PWM a DC Motor

[SteveMeisel]

I am working on a project in which I need to PWM a small DC permenant magnet motor. Is there a way to calculate the optimum frequency to get a decent torque out of it. I need to be able to fine tune it's angular positon. I am having trouble with the torque at low rpm's. Motor's spec's are .5mH @ 1Khz and 1.3 ohms.

Thanks,

Steve

 

[Skogsgurra]

The armature time constant is .5 mH divided into 1.3 ohms. That is around .4 milliseconds. You need a carrier frequency that is several times faster than 1/.4 ms or 2500 Hz if you need to keep current ripple down. I would go for 10 kHz. If that is a problem, you can add an inductance in the armature circuit. But I would not do that, it makes control more demanding (more voltage needed for fast response).

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[coconutalley]

Another issue that you may have: PM DC motors typically have high rpm to operate optimally. Gear reducers are used to get more torque at low rpm.

 

[OperaHouse]

"I need to be able to fine tune it's angular position."

Sometimes the PWM rate is dropped much lower to provide a bumping like a dither to overcome static friction. This can also create other problems.

 

[SteveMeisel]

Thanks all.
The only source that I can find on the web that seems to be a good source is the Seattle robotics club. They metioned about the time constant also. I was hoping that there would be a better way as I am limited to 10KHz max. I tried @ 9KHz but I had to get the duty cycle up to overcome the static fricion. A low duty cycle @ 150 Hz seems to work ok but I cannot get a fine control. We may need to backup and take another look at our design. Thanks again.

 

[coconutalley]

All the pwm does is effectively average the voltage across the motor. I think you are confusing duty cycle with frequency modulation. Pulse width modulation is varying the width of the ON pulses (duty) but keep the frequency constant (or period constant). If you are lowering the frequency to 150 Hz, you are frequency modulating which is not a good way to control dc motor. It works for AC induction motor.

To get more torque you need to up the current in windings which means upping the average voltage. But, as mentioned earlier, if the motor is not designed for low rpm [like most dc motors], it would still be hard to get fine control. You'll need gear reduction to scale down the "linear range" of the motor. As a ficticious example, a DC motor may be controllable between say 2000 rpm to 4000 rpm but outside this window it would not be linear with voltage.

Alternative might be a brushless dc motor but controller would be more complex.

 

[MikeHalloran]

Before all the brass hats went to digital controls, model railroad power supplies included "pulse power" that made them capable of very fine control, i.e. slow speed without noticeable stick-slip. They weren't expensive enough to be real fancy inside. I suggest you buy one and measure its output behavior under load.



Mike Halloran
Pembroke Pines, FL, USA

 

[felixc]

It this is going to get into a commercial product, make sure that the noise does not become an issue. I've had product returns because of a PWM controlled DC motor that was operating at 7.5 KHz. The high-pitch was driving the people crazy. Rising it up to the 15KHz range stopped the complaints.

 

[MikeHalloran]

I used to be able to hear the ~15kHz of a TV's horizontal oscillator. It was very annoying, and they were everywhere.

They seem to have gotten quieter in recent decades. ;-)



Mike Halloran
Pembroke Pines, FL, USA

 

[SteveMeisel]

I was only trying to state that I was using differant frequencies to try to obtain a good torque at a low rpm. I understand fully what PWM is, how and why it works. What I found was at a low freq (150 Hz) a 20% duty cycle had some torque but the rpm was to high, and at 9KHz I had to go to 60% duty cycle just to overcome the static friction. Though it did have some pretty cool harmonics comming out of it.
I was hoping to be able to slow the motor as I reached the targeted position to prevent overshoot. Then maintaining the position I may have to bump the motor one direction or the other. A gear reduction may be the best way to go.

 

[MikeHalloran]

What happens at 10 Hz?



Mike Halloran
Pembroke Pines, FL, USA

 

[SteveMeisel]

H-bridge data sheet states that the max PWM frequency is 10khz.

 

[itsmoked]

Mike it's very hard to hear that from your Flat Panel LCD.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[SteveMeisel]

Thanks Mike.
Should have read your post closer.
Maybe there is a software solution. This will make the design guys happy. I hadn't thought about going to that extreme.

Steve

 

[Skogsgurra]

"I hadn't thought about going to that extreme"

Know what you mean. There's nothing as hard as software...

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[sreid]

Do you have a "Free Wheeling Diode" across the motor?

 

[Skogsgurra]

Can't have that if you are using an H bridge since it outputs a bipolar voltage. Diodes are normally incorporated in the bridge. Usually integral to the switches.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[sreid]

Sorry, I missed where Steve said "H-Bridge" in a recent post.

Is there a lot of friction-stiction in the system you are driving?