Presenting the wrong CMOS battery value for MITX-CLS0

Hardware: Quanmax Inc MITX-CLS0, I/O controller Nuvoton NCT6106D
Hardware report: LibreHardwareMonitor.Report.txt

Hello! I believe there is a bug when presenting CMOS Battery/VBAT value for this hardware. I noticed that the battery voltage value did not seem to make sense. For working CMOS batteries I would expect a higher voltage.

I tested a couple of different cards I had laying around and got the same issue and checked by other means what the value should be. So, I dug a little more and looked at Nct677X.cs, here NCT6106D maps to NCT610XD (which covers NCT6102D, NCT6104D, NCT6106D).

When reading the voltage values, it reads out 9 values, from 0x300 to 0x309. But crucially it skips 0x306.

Nct677X.cs line 451

I looked into Nuvoton’s documentation for the chip and register 06h is reserved there:

So, it skips 0x306. However, in SuperIOHardware.cs, for NCT610XD it assumes that the reserved value was read(even if it ends up not displaying it, line 385):

SuperIOHardware.cs line 387

I hope I’ve understood how this code functions, but it seems to me that when presenting the battery voltage value, it is presenting the VTT register as if it was the VBAT one.

Start with the voltage-reading logic around line 451 of LibreHardwareMonitorLib/Hardware/Motherboard/Lpc/Nct677X.cs and the NCT610XD mapping around line 387 of SuperIOHardware.cs. Compare the skipped 0x306 register with the expected VBAT position using the attached hardware report and Nuvoton register documentation. Done means the MITX-CLS0 presents the CMOS battery voltage from the correct register rather than VTT.