Fans stop spinning after running HWinfo

[d4rkAlf]

I've run into a very strange bug. Everything works fine until launching the sensors. When I do, the chassis fans running off the Asus P7P55D-E stop spinning. HWinfo manages to get one or a couple of readings on the fans, reporting whatever value they were spinning at before getting stopped by HWinfo. After being stopped they will naturally show 0 as RPM.

I encountered the issue at first while running version 448, upgrading to 450 did not help. The fans will spin briefly during POST but will shortly afterwards stop spinning again, even if HWinfo isn't launched. I've tried running HWinfo without reading the EC sensor, but the same issue manifests itself. The only way I've found so far to get them running again is to jump into the BIOS and disable and then re-enable Asus Q-Fan control.

The machine is a simple home server that I connect to using Remote Desktop.

 

[Martin]

Which fans are those that stop, those seen under the "ASUS P7P55D EC" sensor ?
I'd suggest the following:
1. Disable the entire "EC Support" in HWiNFO (in the Settings options before scan)
2. Perform a power off (you might also remove the power cord for a minute if possible)
3. Then start again the system and HWiNFO and check again
In this case you won't see any "ASUS P7P55D EC", it's just to prove if this issue is related to that sensor. So you'll have to check the fans directly if they are spinning.

 

[d4rkAlf]

Which fans are those that stop, those seen under the "ASUS P7P55D EC" sensor ?
I'd suggest the following:
1. Disable the entire "EC Support" in HWiNFO (in the Settings options before scan)
2. Perform a power off (you might also remove the power cord for a minute if possible)
3. Then start again the system and HWiNFO and check again
In this case you won't see any "ASUS P7P55D EC", it's just to prove if this issue is related to that sensor. So you'll have to check the fans directly if they are spinning.

Yeah, it's the fans Chassis1 and Chassis2 that are affected. They show up under the "ASUS P7P55D EC" sensor but also under the "Winbond/Nuvoton W83667HG" sensor. As far as I can tell the two sensors mirror each other, except that the EC sensor seems to be "frozen" and not update any of its values.
Hmmm, this time when I did as you asked they spin up to full speed instead. The "Winbond/Nuvoton W83667HG" sensor is still showing up and reporting the (seemingly) correct values. I've made a new report and debug file from this.

I tried it one more time after making this post and this time the fans stopped again. I did have debug mode disabled this time though.

 

[Martin]

Hm, that's interesting
Please try this build and let me know the result: www.hwinfo.com/beta/hw64_451_2416.zip

 

[d4rkAlf]

Hm, that's interesting
Please try this build and let me know the result: www.hwinfo.com/beta/hw64_451_2416.zip

The fans went to full speed running that build. I disabled EC support before launching the sensors.

I just realized the the operating system might not get included in the report. The server is running Windows Server 2012 R2 x64, in case that makes any difference.

Oh, another thing that I don't know if it makes any difference is that Asus haven't released any Windows 8/2012 drivers for this motherboard.

 

[Martin]

And do the fans stay then at full speed ?

 

[d4rkAlf]

And do the fans stay then at full speed ?

Yeah, after that they stay at full speed until I shut down the computer and then enter BIOS. Actually switching back and forth between Q-Fan enabled and disabled doesn't seem to be necessary.

The fans would stay off when I rebooted without entering the BIOS. I haven't tried just rebooting when the fans went to full speed.

 

[Martin]

Please attach a new Debug File from the latest build for analysis.

 

[d4rkAlf]

Please attach a new Debug File from the latest build for analysis.

I'm sorry but I've moved on to using a hardware based fan speed controller (since I didn't find Asus Q-Fan adequate). So I can't really help that much any longer.

Don't sweat about it though, I'm guessing this is a bug that has to do with this specific controller when running HWinfo under WS2012. I can't remember this behaviour in the past.

 

[Martin]

OK
Indeed, there was something weird and I was running out of ideas