Reverse engineering a three phase brushless motor fan speed controller

[plantgrower]

I wish to reverse engineer the controller for this 12 pole, 3 phase brushless motor fan http://www.hyper-fans.com/shop/bybrand/hyper-fan/hyper-fan conytoller [url=http://www.hyper-fans.com/shop/bybrand/hyper-fan/hyper-fan-speed-controller-and-extension-cable]http://www.hyper-fans.com/shop/bybrand/hyper-fan/hyper-fan-speed-controller-and-extension-cable  [/url]

The fan can be controlled between 35-100% of output. When the controller is removed, the fan goes to 100%.

Any ideas on what the fan controller is doing to control the fan, PWM or something else ?

I don't yet own one of these fans.

Would a Rigol 1054z be of any use in this situation ?

What connector is the fan using ? [url=http://i.ebayimg.com/00/s/MTEwMlgxNjAw/z/FpkAAOSw8d9UxsJq/]http://i.ebayimg.com/00/s/MTEwMlgxNjAw/z/FpkAAOSw8d9UxsJq/$_57.JPG  [/url]

[free_electron]

probably simple phase chopping

[NiHaoMike]

If it's anything like the common variable speed Delta fans, there's a full blown VFD inside with a DSP driving it.

[jaxbird]

The pic16f178x does 3 phase in hardware, you just need to confirgure it and buffer the outputs correctly.

It'll only set you back about $2 or so.

[Jeroen3]

The speed controller looks like an expensive 10k potentiometer at the control signal of the internal controller chip.

[plantgrower]

The speed controller looks like an expensive 10k potentiometer at the control signal of the internal controller chip.

Are there any guides on how to reverse engineer the signal they are sending?

[Jeroen3]

Guides? Most industrial control thingys with potentiometers use a current loop or a potentiometer as small voltage s4etpoint.
One leg to +10, +24 or Supply positive. One leg to GND or Supply negative. And one leg to Analog in.

But, you'd need to tear it down to know for sure that it is an potentiometer in a fancy box.

[scrat]

The motor is a permanent magnet brushless motor with sinusoidal or trapezoidal back-EMF.

There is some confusion about the terms, however cheap permanent magnet brushless motors are often called "Brushless DC" (BLDC), which is also the way they are controlled.

The simplest control method is brushless DC, where two phases are conducting at each time, resulting in a phase current which is a squared + -> zero -> - pattern. The commutation between phases is synchronized with rotor position, and the right instant for commutation is usually estimated by reading the voltage (back-EMF) on the non-conducting phase.
A rough speed regulation can be done simply imposing a certain PWM duty-cycle when the phase is conducting.
This kind of control requires a 3-phase inverter and a simple microcontroller capable of generating 3 complementary PWM signals (as someone already said).

At zero or low speed, commutation instant cannot be detected by means of back-EMF, so an open-loop arbitrary commutation is imposed (with a high current, and then with low efficiency), relying on the fact that the rotor will align to the phase having positive current. This could explain why speed is limited to >35%.

[nowlan]

I would have thought it was a simple triac controller, like the normal speed controllers everyone else uses.
http://www.harborfreight.com/router-speed-control-43060.html
^ clearly its not.

chopper
http://www.freescale.com/training/classroom-training-events/motor-tutorial/phase-angle-control:WBT_MOTORPACTUT_WP

3 phase pwm
http://www.freescale.com/training/classroom-training-events/motor-tutorial/pulse-width-modulation:WBT_MOTORPWMTUT_WP

Does the fan not come with the controller? The fan is the newer EC fans vs typical AC fans. You wont be able to use the controller on them.

My fishtanks MP40 pump has 3 phase freescale chip inside. Freescale have a lot of good documentation explaining how to drive chips + motors.

[plantgrower]

At zero or low speed, commutation instant cannot be detected by means of back-EMF, so an open-loop arbitrary commutation is imposed (with a high current, and then with low efficiency), relying on the fact that the rotor will align to the phase having positive current. This could explain why speed is limited to >35%.

According to video reviews for this fan the 120v versions of this fan use approximately the following power  :-

The 6" version of this fan uses 12w @ 35%  48w @ 50%  55w @ 100%
The 8" version of this fan uses   9w @ 35%  55w @ 50%  85w @ 100%

The company that makes this fan is coming out with an automatic temperature controlled fan controller, if it is controlled by a potentiometer I presume that it would have to use a digital potentiometer. Which digital potentiometers would be adequate to control this fan ? Does anyone know of any examples or articles in using a digital potentiometer with a pid control algorithm ?