Need help identifying an issue

unclesnu

New Member
So I've recently came across an issue with my rig in which when I play a graphically intensive game my monitors will go black showing no signal, I'll still hear sound, the fans keep spinning and the only fix is to reboot. Sometimes it reboots itself.

I used HWmonitor to make a log and it happened while I was playing a game so I have the .csv file. I downloaded a log viewer but I simply can't identify why this issue is happening. The temperatures don't seem high enough to shut my GPU down and I'm not really clued up on where to look for voltage or PSU issues. I can't think of what else would be causing this since it seems like a GPU failure.

When attaching the CSV log file I get "The uploaded file is too large for the server to process". I'm not sure how I can share this file, unsure if I'm allowed to post links to a download it elsewhere. If anyone can help I'd be forever grateful because I miss gaming! :(

My setup is as follows:

Displays
180hz 1440p Main Monitor - Display port
69hz 1080p Secondary Monitor - HDMI

Rig
Windows 10
i7-9700k CPU @ 3.60GHz
32 GB RAM DDR4
GeForce RTX 3060 Ti
Corsair 750W RM750X, 80+ Gold PSU
 
More like damage to the driver. Try to remove it using DDU and install it again.
Hey, I tried this and still have the issue. I've also tried swapping out my cables for my monitors to no avail :(

I saw someone on this forum with a similar issue but his fix was he was using a third party cable extender to his GPU yet I'm just using the cable that came with my PSU so I can't see how that could be the issue.
 
It seems that PSU power should be enough, and there should even remain a supply of 40 - 50 watts, but it is possible that the video card with high load simply does not have enough current through the EPS12V circuit and it turns off. You can try to take a more powerful PSU, for example, like my Aerocool Aero Bronze 850M with a capacity of 850 watts. I have enough of them to power about 720 watts of workstations have:

Xeon E5-2697 v2 / Intel C602 chipset / 128 Gb LRDIMM DDR3-1866 Quad Channel / GTX 1660 Ti / 4 x 3,5" int + 6 * 2,5" ext SATA 7200 rpm server HDD / DVD-RAM / USB 3.0 4 port PCIe card / USB 3.0 card-reader / USB 5 x 3.0 + 2 x 2.0 ports active 5,25" int HUB / AeroCool Aero BRONZE 850M (850VA) PSU

, and you have evaluative consumption (according to the elements you indicated) should be in the region of 550 W, so its capacity should be enough. His cables have long, strip, designed for the lower location of the BP in the midi tower case, and the EPS12V cable is in the factory kit. It is designed for power network used ~230V, and although in terms of its parameters, it corresponds to the requirements of the 80+ Gold standard, but since it does not support required on this standard dual switching power voltage ~110V or ~230V, it is formally certified according to the requirements of the 80+ Bronze standard, which does not interfere with its normal operation.
 
Last edited:
It seems that PSU power should be enough, and there should even remain a supply of 40 - 50 watts, but it is possible that the video card with high load simply does not have enough current through the EPS12V circuit and it turns off. You can try to take a more powerful PSU, for example, like my Aerocool Aero Bronze 850M with a capacity of 850 watts. I have enough of them to power about 720 watts of workstations have:

Xeon E5-2697 v2 / Intel C602 chipset / 128 Gb LRDIMM DDR3-1866 Quad Channel / GTX 1660 Ti / 4 x 3,5" int + 6 * 2,5" ext SATA 7200 rpm server HDD / DVD-RAM / USB 3.0 4 port PCIe card / USB 3.0 card-reader / USB 5 x 3.0 + 2 x 2.0 ports active 5,25" int HUB / AeroCool Aero BRONZE 850M (850VA) PSU

, and you have evaluative consumption (according to the elements you indicated) should be in the region of 550 W, so its capacity should be enough. His cables have long, strip, designed for the lower location of the BP in the midi tower case, and the EPS12V cable is in the factory kit. It is designed for power network used ~230V, and although in terms of its parameters, it corresponds to the requirements of the 80+ Gold standard, but since it does not support required on this standard dual switching power voltage ~110V or ~230V, it is formally certified according to the requirements of the 80+ Bronze standard, which does not interfere with its normal operation.

Sorry, forgive my understanding but that was a lot of info on PSU's which I'm not super clued up on. So are you suggesting I get a new PSU, that it could be a case that the card is unable to draw enough power through the cables?
 
I can’t say without dash measurements from afar what exactly is happening in the scheme - this is how the “exact” answer can give you only an ignoramus or fortuneteller ... with accuracy plus - minus infinity. And I am an experienced engineer and in my conclusions I rely exclusively on reliable facts, and in this case, I suggested the assumption based on the calculation of evaluative consumption currents in the configuration described by you. And my calculation shows that peak currents of consumption of this scheme can exceed the threshold of operation of protection against overloading by current in the PSU power used, which means that it makes sense to put a more powerful PSU.

After conducting calculations, we get the valuation capacity of the PSU of about 810 - 820 W, and to save money, take the low cost variant of needed power 850 W (> 820 W), and then we select a specific brand for criteria "parameters, no worse than the required, reliability and percentage of factory marriage that we watch that we watch for forums and reviews of repairmen - the fewer failures are found - the better, the minimum cost of ownership that consists of the purchase and operational costs. ” And it doesn’t matter which plant will make the product, it is important that it meets the entire set of requirements. This is a professional approach, lovers choose in the sequence "Advertising, manufacturer, series, power, minimum purchase price" and often get problems in the operation of the system as a whole. Do we need them with you?

As for the assumption of a shortage of power, everything is simple - we take the P0 power consumed for the circuit (is given for the rated frequency F0) and taking into account the dependence of the consumption current CMOS circuits on the frequency of work, we count the GPU consumption current at the maximum frequency F1 of the work according to the simple formula p = (1.1*P0*((F1/F0 +1)**0.5)/12 (**0.5 - square root - function squrt(x)). And then we look, this PSU can give such a current along the line +12V (if the parameters of the PSU marked something like "+12V1 20A, +12V 25A, +12V3 17A" it is necessary to smoke the current along all lines +12V, and look at the minimum figure since in such PSU to a common source +12V there are several wires to different currents , and we do not know which wire the consumer will be connected inside the PSU, well, it is advisable to avoid such PSU options - calmer).
 
I've uploaded two logs. One was taken with HWinfo and one was using MSI afterburner. Are you able to look at them to find the issue at all?
 

Attachments

After the log, your PSU works with overloading along the channel +12V because when the video card in a simple takes about 35 W, the voltage on the +12V channel slightly exceeds 11.9V, and with an increase in consumption, the voltage drops to the minimum permissible value. Personally, I would replace such a PSU - he has too weak a channel +12V with great internal resistance, which can be the root cause of the problems.
 
Back
Top