DC servos and ferrite rings: on the motor, on the encoder, or both?

[howiemnet]

I've bought several little DC servos from eBay over the years, some have the motor wires wound once or twice through a ferrite ring, some arrive with nothing on the motor wires, but the encoder wires are wound through a ring, some have neither.

My Teensys seem to cope quite well with running big motors with high-frequency encoders even if I don't put rings on the wires, but what *should* I be doing?

(see the way I squeezed Teensys into a slightly off-topic question )

 

[PaulStoffregen]

I like the Teensy reference, but really, posting good photos of the actual motors would probably help you more....

 

[howiemnet]

You ask and I obey: here's a quick selection out of the drawer:

OK, I guess the tiny one gets by without any ferrite rings, but the bigger ones often have a ring on one or t'other of their sets of wires. But there's no consistency to it.

If it helps, I tend to use higher frequency PWM (25kHz) to drive the smaller motors to keep the audible noise down, lower frequency (< 1kHz) for the bigger ones as it seems to slightly help the motors' fight against friction at low speeds.

Guess the general question I'm asking is: are the ferrite rings supposed to be on the wires generating the noise, or the ones you want to protect from it?

 

[jwmelvin]

When I built a large machine tool using brushless motors (grad school) we used ferries on the motor leads to reduce radiated noise. They were quite effective. I imagine, though, that ferrites on the signal wires may also help by suppressing the induced noise. One would think that they should be separate ferrites, right?