PWM mode - 'fast pwm', 'phase correct pwm'

Hello all. I've been reading details on the internet about these 'modes' of pwm. For driving dc motors, I've been using the mode called 'fast pwm'. I'm sticking to 'fast pwm' (regular mode) for my projects right now. Normally there's no need to consider 'phase correct pwm' modes right?

Has anybody ever needed to use 'phase correct' mode?

Actually, 'phase correct' doesn't even sound right. The documents are saying that the middle of the pwm pulses (along the time axis) remain evenly spaced. So when the pulse width changes, this is no longer phase correct ...... more like 'phase incorrect', right?

Eg. the leading edge (eg. rising edge for regular pwm - non-inverted mode) of the next pwm pulse (after a change in duty cycle) is no longer in-phase with leading edge of the currently occurring pulse. They probably should have stuck with 'symmetric mode' or something. The fast pwm mode should probably be called 'phase correct' instead, or just 'regular mode'.

Phase is not related to the edge of a pulse, but to an invariant point of the base frequency. The maximum amplitude will always reside in the middle of the pulse, and can be used for the point of reference for the computation or measurement of phase differences.

You'll understand the difference between the modes if you compare two or more PWM signals of different and varying duty cycle (the Mega allows for 3 PWM signals from each timer).

DrDiettrich:
Phase is not related to the edge of a pulse, but to an invariant point of the base frequency. The maximum amplitude will always reside in the middle of the pulse, and can be used for the point of reference for the computation or measurement of phase differences.

You'll understand the difference between the modes if you compare two or more PWM signals of different and varying duty cycle (the Mega allows for 3 PWM signals from each timer).

Thanks DrDiettrich! I was thinking that the leading edge of a regular 'fast pwm' pulse would provide the correct phase reference - relative to some starting point in time. This is assuming that the base pwm frequency remains the same.

I'll read up more on those modes, and do those comparisons you recommended. Thanks again!

[update - looks like we just have to interpret the 'phase correct pwm' label as just a name - that simply refers to the mode where the middle part of one pulse and the middle part of an immediate neighbouring pulse (regardless of pulse width) is always spaced the same distance apart (in time). Maybe helpful to not think in terms of phase..... maybe call it mid-pulse-aligned, or symmetric mode or something. Alternatively, if - for some reason - a pwm generator is generating several individual pwm waveforms, each having their own individual duty cycles, and if the phasing needs to be purposely made such that the middle of every pulse of every waveform are all aligned, then this could help with understanding the origins of the meaning of 'phase correct pwm'. But....on the other hand, if the same thing was considered for regular pwm mode, then one could consider all leading edges of all pulses to be aligned --- all synchronised.... locked in phase.]

Southpark:
Thanks DrDiettrich! I was thinking that the leading edge of a regular 'fast pwm' pulse would provide the correct phase reference - relative to some starting point in time. This is assuming that the base pwm frequency remains the same.

I'll read up more on those modes, and do those comparisons you recommended. Thanks again!

[update - looks like we just have to interpret the 'phase correct pwm' label as just a name - that simply refers to the mode where the middle part of one pulse and the middle part of an immediate neighbouring pulse (regardless of pulse width) is always spaced the same distance apart (in time). Maybe helpful to not think in terms of phase..... maybe call it mid-pulse-aligned, or symmetric mode or something. Alternatively, if - for some reason - a pwm generator is generating several individual pwm waveforms, each having their own individual duty cycles, and if the phasing needs to be purposely made such that the middle of every pulse of every waveform are all aligned, then this could help with understanding the origins of the meaning of 'phase correct pwm'. But....on the other hand, if the same thing was considered for regular pwm mode, then one could consider all leading edges of all pulses to be aligned --- all synchronised.... locked in phase.]

Or it could be as simple as a phrase created by a marketing department to differentiate their product from the competitions product.

Paul

Paul_KD7HB:
Or it could be as simple as a phrase created by a marketing department to differentiate their product from the competitions product.

Paul

Could well be Paul. Definitely paulsible! If they managed to come up with something as corny-sounding as 'fast pwm', then it's not surprising that they could conjure up 'phase correct pwm'.