start the program. even on the basic preset. When ran after a cold start, it lasts several more minutes. I have not been able to induce a crash through Furmark yet.

 

From this I have 2 theories. My graphics card is overheating and/or failing and the computer is shutting down as a failsafe. The 2nd theory is that for whatever reason, the PSU is unable to supply enough power to the graphics card under load and the computer shuts down.

 

My Problem:

I have no idea how to go about ruling out the PSU as the cause of the crashing. Does anyone have suggestions for what I should do to verify that this issue stems from the GPU or the PSU?

 

System specs: (all components are at default clock speeds)

Motherboard: MSI Z97

CPU: Intel i7 4790k 4.0 GHz

GPU: Nvidia GeForce 980 Ti

PSU: corsair HX750i

let me know if anything else is necessary

 

Thanks in advance!

Link to post
Share on other sites

Check the psu output voltages for all the rails.


PC: CPU - 8700K 4.8ghz -1avx 1.224V  |  GPU - 5700XT 2100mhz 1.145V target  |  Motherboard - Aorus Z390 Pro Wifi | RAM - Ballistix sport LT 3200mhz red @ 3800mhz 15-20-20-35 1T 1.428V  |  Cooler - nh-d15  |  SSD - 960 evo 250GB, mx300 750GB, mx500 500GB  |  PSU - 750W G2  |   Case - Cooler Master MasterBox 5

Link to post
Share on other sites

Have you monitored the card temps/power usage?

 

 


CPU  Ryzen 5 1600x  MOBO  Prime X370-Pro  RAM  Hyper X Predator RGB 3000mhz  GPU  Sapphire Nitro+ Vega 56  BOOT Adata XPG SX8200 nvme SSD  Crucial BX 240gb x 2  HDD  WD Blue 500gb x 2  PSU Fractal Edison M 750w  Cooling Corsair H110i GT Extreme 280mm aio with 2 x 140mm Chromax ,Aerocool P7 F12 rgb fan x3 + controller CASE Phanteks P350X 

Link to post
Share on other sites
Posted · Original PosterOP
3 minutes ago, glenalz81 said:

Have you monitored the card temps/power usage?

 

 

During furmark, it can run steady in high 80s and never crash. During unigine, it crashes around high 70s from this last one I just ran. I don't know how to monitor power usage. Do you have any suggestions of software?

Link to post
Share on other sites

Afterburner - HWInfo - HWmonitor

 

All can show power usage.


CPU  Ryzen 5 1600x  MOBO  Prime X370-Pro  RAM  Hyper X Predator RGB 3000mhz  GPU  Sapphire Nitro+ Vega 56  BOOT Adata XPG SX8200 nvme SSD  Crucial BX 240gb x 2  HDD  WD Blue 500gb x 2  PSU Fractal Edison M 750w  Cooling Corsair H110i GT Extreme 280mm aio with 2 x 140mm Chromax ,Aerocool P7 F12 rgb fan x3 + controller CASE Phanteks P350X 

Link to post
Share on other sites
Posted · Original PosterOP
7 minutes ago, alatron978 said:

Check the psu output voltages for all the rails.

I have a HDE power supply tester but I've never used it before. Do I plug in each of the cables to the tester and check the voltage to verify its within reasonable range of 12, 5, or 3.3 Volts?

Link to post
Share on other sites
Just now, Lasne said:

I have a HDE power supply tester but I've never used it before. Do I plug in each of the cables to the tester and check the voltage to verify its within reasonable range of 12, 5, or 3.3 Volts?

Yes, the voltages should be slightly over what they are rated, if they are much lower, thats a problem.


PC: CPU - 8700K 4.8ghz -1avx 1.224V  |  GPU - 5700XT 2100mhz 1.145V target  |  Motherboard - Aorus Z390 Pro Wifi | RAM - Ballistix sport LT 3200mhz red @ 3800mhz 15-20-20-35 1T 1.428V  |  Cooler - nh-d15  |  SSD - 960 evo 250GB, mx300 750GB, mx500 500GB  |  PSU - 750W G2  |   Case - Cooler Master MasterBox 5

Link to post
Share on other sites
Posted · Original PosterOP

So, development. I started up unigine to monitor power usage and it shut down almost instantly. Then came the dreaded smell of burning electronics. I turned everything off and took out the GPU. Opened it up and see no visible scorch marks. But the smells seems to be coming from one of the 12V capacitors on the GPU. At the same time, my computer no longer powers on and I think the power supply is dead. So I am left with a dead graphics card most likely and definitely a dead powers supply. I have no way to test the state of my motherboard or processor. 

 

So, is it safe to say at this point that the PSU killed my GPU as it failed this final time? Just for ease of mind I want to know exactly what part kill which.Thanks for your help otherwise. I just came here too late to diagnose the issue before a catastrophic failure.

 

 

Link to post
Share on other sites

'twas likely the PSU at first...but it may've impacted by an on the way to fried GPU....HARD TO DIAGNOSE FROM AFAR...does your system haave onboard video? Try to do some stress testing, benchmarks once you replace the PSU itself...use google to find a PSU checker [input your components to arrive at a minimum desired PSU rating] and ensure you give yourself at least a 15% margin, also aim for at leat a Gold rated PSU from a known high quality manufacturer...

Link to post
Share on other sites
Posted · Original PosterOP

Alright i came back to the broken computer after stepping away from it for a few days. The power supply to my surprise turns back on and does work. I tried plugging my PSU Tester into it. The first picture is the reading from only the 24 pin connector from the motherboard plugged into the tester. the second picture is using the 6 pin connectors that were plugged into my GPU. The GPU normally uses a 6pin and a 6+2 pin. but I have no way to fit the 6+2 pin on the PSU tester I have. Do i need to get a new PSU tester? Also none of the green lights on the side for the 3.3, 5.0, and 12V light up for these PCIe cables.

 

In any event. Ive never used a PSU tester before and im not really sure what im looking at. Sorry if none of this is helpful

69214884_412767479598017_5193194529829683200_n.jpg

69317689_493077538170816_6220660533641609216_n.jpg

Link to post
Share on other sites
Posted · Original PosterOP
On 8/27/2019 at 10:58 PM, Pharr Carnell said:

'twas likely the PSU at first...but it may've impacted by an on the way to fried GPU....HARD TO DIAGNOSE FROM AFAR...does your system haave onboard video? Try to do some stress testing, benchmarks once you replace the PSU itself...use google to find a PSU checker [input your components to arrive at a minimum desired PSU rating] and ensure you give yourself at least a 15% margin, also aim for at leat a Gold rated PSU from a known high quality manufacturer...

I do have onboard graphics and I have stressed tested the processor with the GPU out before it died. I had no issues under load. The PSU still works and I can boot into windows. but I have refrained from doing any stress tests.

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Buy VPN

×