Bug: Incorrect fan speed average calculation when unused sensors report 0 RPM

[erkutkoc]

Has this issue been covered in the Wiki?

• I have read the Wiki and my issue is not covered.

Is there an existing issue reported already?

• I have searched the existing issues, it does not exist.

What is the unexpected behavior?

The application currently calculates the average fan speed by dividing the sum of all detected sensors by the total number of sensors found. On my laptop, the system reports 4 fan sensors, but only 2 are physically present and active. The other 2 stay at 0 RPM, which drags down the average and provides inaccurate monitoring data.
The application should only consider active sensors (RPM > 0) in the average calculation, or allow the user to ignore specific sensors in the settings.
The current "ignore null values" feature doesn't work here because it treats these ghost sensors as valid fans that happen to be "stopped" (0 RPM), rather than non-existent hardware.

As shown in the attached screenshot, the WMI 3 and 4 sensors are incorrectly detected and permanently report 0 RPM :

Steps to reproduce the unexpected behavior.

1. Launch Vitalis on a device where the hardware interface (e.g., WMI) exposes dummy/inactive sensors (as seen in the screenshot for WMI 3 & 4).
2. Observe the Fan Speed
3. Note that the average calculation includes sensors reporting 0 RPM.
4. Apply a high CPU/GPU load to force the active fans (WMI 1 & 2) to spin.
5. Observe that WMI 3 & 4 remain at 0 RPM regardless of system load, proving they are non-existent or "ghost" sensors that should be excluded from the average.
6. Average use 4 fans instead of the 2 actives/connected fans.

Relevant log output

What distribution and version of Linux are you using?

Ubuntu 26.04 LTS, Kernel : 7.0.0-15-generic

What version of Gnome are you using?

GNOME Shell 50.1

[ba2hannn]

Hi @erkutkoc, thanks for reporting this bug! I can confirm this is a real issue.

I've been looking at the code in values.js (around line 253-274) where the fan average is calculated. The current code takes all sensors from this._history['fan'] and averages them, without filtering out 0 RPM ghost sensors.

On my own laptop, I also don't have any fan sensors exposed through hwmon at all (no fan*_input files in /sys/class/hwmon/), so this is a broader problem where:

1. Ghost sensors reporting 0 RPM skew the average
2. Some systems have no fan sensors at all, leaving the fan section empty

I think the fix should:

1. Filter out 0 RPM values when calculating the fan average (unless ALL sensors are at 0, in which case show 0)
2. Optionally: add a per-sensor hide/ignore feature for ghost sensors

I'd be interested in working on a PR for this. Would the maintainers be open to this approach?

[ba2hannn]

Hi @erkutkoc, I've included a fix for this in PR #551. The change filters out 0 RPM values when calculating fan average, min, and max in values.js. If all sensors report 0 RPM, it still shows 0 (preserving expected behavior).

Would be great if you could test and confirm!

[erkutkoc]

Hello @ba2hannn , I have the expected behavior now. Thank you.

Before fix :

With fixed version :