cancel
Showing results for 
Search instead for 
Did you mean: 

Crosshair VIII Hero: Temperature Sensors and Fans Questions

ationfictons
Level 10
Temperature Sensor Issues:
Right now, I'm using HWMonitor to try to see what my fans and temperatures are doing on my Crosshair VIII Hero motherboard. HWMonitor's CPU and Mainboard temperatures pretty clearly match up to the BIOS' CPU and Motherboard temperatures. But, I'm just assuming Package (Node 0) is CPU Package, TMPIN2 is VRM and TMPIN3 is T_Sensor since the temperatures seem pretty close (see below). Does that seem reasonable? Does anyone have any idea what TMPIN4 is and why it never moves?

Background Explanation:
I've got a Crosshair VIII Hero and I've installed an XSPC Wire Sensor 10K

https://www.amazon.com/gp/product/B00CMR38LC/ref=ppx_yo_dt_b_asin_title_o01_s00?ie=UTF8&psc=1

on the T_Sensor connector. The BIOS lists 7 temperature sensors:

  • Motherboard (about 41 C)
  • CPU (about 47 C: swings wildly)
  • CPU Package (about 58 C: changes quite a bit)
  • VRM (about 37 C)
  • T_Sensor (about 38 C)
  • Water In (N/A since I've disabled it)
  • Water Out (N/A since I've disable it)

HWMonitor list 5 temperature sensors under the motherboard area (all temperatures at idle):

  • Mainboard (about 41 - 43 C)
  • CPU (about 44 - 69 C: swings wildly)
  • TMPIN2 (about 41 - 43 C)
  • TMPIN3 (about 43 - 48 C)
  • TMPIN4 (21 C: never changes)

HWMonitor also has a 6th temperature sensor under AMD Ryzen 7 3700x (also at idle):

  • Package (Node 0) (about 44 - 69 C: swings a lot)

Fan Step Up/Down Issue:
I'm assuming that the Step Up and Step Down times in the BIOS' Q-Fan area are to add a delay to how quickly the fans respond to temperature sensor changes. Those delays don't seem to work for me. The fans seem to change almost instantly, no matter what I set those Steps to (which is why I installed that T_Sensor). Is my assumption wrong?

TLDR BIOS Fan Curve Information:
-CPU Fan:

  • Q-Fan Control = PWM Mode
  • Step Up = 2.1 sec
  • Step Down = 25 sec
  • Speed Lower Limit = 200 RPM (defaulted to that)
  • Fan Profile = Manual
  • Upper Temperature = 75 C
  • Max Duty Cycle = 100%
  • Middle Temperature = 60 C
  • Middle Duty Cycle = 40%
  • Lower Temperature = 40 C
  • Min Duty Cycle = 14% (defaulted to that after Optimizing the fans) EDIT: now 20%

-Chassis Fans 1 - 3:

  • Q-Fan Control = PWM Mode
  • Q-Fan Source = T_Sensor
  • Step Up = 12 sec
  • Step Down = 25 sec
  • Speed Lower Limit = 200 RPM (defaulted to that)
  • Fan Profile = Manual
  • Upper Temperature = 70 C
  • Max Duty Cycle = 100%
  • Middle Temperature = 50 C
  • Middle Duty Cycle = 20% (EDIT: Now 40%)
  • Lower Temperature = 40 C
  • Min Duty Cycle = 14% (defaulted to that after Optimizing the fans) (EDIT: now 20%)

High Amp Issue:
When I ran the BIOS' Optimize fans routine, it tested all my fans at set defaults. For some reason, even though nothing is connected to the High Amp motherboard connector (and the fan speed is listed as N/A), it set up a manual DC fan curve for it. Is there some reason for that? Should I just disable everything to do with the High Amp connector? (EDIT: I've now Disabled this)
11,641 Views
7 REPLIES 7

ationfictons
Level 10
I'll just add that I'm trying out HWInfo instead of HWMonitor. It looks like HWInfo exposes more sensor/temperature data. I'm going to have to dig through it and see if that answers some of my earlier questions.

RedSector73
Level 12
ationfictons wrote:
I'm assuming that the Step Up and Step Down times in the BIOS' Q-Fan area are to add a delay to how quickly the fans respond to temperature sensor changes. Those delays don't seem to work for me. The fans seem to change almost instantly, no matter what I set those Steps to (which is why I installed that T_Sensor). Is my assumption wrong?


They are set to measure temp from either CPU or VRM or Motherboard or all, you cant set it to external sensor. The step up delay will delay the time after an trigger event happens but it will still happen once trigger point is reached, the step down means once trigger is reached and assuming no other trigger is presented, scale down after x amount of time.


ationfictons wrote:
Min Duty Cycle = 14% (defaulted to that after Optimizing the fans)


I would not have min duty cycle of 14% regardless of what it found, the fan cools it own mechanism and if you scale below 20% (decent quality fans, around 40% for cheap ones) without switching off the fan, then your asking for trouble, in that the fan will produce heat it cant easily get rid of and this can lead to excessive wear of the fan / death of the fan. I would recommend baselines above 30% even for something like noctua fans.

ationfictons wrote:

High Amp Issue:
When I ran the BIOS' Optimize fans routine, it tested all my fans at set defaults. For some reason, even though nothing is connected to the High Amp motherboard connector (and the fan speed is listed as N/A), it set up a manual DC fan curve for it. Is there some reason for that? Should I just disable everything to do with the High Amp connector?


If there isn't anything in it and it's disabled, not sure why this really matters ?

Thanks for confirming Step Up and Down. But, they don't seem to be working. That's OK for me now, since I'm using the ambient air in the case via the T_Sensor as the controlling temperature and that doesn't change very fast.

I'm still in the process of tuning the three case fans (Noctua NF-A14 PWM fans), so the Min Duty Cycle = 14% is just a first stab. They're actually keeping everything at good temperatures at idle. As you say, 20% is probably more reasonable. (EDIT: I've now bumped the Min Duty Cycle for all the fans (including the CPU fan) up to 20% -- hasn't really made a difference, but it give me a bit of a buffer, I suppose)

Regarding the High Amp thing, I just don't understand why the BIOS thought it was testing a fan on that connector and setting a manual fan curve. I was wondering if there was some hidden reason. I guess I'll disable that in Q-Fan and see what happens. (EDIT: disabled -- no difference that I can see)

82766

My board has the same issue for Step Up and Down (bios 1001). There is a new BIOS due out next month.

I'm still in the "rough-in" stage of tuning these fans/temperatures. But, with the addition of the thermal sensor, I'm much happier with the fan noise variance. I'm using a Define R6 USB-C case with all Noctua fans, so the absolute noise isn't a problem. It's just that with the CPU temperature swinging wildly at all times (especially at idle), the case fans having been tied to the CPU before, and the Step Up/Down not working, the CHANGE in noise was the problem. Now, I need to run some games and make sure the fan curves are OK at load.

BTW: I've switched over from HWMonitor to HWInfo since it finds and reports all the sensors/fans (I think). It's odd that the PCH temperature is 65 C at idle with the PCH fan spinning madly at over 3,100 RPM. That's over 30 C above the in-case air temperature (as reported by the T-Sensor (which is actually resting just above my video card and next to my Noctual NH-D15S CPU cooler)). At idle, that's the hottest thing in the whole system, by far.

might want to try the BIOS here https://rog.asus.com/forum/showthread.php?114673-New-AGESA-1-0-0-4-BIOS-download
and and see if they have corrected step/down. Once thing I noticed was that you can now select a probe to monitor temp / fan settings off.

Hi,

I've had similar problems with fan step down-up with my C8I, also wasn't able to control fans by temp sensors, but few day after opening a ticket with Asus, a new BIOS (1302) popped up on the support site and that's fixed my problem, fan control works as expected now, including step down-up.

You have the same version available for your board as well, give it a go if you haven't done it yet.