DPRG List  

[DPRG] grounding motor cases

Subject: [DPRG] grounding motor cases
From: Bill Boyer daweasel at swbell.net
Date: Tue Jul 5 16:59:09 CDT 2005

Possibly use shielded cable to run the motor power with the shield grounded 
only on the motor control end?

If you have a metal chassis and the motors are bolted to it, you may wind up 
with a ground loop the chassis is already ground and you run a ground wire 
to the motor.

Best solution may be to put the sonar stuff in a metal box and only run 
digital out of the box.

----- Original Message ----- 

>> I've been chasing noise related to the new Polaroid sonar
>> modules on the 6-wheel robot, and it appears that I can
>> reduce it substantially by placing a grounded shield
>> between the motors and the sonar controller boards.
>> I had not noticed any noise problems before, but it
>> seems that the high gain analog amplifiers in the sonar
>> controllers are particularly sensitive.
> That is exactly the place the RF noise would show up. You are amplifying a
> small signal and small noise the same amount.  In regular digital circuits
> the noise is small relative to the digital signal, so it is not noticed.
>> By powering the sonars from a separate bench supply,
>> I was able to narrow the noise problems in the sonar
>> to the 24 DC motors that are mounted nearby.  Placing
>> a sheet of aluminum between the sonars and the motors
>> had no effect, until I ran a clip lead to the aluminium
>> sheet and grounded it to the robot power supply.  At that
>> point, the noise disappeared almost completely.  RF?
> A shield needs to be grounded to be effective. Think of it as the signal 
> can
> transmit through metal a lot easier than in air. By grounding it, you are
> not allowing any signal to transmit through it because any potential
> different than ground gets sucked to ground.
>> I had previously added the recommended trio of noise-
>> suppressing .1 uf caps on the motor termimals and
>> between the terminals and motor case for each motor,
>> but I noticed that the motor case itself is not
>> grounded.  When I added a ground wire to the motor
>> case as a test, it seemed to make the noise problems
>> _worse_ rather than better, and I removed it.
> That seems weird to me.  Did you ground the case to the battery terminal
> directly, or through the chassis, or other current carrying wire?  I would
> guess you need to ground the motor case directly using a non-current
> carrying wire.
> I would remove the capacitors between the motor power terminals and the
> motor case.  They are acting like high pass filters (letting high 
> frequency
> noise go out). Then you have high frequency noise travel down the ground
> wire (which has some inductance) radiating like an antenna.
>> So my question to the experts in this group is, do you
>> routinely ground the motor cases of your robot motors?
> Usually the motors are grounded because the chassis is grounded. It is 
> like
> a big wire with very low inductance. This does not work for plastic 
> chassis.

More information about the DPRG mailing list