RF shielding

The encoder wires on my Rover 5 seem a little short.  The ground wire seems to be a shield around the other three conductors.  Also, I notice that the pin groupings on my 4 channel motor controller have a ground.  Makes me think a little.

I'm about to order a whole bunch of three pin to separated single pin cables to accomplish the motor control and current sensing I'll be doing in my current project.  I also want to open up a bunch of servos and add a tap to the wiper.  While I'm at it maybe I should extend the cables coming out of those previously metioned encoders.

But those encoder wires make me stop and reconsider.  I'm going to be connecting a bunch of (cheap, probably noisy) servos, over long cables, to an Arduino Mega, and then, ultimately, a Red Back Spider Controller.  And someday, there will be an ARM or Atom based SBC in there.

I haven't been giving much thought to those ground pins.  But I'm sure that Oddbot, in his wisdom, put them there for a reason.

SO, I'm thinking that perhaps I should get a couple dozen feet of 3 conductor shielded cable, (I'm leaning towards 22 gauge.) Some pins and 4 pin, 3 pin, and 1 pin housings and make a bunch of custom length shielded cables to do all my device to device wiring with.  (Including extended servo cables, with an extra wire for position sensing.)

But I'm a little concerned about the nebulous concept of "ground loops" and my lazy self is poking me in the ribs trying to convice me that it's not necessary.  And where IS my crimper anyway?

Still. . .there are those shielded encoder cables.

Anyone care to chime in about chasing noise issues in your own projects?  I'm looking for a little advice on rf noise control.

Seconded, caps near noisy

Seconded, caps near noisy things like servos will cut the noise they spit out.

I only use shielded cables for analog lines and reasonably high-speed comms - the shield on the encoder lines will help prevent false triggers, so I’d suggest keeping it.