CPU Fan Speed

[cwinson]

I was checking the status of my HTPC yesterday using the AsusProbe tool and noticed that the CPU fan speed was registered as 0, even thought the fan was visually purring away.:confused:

I have a Panaflo 92mm low volume fan as my CPU fan (on top of SLK-800 heatsink), connecting into an A7N8X motherboard. Thats the only fan in the box (apart from PSU)

Should I expect the CPU fan speed to be monitored with this config, or do I need further connections for feedback ?

Any help would be much appreciated

Chris

[JohnS]

If the panaflo is connected with a 3 pin cable to the motherboard it should be able to be monitored by the motherboard but my doesn't either for some reason even though it did with a Papst fan I used.

[cwinson]

John,

Thanks for the reply - at least I'm not the only one who's seen the problem.

BTW - will let you know how the rest of the build goes - you may find my list of components I'm using somewhat familiar :blush:

- Kanam Accent HT-200 Case (new version of Dign case)
- Asus A7N8X (non-deluxe) motherboard
- AMD 2100+ Thoroghbred B core
- Seagate V Baracuda 120GB IDE 7200RPM
- Sony DRU500AX DVD drive
- Corsair Twin XMS 512MB DDR PC3200 (2 x 256MB)
- Sapphire Radeon 9000 Atlantis 129MB
- ThermalRight SLK-800 heatsink + Panaflo 92mm Low Volume fan
- Packard Bell TV-100
- Gyration Ultra Suite (RF keyboard & mouse)

You may find some more questions heading your way....:)

Cheers,

Chris

[mephistopheles]

The 'standard' Panaflo, incl. those available at "The Overclocking Store", don't have an rpm signal; for this feature the manufactureer part number has to have a '1BX' suffix, eg FBA09A12L1BX for the low volume 90x25mm. 'Standard' Panaflos don't have the final 3 character suffix.

FWIW, there are Panaflos with a thermal sensor in the hub for variable speed, which have a suffix '1B1', and others with a locked rotor arm sensor with suffix '1CS'.

Papst fans have rpm signal as standard issue, but there is a Papst 'Variofan' with a third wire connected to a NTC temperature sensor and onboard voltage regulation circuitry, which causes the fan to spin at between 50-100% in the 25/30-50C range, depending on the fan size. These NTC modules can be located either directly in the hub or remotely, and would seem to be an ideal chassis fan solution. Additionally, the Variofans can be undervolted, as long as it is by no more than the minimum voltage per spec, and it can be by means of pulse width modulation as long as it is not more than 55%.

[JohnS]

The 'standard' Panaflo, incl. those available at "The Overclocking Store", don't have an rpm signal;

Well that explains it:)

[cwinson]

Wow mephistopheles, that's a fairly comprehensive answer :eek:

These fans are more complex than I gave them credit for :blush:

With regard to the 'thermal sensor in the hub for variable speed' functionality, I presume this is similar to the Q-Fan technology in my A7N8X motherboard. In fact, I guess the rpm monitor my fan is missing would really only give me positive proof that the Q-Fan technology is working (fan slowed under lower temps) - something I'm happy to assume for the moment ;)

Chris

[JohnS]

I guess the rpm monitor my fan is missing would really only give me positive proof that the Q-Fan technology is working (fan slowed under lower temps)

I must admit that was the only thing I wanted it for but its so quiet even on the quickest setting I haven't bothered with the MB control on this machine. I use it on my desktop machine though with a stock fan and it slow it from 5000 to 4000rpm.

[Underscore]

I tried the Q-fan on my A7V8X when I first built my HTPC. Like John, I found that it did slow the fan a little, but nowhere near enough. I ended up disabling it and using the fanmate that came with my flower cooler. Even at the lowest setting it provides plenty of air flow for my system - though the ASUSProbe starts whinging due to the slow (1400-1500rpm) fan speed.

_

[mephistopheles]

Although it's a little off topic, it might be worth mentioning the noise issue with regard to fan speed, since that is presumably what we're all seeking to eliminate.

Whilst the blade rotation itself will obviously output at least some noise, the total noise output in any given system is mostly attributable to vibration and resonance. However quiet a fan might be in its own right, so to speak, if it is not fastened to rigid mountings and prevented from vibrating, it will set up resonance. PC cases, unless they are very strongly constructed from heavy gauge sheet (aluminium is ideal), are insufficiently rigid in this respect.

With the help of a Papst technical engineer, I have located a manufacturer of specially damped mountings which are made from an acoustic barrier material, which I shall be testing out. The same manufacturer also supplies rather interesting acoustic insulation materials, which, from the specs, would make a highly effective case lining. It's probably similar to the Akasa type material available in kit form, but more efficient and cheaper.

I'll be checking all this out shortly for the mods to my CRT monitor and will report back any worthwhile discoveries!