I would like to know if it is possible to replace the rotaty dimmer by a digital potentiometer controlled by i2c.
Yes - that should be possible, provided your digital pot has the same ohm range as the physical pot. The only possible problem might be that the load of the digipot and the load of the physical pot are different enough to upset the operation of the rest of the circuit in the dimmer - in which case it won't likely work right. But - if you think you can hook it up to try it, it is certainly worth a shot.
Before you attempt this, what kind of AC motor do you have? If its an induction or synchronous
motor then it simply may not be suitable for speed control at all (and your current method of
control could cause it to overheat or even catch fire).
Ok, but is there any digital potentiometer which can do that?
Because all digital potentiometer work at about 5-15VDC and in my case it is 220V AC...
Concerning my AC motor, i don't know, how can i know that?
Because the Motor/fan is in the chimney. But i can unmount the actual controller and take pictures (perhaps this controller is dedicated to a determined AC motor type)
Well it appears to be a standard triac circuit, which suggests the motor is a universal motor, or
perhaps a low power torque-motor, which can be power/speed controlled this way.
The point about induction motors is that they need voltage and frequency to be scaled together
to control speed, which requires a V/f inverter-based controller. Standard induction motors overheat
if abused. Synchronous motors depend on the line frequency as they are locked to it.
I'd check what circuit the controller is actually using - checkout the active device part number if
it has one to make sure its a triac fro instance.
Maybe a servo to turn the existing mechanical pot. Might be an electrically safer thing to do.
I thought that it could be a possibility but i'm trying to find a solution without mecanical parts.
I understand the digital potentiometer is not possible but i think some electronic solutions would be available...
Just a note re induction motors and variable speed:
An induction motor is inherently a fixed speed or near fixed speed machine when run at its nameplate speed, frequency and voltage. When driving a fan, however It can be (and has been) run at reduced variable voltage and nameplate frequency to get a reasonably wide range of speed control. This is possible because the fan load torque falls off rapidly as the speed is reduced and so the motor current is reduced enough to save it from burning up. Of course the power factor and efficiency are very poor in this mode.
In using it this way you almost have to do a case by case examination of the motor and fan speed-torque curves to make sure the motor does not burn up and the speed stays more or less under control.
This was used more in the past when nothing better was available; now there are several well designed alternatives so I see little reason to use it this way