Originally Posted by jeffs555
When you use hall sensors to control a motor, you are just replacing the mechanical commutation on a brushed motor with more efficient electronic commutation, so changing the voltage to the motor will change the speed just like on a brushed motor. Here is a page that tells what you should need to design a sensored controller. http://www.allegromicro.com/techpub2/hurst/bldcmot.pdf
I am new to RC but not to motor control and I am working on a project where efficiency at less than full power is the most important criteria. I hope I am not out of like reviving an old thread.
Has anyone tried or considered optically sensing the potion and speed of the motor on motors use on RC air craft and calibrating that instead of Hall effect sensors or unscrambling back EMF wave forms both which in addition to their obvious problems of not having a clear transition have even more when on the same board with the noise of FETs switching hundreds of watts of power on and off very fast.
I realize that optical sensing and making any real use of more accurate information you get requires a CPU that is good deal more expensive than the ones in use on controllers to day. And every installation o a problem of installing the optical sensors in a rather hostile environment. Cost is not much of a consideration on this project. With the current state of embedded computer chips it doesn't have to be that expensive. High motor efficiency at reduced speed is important and no current controller is as good as I think they could be if the rotor position was more accurately known and speed change were spread over several revolutions instead of controller that reacts by changing right now to what I believe to be motor position data that has some error in it.
I will have a better idea of the error in the back EMF controller when I get a recording digital oscilloscope on the motor and hold it at one speed at reduced power and see how stable the position signal is off the best controllers.