PI control of a PMDC motor at low speeds

  • Thread starter Thread starter awaiting
  • Start date Start date
  • Tags Tags
    Control Motor Pi
AI Thread Summary
The discussion centers on challenges faced in implementing PI control for a PMDC motor at low speeds, specifically at 5 rad/s, where fluctuations in motion occur despite successful operation at higher speeds. The user suspects the issue may stem from the motor's commutation characteristics, given its 30 commutator blocks and 2 poles, which contribute to a rumbling effect at low speeds. Suggestions include testing the motor in open loop to isolate the problem and ensuring proper gain settings on the integrator within the control system. Increasing the PWM frequency is deemed unlikely to resolve the fluctuations, and mechanical solutions like adding a flywheel are considered but viewed as extreme. Overall, the noise from the commutator and potential control system limitations are identified as key factors affecting performance at low speeds.
awaiting
Messages
24
Reaction score
0
Hi everybody,

I have implemented PI control of a DC motor and it works properly at high speeds (50rad/s) but if I lower the reference speed to 5 rad/s, there are fluctuations in its motion. Initially, I thought it could be due to the limitation of my current sensor (LTS 6 np) but I added some load, so that current is at a higher value (around 30mA) but the fluctuations are still present. I tried to increase the frequency of the PWM but due to limitations in my board, could not test beyond 20KHz.

Will increasing the PWM frequency have any positive effect at low frequencies?

Your comments and suggestions are welcome.

Thanks,
 

Attachments

  • lts6-np.pdf
    lts6-np.pdf
    505.6 KB · Views: 338
  • cct.jpg
    cct.jpg
    41.2 KB · Views: 422
Engineering news on Phys.org
How many poles, commutator blocks per revolution?
 
It has 30 commutator blocks and 2 poles.
 
awaiting said:
Will increasing the PWM frequency have any positive effect at low frequencies?
No, I don't think so
 
awaiting said:
It has 30 commutator blocks and 2 poles.
At 5 rad/s, you've got a 15 Hz rumble that's just not going to be controlled with an analog PI.

Added comment on edit: You could iron it out mechanically with some monstrous fly wheel, but that's getting pretty desperate.
 
Thanks for the reply!

Could you please explain how did you calculate the 15Hz value. I would also like to try the flywheel option. I am using 1N5822 diode across both the sides of the h-bridge, do you know of any other diode that could be useful.
 
Thirty commutator contacts may be 15 windings on the rotor. As the motor rotates and switches from one to the next, you get a bit of a "jump" that's probably just too rough to iron out with the controller. I guarantee nothing. I'm making a "remote" guess at what might be giving you problems at low speeds.
 
Sounds like at low speeds the noise of the commutator is causing the rumbling effect. Why don't you try to run the motor open loop i.e. since the speed depends on the voltage applied, just start off by applying a very small voltage using a DC source and listen to the rumble and then increase the voltage for higher speeds. If you hear the rumble then it's the characteristic of the motor.

If not, I suspect you don't have enough gain on the integrator. Make sure that the pole is at s= 0 and not higher like s= 1kHz. Just a guess as I can't see any control schematic.
 
Back
Top