Motor speed PWM torque boost at low RPMs

[throbscottle]

I started thinking about DC motor speed PWM controllers. Simple PWM still loses torque at low RPM, and so I was wondering is there's an adaptive version that can sync with the motor's brush contact cycle, or maybe provide more voltage at lower speeds. Does anything like this exist?

 

[Papabravo]

Generally, torque in a DC motor will be proportional to current. It will be a maximum at low speed. Voltage will be related to the time rate of change in current. The torque will be unaffected by a change in voltage.

 

[rjenkinsgb]

The unit needs to be closed loop - speed feedback at a minimum, and preferably current loop control.

That allows full torque down to very low speeds; it's how machine tool servo drives worked prior to brushless motors being introduced.

The speed feedback can come from armature voltage sampling, for the back EMF - but that's not very accurate as it is somewhat load dependant.
A DC tachometer is used for high accuracy and can give better than 100:1 speed control range.

From a quick search, PWM with back EMF feedback is often used for model railway speed control.
That in turn brings up this - a design in Silicon Chip magazine, in an old issue backed up on archive.org

It is a big file and takes a few minutes to download.. See page 19.

https://ia902509.us.archive.org/17/items/silicon-chip/Silicon%20Chip/2008/Silicon%20Chip%202008%2009.pdf

That covers the details of how such a system works.


An industrial drive typically has two separate feedback loops. The motor current is controlled by the inner loop which compares motor current with a current setpoint signal, with the setpoint to that from the speed loop which compares the speed feedback signal to the commanded speed signal.

Both loops use some combination of PID elements.

 

[ZipZapOuch]

The torque will be unaffected by a change in voltage.

Ohms law applies to motors. Raise the voltage, the current through the coil increases, the magnetic field increases, the torque increases.

 

[Papabravo]

Ohms law applies to motors. Raise the voltage, the current through the coil increases, the magnetic field increases, the torque increases.

Only if the source can supply the required current. Power supplies have limits and you can't increase the output voltage without limit. Motors run on current, not voltage. It matters not how much voltage comes with the current used. If I have two power supplies, a 20V @ 4A and a 40V @4A, either one will provide the same torque to a motor. One supply has 80 watts of power output and the other has 160 watts of power output. The voltage is irrelevant when it comes to torque production.

 

[ZipZapOuch]

Only if the source can supply the required current. Power supplies have limits and you can't increase the output voltage without limit. Motors run on current, not voltage. It matters not how much voltage comes with the current used. If I have two power supplies, a 20V @ 4A and a 40V @4A, either one will provide the same torque to a motor. One supply has 80 watts of power output and the other has 160 watts of power output. The voltage is irrelevant when it comes to torque production.

You sound like you're grasping at straws to make your ohm's law-violating explanation sound true. It is not true. And you know it because you bring in a red-herring topic of a power supply not being able to deliver additional current. If the power supply isn't able to deliver the current at a given "voltage setting", then the power supply cannot deliver the "output voltage" for a given load - in the end, the torque is a function of the input voltage (but as you said, not necessarily the input setting - but nobody expects a power supply to deliver more current than it is designed for, right? (Because ohms law is always true).

In the end, raising the voltage actually put into a motor increases the torque by the cascade of logic I used above.

Raise the voltage, the current through the coil increases, the magnetic field increases, the torque increases.

Similarly, decreasing voltage applied to a motor lowers the torque of a more. Right?

 

[MrAl]

I started thinking about DC motor speed PWM controllers. Simple PWM still loses torque at low RPM, and so I was wondering is there's an adaptive version that can sync with the motor's brush contact cycle, or maybe provide more voltage at lower speeds. Does anything like this exist?

Hi,

As others have said, to get high torque at low speed you have to use a closed loop speed regulator. What that does is keeps the speed relatively constant as you dig into the work harder. That means high torque at low speed.

There are limits of course, the most notable is the max torque of the motor. You can not get higher than that except maybe for short time periods, but you do risk harming the brushes.

Also, you do not want the speed to be perfectly constant. You want the user to notice a 'slight' speed decrease so they get a feel for how the tool is reacting. If you do not do that, the tool will stall in the work and it will stall all of a sudden with no warning. That's not desirable. Better is to let the motor speed decrease a little so the operator can judge when they are applying too much pressure. If it is an automatic machine it may be different. You may want very good speed regulation.

I actually did this about 40 years ago with an old Dremel tool. Inside I found the motor ran on DC, so I was able to set up a back EMF type of speed regulation. It works very well although like I said, you only get the max torque of the motor not any higher. That still made the tool much nicer to use because I liked working with low speeds.

 

[MrAl]

Only if the source can supply the required current. Power supplies have limits and you can't increase the output voltage without limit. Motors run on current, not voltage. It matters not how much voltage comes with the current used. If I have two power supplies, a 20V @ 4A and a 40V @4A, either one will provide the same torque to a motor. One supply has 80 watts of power output and the other has 160 watts of power output. The voltage is irrelevant when it comes to torque production.

Hi there,

I am not sure that is an accurate account of how a motor works, although I have a feeling you just did not mention everything about the motor you were running.

For example, a 160 watt motor will produce more torque than an 80 watt motor if they were both build with the same technology, and if different, I would think either one would produce more torque, more speed, or more of both.

When they advertise certain motorized tools they often mention the wattage in order to attract attention to the power of the tool.

However, I am guessing that you were running a 20v motor only or maybe something less than that like 12v. If you run a 12v motor at 12v you get the max torque and that's about the size of it. If you use a 100v power supply you have to set it to 12v anyway.

 

[schmitt trigger]

Is the motor series-wound, shunt-wound or permanent magnet?

A series wound motor automatically increases its torque with load.

 

[throbscottle]

I'm back. Been busy! How did this get on to power tool motors? Ah well, joys of the internet!

I've discovered another aspect of my problem - that of how well the PWM driver works. It's only a small, slowish motor from a printer or something, but using a proper driver (ie, not one made by me!) A4950 in this case, makes a massive difference. So it's got enough juice now and good braking (too good, in fact).

rjenkinsgb - that's an inspirational article (actually the whole magazine is interesting!). I wonder if I can implement that inertia control in hardware alone (since my PWM source is a 555 timer)? It would only need a very brief slowdown period. Hmmm...

 

[MaxHeadRoom78]

The unit needs to be closed loop - speed feedback at a minimum, and preferably current loop control.

That allows full torque down to very low speeds; it's how machine tool servo drives worked prior to brushless motors being introduced.

An industrial drive typically has two separate feedback loops. The motor current is controlled by the inner loop which compares motor current with a current setpoint signal, with the setpoint to that from the speed loop which compares the speed feedback signal to the commanded speed signal.

Both loops use some combination of PID elements.

The BLDC/AC versions still use closed loop.
There is now now longer two loops needed, inner and outer.
Torque mode drives just need the encoder to the PID processor controller.
The old systems used both drive and PID controller, now the drives can be simple transconductance, non-intelligent torque mode drives.
All the control is in the one PID loop.
I use the Galil PC boards this way.
Easy to mix and match drives and motors.

 

[Pommie]

I like to think of motors as,
Voltage (including back EMF) = speed,
Current = torque
Hence when speed = zero (stalled) current (and torque) is at a maximum.
When no load the speed will increase until Volts applied-back EMF is almost zero and hence current is also very low.
This also explains why motors are often specified as having a kV or how many 1000 revs it will develop for each volt applied.

This helps me better understand them.

Mike.
Edit, this also means that ohms law does apply but back EMF has to be taken into consideration.