Brushed DC motor encoder oscillations

[greg2835]

I purchased a Pittman brushed DC motor with a 500 count/rev encoder. I noticed quite a bit of oscillation in the encoder signal when the shaft is rotating at a constant rate. After running some tests I started to notice a trend. For instance, when I ran the motor at an average 600 RPM (averaged from the encoder oscillations), or 10 rev/s I measured a 10 Hz oscillation in the encoder signal. Running at an average 1200 RPM, or 20 rev/s I measured a 20 Hz oscillation in the encoder signal. I performed this test at various speeds and always found the oscillations to equal the motor's frequency. I tried running the motor at a very low rate (~1 Hz) and can actually see/hear the motor slow down at a certain point once every revolution.

I am speculating that perhaps there is a commutation issue with one of the brushes. I am curious if anyone else has some experience with this and if they have any suggestions. I ultimately want to use the encoder for feedback control and would like to solve the issue itself rather than just adding filters. Any recommendations would be greatly appreciated.

:LabGuy330

 

[djs]

Monitor the armature current with an oscope. If there is any variation in speed due to loading or commutator issues it will show up there. If not then you may have a bad encoder.

 

[sreid]

Optical encoders have a once per revolution position error due to eccentricity of the encoder disk. Disks in small motors often have higher eccentricities.

 

[greg2835]

Thank you both for your responses. I in fact did measure the current to the armature and saw oscillations. This morning I contacted Pittman motors and spoke with one of there engineers. He informed me that these oscillations are due to cogging (unknown to me before) due to an iron armature. Apparently all Pittman motors experience this cogging, though they try to minimize it as much as possible. He did recommend other companies that sell ironless armature (zero-cogg) motors: Maxon and Micro Mo. According to him these motors will not exhibit the oscillations I am seeing. If anyone has some experience with such coggless motors I am all ears. Would really like some input before I make another purchase. Many thanks.

:LabGuy330

 

[MikeHalloran]

The MicroMo motors have ironless cores, and no physical structures that could cause cogging. They don't cog. They may cost a bit more than the Pittman motors.

You may be able to improve the performance of your system even with the Pittman motors by adjusting the phasing between the encoder and the motor's commutator.



Mike Halloran
Pembroke Pines, FL, USA

 

[IRstuff]

How are you controlling the motor, using the encoder tach output? If so, then the eccentricity of the encoder will affect the motor control. You possibly need to try and run the motor open loop or with a different tach to see if it is the motor by itself.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[greg2835]

Valid question. I am running the motor in open loop and just monitoring the encoder readings. My plan was to step test the motor to empirically find the mechanical time constant, and then I noticed the trend in encoder oscillations. Encoder eccentricity is also something new to me and I will look much more into this as well. Thank you for everyone's insights.

:LabGuy330

 

[mikekilroy]

"running open loop"

what exactly does that mean?

your motor does NOT - note - NOT - have a once per rev cogging - it is impossible.

SO. AGAIN, how are you running this motor where is the voltage and current coming from?

please do not jump to some conclusion about cogging when your data makes that not a possible cause.

 

[mikekilroy]

i suppose I got too excited in hitting submit button wo an edit button. My factual statement that you cannot have a 1/rev cogging was in answer to your and other suggestions that it was caused by the motor design. Know that a motor has an even no of poles so you MUST have at least 2 'cogs' per mechanical rev if it is caused by the motor design. buying an ironless motor may fix your problem but not because this one has a 1/rev cog due to iron in its rotor.

 

[greg2835]

The motor is hooked up to a power supply. When I said 'running open loop' I simply meant the encoder is not being used in feedback, just for visual monitoring.

No conclusions have been made. This is a discussion. I only relayed the information I received after calling Pittman motors. I am still very open to the advice from others.

:LabGuy330

 

[djs]

Maybe the armature is out of round? Off center?