Three Phase Synthesis for brushless DC motors
LoopyByteloose
Posts: 12,537
Brushess DC motors have taken over r/c airplanes as they output more power than gas motors.
These little demons use 3-phase that is constructed from a micro-controller.· While a PIC is often used, I would prefer to figure out a way to do it with the SX.· The Microchip PIC site has application docs supporting their own technology in larger, more powerful motors, but I seem to be looking at a few pieces that are missing.
1. Conversition of a PWM input (from the reciever) to a speed control code.
2. An all software three 120 degree out of phase PWM output that is contoled by the r/c and possibly motor feedback.
3. A constant torque at variable speed (via mixing the output voltage and the output frequence) calculation
Since this would be an open loop speed controller, it may be best to measure the back EMF of the motor.· I am not sure how to do that either.
Advertisers have other nice features, such as.
1. low battery detection
2. peak output control
3. specific battery profiles
and more.
But, I would just really like to get the basic concept and not all those goodies.
It seems to me that the brushless motor's hefty power output could be equally useful to robotics if it were geared down.· We always seem to be talking about too much weight when we discuss batteries or motors.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
G. Herzog in Taiwan
Post Edited (Kramer) : 4/17/2005 3:51:19 PM GMT
These little demons use 3-phase that is constructed from a micro-controller.· While a PIC is often used, I would prefer to figure out a way to do it with the SX.· The Microchip PIC site has application docs supporting their own technology in larger, more powerful motors, but I seem to be looking at a few pieces that are missing.
1. Conversition of a PWM input (from the reciever) to a speed control code.
2. An all software three 120 degree out of phase PWM output that is contoled by the r/c and possibly motor feedback.
3. A constant torque at variable speed (via mixing the output voltage and the output frequence) calculation
Since this would be an open loop speed controller, it may be best to measure the back EMF of the motor.· I am not sure how to do that either.
Advertisers have other nice features, such as.
1. low battery detection
2. peak output control
3. specific battery profiles
and more.
But, I would just really like to get the basic concept and not all those goodies.
It seems to me that the brushless motor's hefty power output could be equally useful to robotics if it were geared down.· We always seem to be talking about too much weight when we discuss batteries or motors.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
G. Herzog in Taiwan
Post Edited (Kramer) : 4/17/2005 3:51:19 PM GMT
Comments
Back EMF measurement using the zero cross method is fairly well explained, you monitor the EMF field of an inactive phase·of the motor by sensing the voltage across the winding since the permanent magnet rotor will act like a generator in proximity to the undriven phase. When the voltage switches polarity you know where the rotor's position is and you derive you control signals from that.
They mention this technique does not work for low RPM (and stationary motors) so you may need to determine a start sequence to get the thing moving before you turn over control to the controller.
Post Edited (Paul Baker) : 4/17/2005 12:36:52 AM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
G. Herzog in Taiwan
Take a peek at my contest entry "SX Hardware Prototype Board" for some ideas on generating quality sine waves. For generating three phases you simply need to run three (instead of the two I have shown) generators all linked to the same frequency time tick, but staggered 120 degrees. So there will need to be a counter divisible by three.
Should be real easy to adapt.
Have fun,
Peter (pjv)
Just My two centavos.
Paul