And since BIOS reports 12v, and now HWMonitor only reports an almost correct voltage, it confirms for me that it is a windows monitoring issue. IE: the sensors cannot be monitored correctly. A Guy
I'm disappointed that Aida64's trial version has certain functionality disabled, and in particular apparently the voltages (or at least 12V) is one of those that does not perform as it will in the licensed version. I'd like to share my own license key just for the test on the problem machine, but that would be a no-no.
But if we had at least one (or more) other hardware monitor product (like Aida64) that DID show 12V results on this particular motherboard and PSU (and it would be helpful to see what value displayed), that would be very interesting and informative.
In my experience using Aida64 on a number of different Win7 machines (the most recent just this past week, on a Lenovo K330 IdeaCentre i3 desktop) the sensor voltage readings have never failed to display correctly. I suspect the licensed Aida64 version would, like HWMonitor, show correct 12V voltage values for the current problem machine, but of course that's just speculation. I don't consider Speedfan to be in the same class as these products.
(I am particularly a fan of Aida64 because of its configurable OSD display of lots of hardware information in a small desktop object, as shown by my screenshot earlier. It also has extensive larger system information presentations like HWMonitor if I want to probe very deeply with large open windows, but for a 24/7 desktop "gadget" this little OSD is ideal for me.)
It's just honestly hard for me to believe that this problem is actually rooted in Win7, rather than something odd or unusual in the motherboard hardware or BIOS. I had a problem on my Supermicro C2SBX mobo a few years back with temperatures being reported by Aida64/Everest way off after I upgraded BIOS (because I felt I should "keep current"). Speaking with their tech support, they contended the problem came from Intel, with a new method of reporting temperatures being simply "low, medium or high" rather than the true numeric values. Reverting to the previous BIOS of course fixed the problem and the temperatures were now once again reported accurately... though I really never knew if it was Intel or Supermicro who'd messed up for real in that updated BIOS version. I'm still running just fine with the original prior BIOS version.
I've got to believe it's something like that in the current story.