I am doing this from memory so bear with me...
For some good explanation why motor drivers use fast decay mode and slow decay mode and mixed decay mode see these great application notes from ST to explain about steppermotors and from texas instruments to explain about dc brushed motors.
As written in post #49, the user of a steppermotor really wants the energy stored in the coil /inductor during a step, and still present in te coil / inductor to be removed. This energy present itself as current still stored as magnetic field and creating backEMF and you want to drain the magnetic field by draining the current. That is where the decay modes are for.
Why drain the current ? often, you have to reverse polarity for the next step and when there is still current flowing in the wrong direction , you will not be able to drain the current during the step.
The current you expected to flow will not flow and therefore the magnetic field strength you hoped to generate does not appear and you will not be able to generate proper torque or speed.
So, take that current out soldier !
In essence you short circuit the coil (which is just an inductor) for a moment or you apply the opposite voltage for a very short moment.
This puts quite a load on your much needed electrolytic low impedance buffer capacitors. So you have to make sure that you take the the maximum ripple current into account and do not forget about it or your electrolytic low impedance buffer capacitors will turn into POP ! corn.
And these electrolytic low impedance buffer capacitors are also great to absorb and level any load dump alike situations when properly chosen.
When you have chosen your electrolytic low impedance buffer capacitors correctly, you can also put a few in parallel but you have to take the pcb traces into account because of the inductance of the copper traces, therefore , make a low impedance and low inductance copper plane.
The fun part with a copper plane is that when you have a ground plane and above it a powerplane, that you get free capacitance also for high frequencies, not ideal but better than inductance.
In practice you always get what is called parasitic capacitance and parasitic inductance and of course normal resistance which for you as designer may also act parasitic.
Parasitic capacitance and parasitic inductance are in the pF, range respectively the nH range and together this can act as resonance tanks and oscillate at very high frequency.
For a example what can happen if you do not take this into account : Which may be a cause why the circuit is emitting EMI at a too high level to pass any EMC and EMI certification.
But that is where good pcb design comes in and at the right places EMI ferrite beads and ceramic capacitors ideal to be relatively low impedance for very high frequencies in the MHz range up to the GHz range.
This document, see page 16 :
With a dc brushed motor it is in essence the same. But you are using PWM here and you want to take out the current which is momentarily stored as a magnetic field that will present itself and the back EMF before the next pwm pulse.
In essence you short circuit the motor for a moment or you apply the opposite voltage for a very short moment.
Why do this all ?
Well, most of the time you do not have to do this, the logic inside the Hbridge dc brushed motor driver or the steppermotor driver takes care of this all. But you must be able to understand why it does do fast decay or slow decay. So you end up with a strong and speedy motor.
With steppermotors there is of course the hold current which is often a fraction of the rated current during stepping.
When you hold a steppermotor in a position , in essence the coil/inductor becomes together with an intelligent steppermotor driver, a sort of SMPS (switch mode power supply) that regulates on the hold current through the coil/inductor from the steppermotor. The holdcurrent looks on an oscilloscope, like a current which you can see as a DC current but with an AC triangle current placed on top of it. That triangle shaped current is the result of the current regulation of the steppermotor driver.
With dc brushless motors, this is the same. These motors have 3 phases, but the idea is similar like a steppermotor.
I know some theory but to be honest, i never had the chance to work with brushless motors yet. So no practical experience but as mentioned above, similar to steppermotors but different amount of phases, different amount of poles, a bit different construction and some more differences.