[SOLVED] GPU going over voltage limit and turning off screen, Don't know how to fix.

Dec 16, 2020
18
0
10
So I have just recently built my first new computer everything seemed fine at first, but once I downloaded and began to play a game, the game would have its first loading screen and then my screen would just turn off/lose connection. I could still hear the game audio and keyboard and mouse were both still responsive. Once I closed the game, my screen would soon turn back on and not have any problems. I had tested my GPU before on a friends computer and it worked fine, but I still assumed this was GPU problem and checked both GPUZ and MSI Afterburner and saw that my graphics card was hitting 1.062v while the limit was 1v, according to Afterburner. I have tried lowering my core and memory clock speed, a small overclock, reinstalling my drivers, and undervolting my GPU whilst underclocking(This one slightly worked, but every now and then my screen would still turn off). I have been searching up many things related to this for 2 days now and I do not what else to do. I feel like it might be a hardware issue (either GPU or PSU).

My Specs are:
GTX 1070 OC (Aero ITX)
Ryzen 5 3600
G. Skill 2x8gb 3600mhz CL16
Gigabyte B550m DS3H
WD SN550 500gb NVMe
EVGA 600w 80+ bronze (got off of B-stock)
Windows 10

Just checked again, If I set VSYNC on it becomes more stable, but it doesn't fully fix it.
 
Last edited:
Solution
Update: So I do have another power supply, but its a SENTEY 550w 80+ BRONZE PSU and it looks like it could be worse than the PSU I have right now. My friend's PSU, the one I tested my GPU with, was a Thermaltake 500w 80+ White PSU. I'm not sure if I should take the risk of testing my other PSU, or just investing some money for a brand new one.

These are all pretty bad PSUs.

Since these are not good PSUs, you should invest some money in a decent one, regardless of whether it's the cause of this particular problem. It's better to buy a high-quality PSU once every 10 years rather than 2-4 low-quality PSUs over the same timeframe.
Yup, swapping out the PSU is the best idea here. Not sure why this one was used in the first place; it's not an appropriate model for these components. On first glance, this appears to be a case of cheaping out on the least-fun but most-important component.
 
Update: So I do have another power supply, but its a SENTEY 550w 80+ BRONZE PSU and it looks like it could be worse than the PSU I have right now. My friend's PSU, the one I tested my GPU with, was a Thermaltake 500w 80+ White PSU. I'm not sure if I should take the risk of testing my other PSU, or just investing some money for a brand new one.
 
Update #2: Monitor now turns off sometimes when Alt-Tabbing. I've reinstalled my GPU drivers again and set my speeds to be a factory settings.
 
@Knix
That probably isn't any better.
An alternative you have with the current psu: In either Msi Afterburner or EVGA Precision X, test with different power limit settings: 75%, 50%, 25%, 0%.

Update #2: Got a faulty monitor cable there? Have you tried different ones?


@MacoyGG
You should always clean install Windows after a motherboard change: https://forums.tomshardware.com/faq/windows-10-clean-install-tutorial.3170366/
After that, update all your drivers - chipset, sata, lan, audio, etc.
 
@Knix
That probably isn't any better.
An alternative you have with the current psu: In either Msi Afterburner or EVGA Precision X, test with different power limit settings: 75%, 50%, 25%, 0%.

Update #2: Got a faulty monitor cable there? Have you tried different ones?

1: Ok will do and give an update on it. I have tried increasing my power limit as well, but it didn't work either.

2: I have tried 2 different HDMI cables and my 2 different HDMI ports on my GPU and I'll be trying to test my HDMI cables on a different output(My TV in this case) to see if it might be a monitor problem.
 
Update #3(Hefty one): So I have a DVI-HDMI Adapter that I am using now and it has gotten rid of some red lines and most black screens that used to appear if I had my HDMI cable directly connected to my GPU. I loaded up a game with VSYNC on and it was working well, but as soon as I turned it off(VSYNC), the screen went black and I saw that my PerfCap reason on GPU-Z was VRel.

Also have to mention that I thought before that the clock speeds/voltage was the problem, now that does not seem to be the case. When I was testing the game with VSYNC and Power limit 75%, My highest clock speed was around 1840(was around 1531 before turning off screen earlier) and my highest voltage was around 1.062 (Could not see anything higher than 0.9v before screen turning off earlier), so now I am almost completely lost on what it could be that's turning off my screen, but I do still have some speculations.

1. I could try removing the power limit cap as I tried testing on another game and the PerfCap was "Pwr".

2. I could try to lower my clock speed again with this new adapter.
 
Update: So I do have another power supply, but its a SENTEY 550w 80+ BRONZE PSU and it looks like it could be worse than the PSU I have right now. My friend's PSU, the one I tested my GPU with, was a Thermaltake 500w 80+ White PSU. I'm not sure if I should take the risk of testing my other PSU, or just investing some money for a brand new one.

These are all pretty bad PSUs.

Since these are not good PSUs, you should invest some money in a decent one, regardless of whether it's the cause of this particular problem. It's better to buy a high-quality PSU once every 10 years rather than 2-4 low-quality PSUs over the same timeframe.
 
Solution
Update #4: I tried to load up another game without VSYNC and, as expected, the screen almost instantly went black. PerfCap reason was VRel.

But theres something I don't understand. On the game with VSYNC on and power limit at stock, I played one match which went smoothly(no black screens), but once I loaded back into the lobby, my screen began to turn off and on again/continued losing connection every 2-ish seconds. I checked GPU-Z and nothing seemed out of the ordinary. There was no PerfCap, core clock spiked up from ~1600 to ~1800 and the voltage spiked up from ~0.6 to ~0.8 (which I don't think should cause the screen turning off). I thought I was beginning to narrow down my problems to it being my display ports, but this just completely messed me up and I feel like I'm back to square one.