RPM fan monitoring signal

[Goi]

Hi,
I was just wondering, for the 3 pin fans where 1 pin is for RPM fan speed monitoring, what kind of voltages does that pin carry? I'm assuming it's just a voltage level between 0 and some Vcc, where 0V indicates a fan that's not spinning and Vcc(12V?) is full speed, but if that's the case, then wouldn't fans that inherently spin slow(larger fans for example) never reach Vcc?

I'd like to know exactly how it's done, and I've searched some of the major fan manufacturer's sites(Sanyo Denki, Sunon, YStech) but couldn't find any relevant information.

Any detailed description and/or links would be appreciated, thanks!

 

[Goi]

anyone? I googled a bit more and apparently they carry an AC square wave, sorta like a PWM signal, and the frequency corresponds to the RPM. How does that work though?

 

[Goi]

bump

 

[harrkev]

I do not know this first hand, but I believe that the circuit is open-collector, one pulse per revolution.

 

[Zepper]

Info on sensor:

http://www.comairrotron.com/pdf/Tachometer%20Output%20Option.pdf

.bh.

 

[Goi]

Thanks! That is great info. Seems like most PC fans are non-isolated.

 

[Goi]

Hi,
According to the pdf, the tachometer output should be a square wave of 50% duty cycle. However, when I use an oscilloscope to scope the fan tachometer output using the fan ground as scope ground, what I got was a ~200mV pulse signal that jumps to 0V at equal intervals. The first half of the period is fluctuating with an average at ~200mV while the 2nd half of the period is steady at ~200mV. This is very different from what the pdf shows. Am I scoping it wrong, or is my fan bad, or can the signal I obtained be explained in some other way?

 

[harrkev]

Try adding a pull-up resistor. Experiment a little, and try using a 1K, 5K, and 10K resistor. Choose the largest one that still gives a good wave-form.

 

[Goi]

Thanks. I tried a 330-5K pullup resistor to 12V and it gave me a waveform that's slightly above 1V peak voltage. I was under the impression that it would be 12V though.

 

[harrkev]

If you are hooking this thing up to digital electronics, then pull-up to 5V or less. My guess is that the reason that they use open-collector is so that you can choose how high the signal can theoretically go.

You can always use an op-amp to amplify the signal. You can put the negative terminal at 0.5V and run the fan input into the positive terminal open-loop. Might be noisy on transition, but should work. Hook up a large feedback resistor back to the positive terminal to prevent glitching (but then you have to run your input through a small resistor).