I presume this is a brushed DC motor you are driving
It is, yes.
in which the waveform looks perfectly plausible.
The regular spikey bits are the commutator switching transients, the undulations in the back EMF from the
armature are due to the geometry of the motor.
The 0V segments are when the PWM is on, the high segments are the inductive kick-back taking the
bridge into fast-decay mode at the end of an on period. Only after fast-decay is finished does the
back EMF signal become visible (and the switching transients which presumably are inter-winding
currents being interrupted as the brush passes over a gap in the commutator)
I see, however - the very same motor produced the following waveform when being driven by an LMD18200 device, as opposed to the VNH2SP30. Admittedly, one is an analog scope - the other is digital, not sure if that should make a difference?
I drove another larger motor today that draws roughly twice the current, and got a very similar waveform.
...and if I stall the motor, I get a very squarewave like waveform.
If I enable the recorded noise band on my digital scope, it looks like there is a serious amount of noise being thrown around - now, I'm not sure if this is the sort of noise you'd visibly see on an analog scope? So everything I've seen in the past may have had these noise components, I've just never seen them.
You might try spinning the motor by hand while looking at its output on the scope.
Perhaps I can't quite spin it long enough or fast enough by hand...
So, any thoughts on; a) why the different waveforms with the same motor, at the same frequency between two different devices? The only thing I can think of, is the LMD18200 was being run in locked anti-phase PWM mode... and b) regarding noise, how to reduce this? I've got a HV 10uF cap across the motor terminals, and 220uF at the 12V supply to the IC.