Fan RPM readout issues.

[TacticalTimbo]

I've got a Gigabyte GA-Z97P-D3 motherboard with identical Noctua 3 pin fans connected to both the CPU and sys1 headers. Until recently they both reported RPM, now it's only the CPU header that reports. I believe this data comes from pin 3...

I've tried switching the fans around, with no change, proving the fan isn't to blame. I even tried connecting the fan to sys2 header, with no RPM reading there either, making the header look blameless.

Finally I tried connecting a 4 pin Corsair fan I had lying around, this time I got a read out from the sys1 fan header...

I don't understand why it was happy pumping out RPM data from my 3 pin Noctua for a time, until all a sudden it needs a 4 pin to do this!? Especially as the CPU header is still happily churning out RPM data using the Identical 3 pin fan!?

 

[TacticalTimbo]

I've managed to find a different Cooler Master 3 pin fan, and this sits pretty stable at between 593-633 RPM, when idle.

I found that the Noctua NF-P12 would show under load with prime95, giving crazy readings between 63-2600RPM (more than the fan's physically capable of producing), though on average being 800ish RPM... Why is this happening?

 

[Mellon Head]

What software are you using to read the fan's RPM?

 

[TacticalTimbo]

What software are you using to read the fan's RPM?

I'm using the PC health section of my Gigabyte's UEFI/BIOS, and HWMonitor; behaviour seems to be reported similarly by both.

 

[Mellon Head]

Sounds to me as though your motherboard's fan sensor is giving the weird readings. What do the fan speeds show in BIOS?

 

[TacticalTimbo]

Sounds to me as though your motherboard's fan sensor is giving the weird readings. What do the fan speeds show in BIOS?

Same thing as HW Monitor app in Windows environment; mostly just 0 RPM, but then after some time, once it's warmed up no doubt, it gives erratic readings between 0-2500 RPM (without the audible revving to back this up) .

 

[TacticalTimbo]

Right now it's calming the RPM is over 4000, simply not possible from this fan, and nowhere near the noise there'd be, if it really was... (see pic)

 

[Mellon Head]

Looks like you have a borked fan sensor then. Swapping out fans proved that it wasn't a fan problem, and it happens in BIOS too. The only thing that I can conclude is a bad sensor on the motherboard.

 

[TacticalTimbo]

Looks like you have a borked fan sensor then. Swapping out fans proved that it wasn't a fan problem, and it happens in BIOS too. The only thing that I can conclude is a bad sensor on the motherboard.

Would you expect both sys1 and sys2 headers to be affected, because they both misbehave with either of my Noctua fans? At the same time both Noctua fans don't misbehave on the CPU fan header.
If the sensor was borked, why would a 4 pin Corsair, and a 3 pin Cooler Master both act normally?

 

[Mellon Head]

Would you expect both sys1 and sys2 headers to be affected, because they both misbehave with either of my Noctua fans? At the same time both Noctua fans don't misbehave on the CPU fan header.
If the sensor was borked, why would a 4 pin Corsair, and a 3 pin Cooler Master both act normally?

Perhaps I misread your posts...

If the Corsair and the Cooler Master fans both work fine, the problem could be in the Noctua fans themselves. Perhaps they are drawing more current than the fan header can supply, or are simply incompatible with the fan sensor chip, which I seriously doubt.

As for why they work on the CPU fan header with no problems, I honestly don't know.