06-03-2017 09:51 AM - last edited on 03-05-2024 07:10 PM by ROGBot
02-02-2018 09:11 AM
Jsunn wrote:
So you are saying that the fan curves in the BIOS are NOT based on CPU temp? This doesn't seem accurate to me. All of my fans will ramp up and down based on CPU temp.
What do you mean by Tcase?
Sorry for the basic question.
Thanks,
Jason
Tcase: Temperature measurement using a thermocouple embedded in the center of the heat spreader. This initial measurement is done at the factory. Post-manufacturing, the BIOS calibrates Tcase. A diode between and below the cores delivers a reading.
02-05-2018 12:59 AM
Silent Scone wrote:
This is a Tcase reading that is utilized for fan curve control.
Aghoithae wrote:
I aquired a Strix Z270F myself, and been suprised by this limitation too, and I have to add that the "CPU fan step up" and "CPU fan step down" settings in the bios are not working (tried with DC and PWM fans).
?
02-05-2018 04:32 AM
RidingTheFlow wrote:
Registered just to chip in, since I've bought an Asus Prime Z270A motherboard recently and was totally baffled when I saw this.
This is not entirely accurate for modern mainboards. Intel got rid of TCase since 7th generation - mainly because it was not possible to use as accurate measurement. TCase required specially-modified heatsink and "software offset" in BIOS never worked that well.
So they switched to using solely TJunction/core temperature sensors. Modern CPU datasheets only spec the TJunction max and not a TCase (since there are no means to even know TCase anymore).
E.g. both Z270 boards I had report "CPU temperature" in BIOS via "c" interface - which is basically average of core temps, which can reach 90-degrees during normal operation without any thermal throttling.
More details here - http://www.tomshardware.co.uk/forum/id-1800828/intel-temperature-guide.html
In any case, 75 degree max for fan curve is just arbitrary and low restriction. Why is it even here? Is it to protect the user from doing stupid things? I remind you this is "gaming" motherboard, and it already have much more dangerous CPU-destroying options like voltage overclocks and so on. Not to mention that its not really possible to damage modern Intel CPU even by switching fan off - as long as heatsink is present, it will simply throttle to low speed (since it has enough TDP on low speed for passive cooling).
I find it ridiculous that Asus gaming motherboard fan control is full of such arbitrary restrictions. E.g. you can only control Pump header speed from CPU temperature - why is that? Its just another header, third party software (like Speedfan or Argus Monitor) can control it like any other header from any source, but no, Asus thinks it must be arbitrary restricted just because they named it "Pump".
Asus, please remove these restrictions. No competitor board I've tested have these. I want to be able to use high temperatures for fan curve points, since air cooling most efficient/quietest with hottest heatsinks. Now I have to use third-party software to override BIOS or make hacked BIOS to to this.
And yeah Asus, please fix fan speed change delays too. They don't really work and don't serve their main purpose - they don't prevent abrupt fan speed changes on fast temperature spikes (common with Kaby lake CPUs for example). This causes unnecessary noise and wear as PWM fans are forced to abruptly change speed.
02-05-2018 04:52 AM
Silent Scone wrote:
PECI is used, however, if the IHS temp is at 75c, core temps will be anywhere between 10 to 15 degrees Celsius higher.
02-05-2018 05:18 AM
02-05-2018 05:42 AM
02-05-2018 06:00 AM
Silent Scone wrote:
The value is averaged.
1) It is sensible to have some buffer to account for variance.
2) You'd struggle to put a temp probe on a GPU in a way where there would be no offset.
03-20-2018 10:34 AM
01-22-2019 02:47 PM
03-08-2019 12:03 AM