Speed of braking methods for PMSMs in VFDs

Question

I am trying to understand the strength of the following braking methods available on VFDs and how they rank in stopping/deceleration speeds:

1. Coast to Stop
2. Ramp to Stop
3. DC Injection Braking
4. Short-Circuit Braking
5. Dynamic Braking with suitable Braking resistor

*Assume that in all the above mentioned stopping method cases, the same load, motor and VFD is used with the rated braking resistor to dissipate all the regenerated energy.

My concerns are the following:

1. Due to the presence of flyback diodes on VFD output circuit, the generated current is directly transferred to the main dc capacitor on the VFD while coasting to stop. So would coasting to stop happen at the same speed as dynamic braking. If not, why?

   • Is current usually limited (maybe with a resistor) through flyback diode? What if this diode can pass through all the current generated by the motor?
   • During deceleration, when overvoltage fault occurs on the drive and turns off output transistors, wouldn't current still flow through the flyback diodes and charge the capacitor further?
   • Or is there a switch which is not usually shown in the drive schematics which cuts the connection (even from flyback diodes) to the capacitor when drive is stopped with fault or in Coast to Stop method?

2. Due to boosting effect with flyback diodes, with dynamic braking dissipating the excess energy in the capacitor, wouldn't transfer of all the energy out of the motor into the main dc capacitor happen at the same rate as short circuit braking and decelerate the motor in the same time? If not why?

3. With a suitable dynamic braking resistor present, Is it possible to configure Ramp to Stop deceleration to happen faster than short-circuit braking and dc injection braking?

   • Can we do this by turning the transistors ON, such that (FOC quadrature current)/torque is applied in the opposite direction to the motor's speed direction?

Answer 1

Great Questions. For your first part on ranking braking methods that is difficult to say because it is dependent on the particulars of the system in question. You can say with confidence that coasting to a stop will be the least powerful since there is no braking.

I also want to point out that "ramp to a stop" doesn't actually say anything about the braking method just that the VFD is controlling the speed ramp down profile.

1. When pairing a Motor and VFD the rated voltages of the motor, VFD and available source are considered. Motor's are rated so that when the applied voltage meets or exceeds the motor rating the motor's back EMF will remain lower than the voltage rating up to the listed rated speed. This means that when running a PMSM with a VFD and suddenly turning off modulation the voltage that develops at the motor windings due to the rotor magnets spinning around will be less then the voltage on the VFD's bus. Since the bus is at a higher voltage no current will flow through the flyback diodes.

Notice this assumes that when you stop modulating the motor the load coasts down to a lower speed. Let's assume you have a load that actually speeds up, say a crane's brake fails and a heavy load spins the motor in excess of its rated speed. Once the Back EMF exceeds the Bus voltage and the flyback diodes voltage drop you are right current will flow from the motor into the VFD's bus! This current will increase the VFD's bus voltage and if there is nowhere to put that energy the bus capacitors could fail.

This is why a "braking resistor" is included in so many VFD applications, that current (and really kinetic energy from the load) can now be safely unloaded from the bus capacitors and dissipated as heat.

2. The energy dissipation is not in the bus capacitors. Well designed VFD's minimize the amount of energy the Capacitor's dissipate. When the VFD is expected to provide braking other components must be added to the system to put the kinetic energy somewhere safely. The two most common methods are using a braking resistor, or another inverter device typically called an Active Front End (AFE) that can put the mechanical energy back onto the grid. Whatever method of braking is used it's the rate of energy transfer (power) through these braking components at sets the braking power of the VFD.

3. Again it is not the method of braking employed so much as the power ratings of the components that handle the braking energy.

Hope this helps.

Amendment based on your comment

You seeing the bus jump up in voltage on a real system reminded me of the other energy your VFD is managing and that is the magnetic fields in the motor itself. You are absolutely right that once the VFD stops modulating there is an initial "jump" in the DC bus voltage. This is from the inductance of the motor, maintaining a stater current while the magnetic fields in the motor collapse. You are correct that while this current flows it will flow through the flyback diodes.

On systems I've worked with the VFD's were sized such that this magnetic energy could be readily absorbed by the DC bus without throwing a fault. But when I would look at a trace of the bus voltage it would briefly bump up when we would shutoff modulation on a running motor. Our VFD's were 200/400Vac so were built with internal discharge resistors to meet DC bus discharge requirements. So that bump of energy would dissipate into those resistors and the VFD's bus voltage would normalize back to whatever the supply voltage would rectify too.

Answer 2

A VFD can be programmed to do pretty much anything the designers could think of, within the rules of physics, and the components available in the particular VFD.

There are only four places to dump the energy stored in a rotating machine as it's brought to rest - as heat in the machine itself or a dump resistor in the VFD, as extra voltage on a DC bus capacitor, or returned to the grid. All of these will have power and/or total energy limits, which will limit the braking torque or the size of the motor that can be braked.

Consult the datasheet for your particular VFD.