Skip to the bottom if you want a quick summary and for specs.
A while ago I changed my GPU thermal paste with a paste that came with my CPU cooler, which I now think was bad quality, and the temperatures for my GPU have recently been higher than I'm used to. So yesterday I replaced the GPU thermal paste with Arctic MX2 from Amazon. This is where it gets weird.
I thought the temperatures were OK yesterday - at a lower rpm, my temperatures seemed to be cooler. When it gets hot, I notice a smell. This morning while playing I noticed that smell (usually noticed over 70/80 celcius), but the reading from the sensor (checked on multiple programs) was only 50 celcius which is a cool temperature. I put my hand on the graphics card and it definitely felt more like 80 degrees.
As a joke, I blew on the GPU to cool it down. As I had closed the video game, it cooled down slowly and the fan (PWM) reduced it's RPM like usual. The temperature shown on the EVGA program now reads -40 somehow... and the fan is automatically at full RPM. HWMonitor, a program which also reads the temperatures, shows the normal temperature of about 40 degrees. So, the EVGA software has triggered something? I'm not sure why it says -40 and is at full RPM.
SUMMARY: I changed my GPU's thermal paste to a better paste because of high temperatures. It seemed fine and the PWM was working, but the temperature reading from the sensor in the GPU was inaccurate. This caused the GPU to get very hot, as the temperature used by the PWM was lower than the actual.
Now, the EVGA software reads -40 degrees celcius by default, and has set the fan's RPM at maximum. What's happened?
GPU: EVGA 1060 6gb (1 fan only)
A while ago I changed my GPU thermal paste with a paste that came with my CPU cooler, which I now think was bad quality, and the temperatures for my GPU have recently been higher than I'm used to. So yesterday I replaced the GPU thermal paste with Arctic MX2 from Amazon. This is where it gets weird.
I thought the temperatures were OK yesterday - at a lower rpm, my temperatures seemed to be cooler. When it gets hot, I notice a smell. This morning while playing I noticed that smell (usually noticed over 70/80 celcius), but the reading from the sensor (checked on multiple programs) was only 50 celcius which is a cool temperature. I put my hand on the graphics card and it definitely felt more like 80 degrees.
As a joke, I blew on the GPU to cool it down. As I had closed the video game, it cooled down slowly and the fan (PWM) reduced it's RPM like usual. The temperature shown on the EVGA program now reads -40 somehow... and the fan is automatically at full RPM. HWMonitor, a program which also reads the temperatures, shows the normal temperature of about 40 degrees. So, the EVGA software has triggered something? I'm not sure why it says -40 and is at full RPM.
SUMMARY: I changed my GPU's thermal paste to a better paste because of high temperatures. It seemed fine and the PWM was working, but the temperature reading from the sensor in the GPU was inaccurate. This caused the GPU to get very hot, as the temperature used by the PWM was lower than the actual.
Now, the EVGA software reads -40 degrees celcius by default, and has set the fan's RPM at maximum. What's happened?
GPU: EVGA 1060 6gb (1 fan only)
Facebook
Twitter
Instagram