View Single Post
Old 09-14-2014, 10:34 PM
nh_790 nh_790 is offline
Registered User
nh_790's PC Specs
Join Date: Mar 2014
Posts: 51
POST ID # = 727844
nh_790 Reputation: 23

Originally Posted by wytnyt View Post
if your basing your curve off the cpu,,erratic fan speed is common which is why basing it off the water temp is recommended ..
theres no way to set an accurate cpu curve simply because different programs produce different heat and loads.
I think I would agree water temps should be a better option but i have avoided it for 2 reasons:

1- the water temp in the h100 (or mine at least) doesn't vary much. I go from 28c or so at idle and never above 35 or 36 during extended prime 95 test (with fans at 1100 rpm, which is quiet enough i dont hear them) That may be simply because the cooler is doing its job, but i'm not sure i want to set a curve with data points that are only 1 or 2 degrees apart.

2- if you wait till the water gets hot to start cooling it (or start cooling it more aggressively via higher radiator fan speeds) it would seem to me that the cooling system has to dig itself out of a hole so to speak. It seems it makes more sense to measure the heat source and keep the water from heating up in the fist place. If my CPU starts getting hot i want to start cooling it as much as possible right away, and not wait till it heats some water first.

wytnyt, could you share your setup and configuration and the logic you used to get to that setup? are you in fact using a 1 or 2 degree change in water temp for the next point to spin the fans up? What are your min/max?
Reply With Quote