I recently upgraded from a GTX 560 to 660 on my sig rig, largely because the 560 had chronic VRM overheat problems, despite adding a huge VRM heatsink and fan.
After extensive load testing the new 660, I installed my Koolance GPU-220 on the die and Enzotech MOS-C1 heatsinks on each of the 8 VRM transistors.
Now I'm worried that the lack of airflow over the VRMs might cook them. I have an infared thermometer, and it says the heatsinks are getting to almost 120C during Furmark! Is that too hot for VRMs?
However, I can't set the emmisivity factor on the thermometer, and its hard-set to .95. But the emmsivity of polished copper is .06. I have no idea how that will affect my readings. I usually use electrical tape on what I'm reading, but the heatsinks are too small for that.
I also have a contact thermocouple thermometer, but I think if I try to use that, I'll just knock the sinks off.
Does anyone know if my infared readings are accurate? Should I even be worried about this? I have no plans on overclocking the 660; I'm CPU bound anyway.
After extensive load testing the new 660, I installed my Koolance GPU-220 on the die and Enzotech MOS-C1 heatsinks on each of the 8 VRM transistors.
Now I'm worried that the lack of airflow over the VRMs might cook them. I have an infared thermometer, and it says the heatsinks are getting to almost 120C during Furmark! Is that too hot for VRMs?
However, I can't set the emmisivity factor on the thermometer, and its hard-set to .95. But the emmsivity of polished copper is .06. I have no idea how that will affect my readings. I usually use electrical tape on what I'm reading, but the heatsinks are too small for that.
I also have a contact thermocouple thermometer, but I think if I try to use that, I'll just knock the sinks off.
Does anyone know if my infared readings are accurate? Should I even be worried about this? I have no plans on overclocking the 660; I'm CPU bound anyway.
Tractorbox
(16 items) |
|
Tractorbox
(16 items) |
