Energy recovery from PWM DC Motor ?

Thread Starter

daba1955

Joined Apr 27, 2019
181
On another thread I have shown the reverse EMF produced by a DC motor running on PWM speed control.

I showed traces of the voltage across the motor terminals with a standard flywheel or flyback diode, a Schottky diode, and with no diode at all. That's the trace I want to discuss, so I'll post it here as well ...

2019-05-27_081028.jpg

The project I am working on is battery powered, and a design brief was to make it as efficient as possible, extending battery life.

The thought has occurred to me that it ought to be possible to capture the energy in that reverse EMF (when the motor switches off), and put it back into the battery, instead of just throwing it away (presumably as heat into the motor windings).

Does anyone know if that is at all possible, and suggest a way to do it ....

TIA
 

DickCappels

Joined Aug 21, 2008
10,153
This sounds like it may be related in a loose way to regenerative braking on electric cars, but if you are only looking at recovering the tiny bit of energy stored in the inductor, that won't be much compared to what you would get when you recover the energy stored in the load. Also faintly resembles damping iwht energy recovery which can be useful in high powered applications in which the current to the inductor is continuously switched on and off. What kind of application are you looking at?
 

Thread Starter

daba1955

Joined Apr 27, 2019
181
This sounds like it may be related in a loose way to regenerative braking on electric cars, but if you are only looking at recovering the tiny bit of energy stored in the inductor, that won't be much compared to what you would get when you recover the energy stored in the load. Also faintly resembles damping iwht energy recovery which can be useful in high powered applications in which the current to the inductor is continuously switched on and off. What kind of application are you looking at?
Without going into much detail, because the application is yet to be patented, the load on the motor is an unbalanced weight, which produces vibration. The trace is telling me there's 25V there, surely that alone couldn't be generated by an unloaded motor, or could it ? I suppose I could take the load off it and see if I still get as much. If I can get a mere 5 to 10% of extended running time, it will help. It does seem a shame to fritter the energy away if it could be recovered somehow. Remember this is PWM, so it is happening on every pulse ....
 

Sensacell

Joined Jun 19, 2012
3,432
Chances are if it's clamped with diodes back to the supply, the current IS flowing back to the energy source.
Just the inductive energy, not the back EMF from the mechanical motion of the rotor.

But we'll never know, due to the proprietary nature of your super-secret about to be patented device.
 

Thread Starter

daba1955

Joined Apr 27, 2019
181
Chances are if it's clamped with diodes back to the supply, the current IS flowing back to the energy source.
Just the inductive energy, not the back EMF from the mechanical motion of the rotor.

But we'll never know, due to the proprietary nature of your super-secret about to be patented device.
Apologies for keeping the actual application secretive, but I am working for someone else, I'm not at liberty to disclose too much, sorry. Let's just say the final product needs to vibrate at varying frequencies and leave it at that for now.

I'm now thinking if we try to regain the energy in the spinning "flywheel", it's got to slow it down, and it's going to need more energy to accelerate it back to the natural speed of the motor, so we might not gain much at all. In fact we might be in danger of the motor over-heating, and the surges on the battery might not be too good for it.
 

Alec_t

Joined Sep 17, 2013
14,280
it ought to be possible to capture the energy in that reverse EMF (when the motor switches off), and put it back into the battery
Don't mistake voltage (emf) for energy. If the back-emf is reduced then the motor draws more current to maintain its speed. If you extract energy from the motor then the motor slows down (aka regenerative braking).
 

Thread Starter

daba1955

Joined Apr 27, 2019
181
Don't mistake voltage (emf) for energy. If the back-emf is reduced then the motor draws more current to maintain its speed. If you extract energy from the motor then the motor slows down (aka regenerative braking).
Not one of my brightest ideas then, we'll leave it as is (with the Schottky diode, not as above).

2019-05-27_072139.jpg
 

avayan

Joined Oct 30, 2015
38
This sounds a lot like regenerative braking where instead of letting the motor coast down, you switch the power stage so the motor current is dumped into the battery. This happens pretty much whenever the motor's BEMF is larger than the applied voltage, and the right current recirculation mode is employed.

In an H Bridge, the PWM duty cycle represents an effective voltage being applied to the motor. For example, if you have a 12V power source, and the H Bridge is running at 50% duty cycle, then the motor sees an equivalence of 6V. When the motor is speeding up, the BEMF is increasing to this command voltage. At this point in time you would expect to see the motor's BEMF at nearly 6V (ideally, it would be 6V, but due to losses, it will be less).

If you now command the H Bridge to 25% duty cycle (equivalent to 3V), then the motor will behave like a generator because its BEMF is larger than the effective applied voltage. Whatever current it has will now go somewhere and this depends on the current recirculation mode. If you use fast decay, then it goes into the power supply. If you use slow decay mode, then it recirculates through the FETs.
 

Thread Starter

daba1955

Joined Apr 27, 2019
181
This sounds a lot like regenerative braking where instead of letting the motor coast down, you switch the power stage so the motor current is dumped into the battery. This happens pretty much whenever the motor's BEMF is larger than the applied voltage, and the right current recirculation mode is employed.

In an H Bridge, the PWM duty cycle represents an effective voltage being applied to the motor. For example, if you have a 12V power source, and the H Bridge is running at 50% duty cycle, then the motor sees an equivalence of 6V. When the motor is speeding up, the BEMF is increasing to this command voltage. At this point in time you would expect to see the motor's BEMF at nearly 6V (ideally, it would be 6V, but due to losses, it will be less).

If you now command the H Bridge to 25% duty cycle (equivalent to 3V), then the motor will behave like a generator because its BEMF is larger than the effective applied voltage. Whatever current it has will now go somewhere and this depends on the current recirculation mode. If you use fast decay, then it goes into the power supply. If you use slow decay mode, then it recirculates through the FETs.
This is not an H=Bridge driver, it is simply a single FET (IRFZ44N) pulling the motor neg. to ground, motor pos. is on V+
 

avayan

Joined Oct 30, 2015
38
I am afraid you won't be able to do any kind of battery charging with a FET and a diode. When the FET disables, the motor current recirculates through the diode, so there is no way to dump it into the battery. You asked for a mechanism to charge the battery. The simplest is an H Bridge where you can control the PWM and the current recirculation mechanism.
 

crutschow

Joined Mar 14, 2008
34,280
As has been noted, any energy you extract from the motor during the OFF part of the cycle will have to be replaced during the ON time.
The diode just allows the motor current to keep flowing during the OFF time.
So you actually want to keep the current flowing during the OFF time with minimum loss, not extract energy from it.
For best efficiency, you want to use a Schottky diode, or even a switched MOSFET (active rectifier) across the motor to carry the OFF time current with a minimum of loss.

The switched MOSFET would be turned ON during the OFF period.
The MOSFET can have a drop of a few tens of mV versus several hundred mV for a Schottky diode.
The energy you would save for a MOSFET compared to the diode would roughly be the motor current times the OFF time divided by the ON time, times the diode voltage drop.
Don't know whether the added complexity of the active rectifier would be worth that power savings.
 
Last edited:

ronv

Joined Nov 12, 2008
3,770
As has been noted, any energy you extract from the motor during the OFF part of the cycle will have to be replaced during the ON time.
The diode just allows the motor current to keep flowing during the OFF time.
So you actually want to keep the current flowing during the OFF time with minimum loss, not extract energy from it.
For best efficiency, you want to use a Schottky diode, or even a switched MOSFET (active rectifier) across the motor to carry the OFF time current with a minimum of loss.

The switched MOSFET would be turned ON during the OFF period.
The MOSFET can have a drop of a few tens of mV versus several hundred mV for a Schottky diode.
The energy you would save for a MOSFET compared to the diode would roughly be the motor current times the OFF time divided by the ON time, times the diode voltage drop.
Don't know whether the added complexity of the active rectifier would be worth that power savings.
The active clamp allows regenerative braking.
If he doesn't want regen, no need.
 
Top