First of all do these motors differ electrically and if yes, how? Can you measure BLDC generator speed using back-emf independent of its waveform type?

  • \$\begingroup\$ The winding is done differently I believe. \$\endgroup\$ – hekete Jun 14 '19 at 13:10

”do these motors differ electrically and if yes, how? “

They (?) may differ according to power ratings and ability to handle harmonics of triangular currents but not necessarily. A square wave voltage commutated by the Bridge driver to load may result in more current generated and thus more torque load than a sine generated voltage but also create more Eddy current losses at higher RPM at max load.

e.g. Such as pulse charging a battery with a FET bridge will create broad spectrum harmonics due to narrow current pulse in the rectifier to low ESR battery capacitance.

”Can you measure BLDC generator speed using back-emf independent of its waveform type?”


If you consider only trapezoid and a sine waves, both can be easily sliced with a comparator to get a square wave to count RPM. The only difference is that a trapezoid or triangle is the integral of a pulse or square wave, if symmetrical.

The trapezoidal-current drive systems are popular because of the simplicity of their control circuits but suffer from a torque ripple problem during commutation. A sine controlled motor or linear loaded generator ( with active PFC) is smoother to control at low RPM.

This is how large VFC motors are controlled with a pseudo-sine wave.

worse than a trapezoid: ...

I can imagine a generator with V/f constant (rated at no load) being overloaded with excessive pulse noise before the FET diode bridge and possibly a voltage controlled LPF filter would be useful.

I don’t have a universal solution but I am sure you can find some that work for linear loads (rectified sine) and non-linear loads ( battery charge) or with MPT converters that vary load impedance or those with trapezoidal waveforms and the source impedance varies with the mechanical power source available. ( torque and RPM)

Signal conditioning requirements for suppressing commutation noise are necessary for generators into battery chargers and may require variable frequency low pass filters that rise with RPM depending on fundamental signal f at max RPM and harmonic noise rejection required at low RPM to prevent false sensing. In this case you would measure the current waveform and determine the signal conditioner requirements over a wide range of loads and RPM to define the design specs.


You do measure zero crossing of back EMF to get the frequency and then you calculate speed. The amplitude and form has no influence on measuring speed from back EMF.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.