Model Forum / Radio Controlled / Air Models / December 2006

The noise reduction capacitor is normally determined simply by the
function of the noise to be reduced and has no bearing on whether it
is used for RC or just general noise reduction.

Here's a fun experiment for you. Run an unprotected and unfiltered
motor next to an AM radio tuned to a dead spot. By dead, I mean no
station. There will always  be a faint hiss in a dead spot. Run your
motor near the radio and listen to the noise it makes in terms of
radio frequency noise, not the whine of the motor itself but what
you hear in the speaker as a result of the motor being on.

Now add a ceramic capacitor of the value .01 microfarad rated at 50
volts or 100 volts across the motor terminals. Smaller is better on
noise capacitors and be warned that you should never use a polarized
capacitor (electrolytic or tantalum are examples of polarized
capacitors) on a motor. Make a note of the difference in noise
levels. Now try the same with a .1 microfarad of the same rating and
see if you can tell the difference. You can do that experiment with
hundreds of values of capacitors until the cows come home and the
results should not differ significantly no matter what value used.
Thus, the de facto standards of .01 or .1 are the most often
selected for the convenience and price. They are used in billions of
noise reducing operations and are found everywhere they sell
electronic components in bags of 100+ for pennies each.

As mentioned earlier, the spark is a noise generator but that can't
be stopped or eliminated by a capacitor no matter how big or how
powerful. The reason is that the spark is a necessary byproduct of a
motor's brushes and emits its own field. Marconi knew this and the
very first radio transmissions across the Atlantic were just noise
from gigantic sparks. This brush noise fact of life led to the
development of the brushless motor which uses pure induction to move
the rotor. It's the effects of the collapsing induction field that
the noise capacitor reduces in both cases, not spark noise. As the
motor rotates away from the point where the electromotive power is
applied, the field in the coils collapses rapidly and can emit a
field of energy around the coil as it does. That capacitor acts as a
shunt for this field and effectively kills it allowing your radio
some level of safety from the motor.

To fully reduce the motor noise as much as is possible, the easiest
method is to have a good ground throughout your electronic motor
subsystem and ensure that your motor case is grounded within that
system. A screw from the case and a wire to the negative terminal of
your battery is normally all that is required. Also, you may want to
cover the gaps that you have in the motor where you can see the
brushes contacting the motor armature. You can add a Faraday type
shield which will cut the emissions a great deal. To find a good and
nearly free one, look for large cables that have ground mesh inside
them. Remove the mesh, expand it so that plenty of air can pass
through and slip it over your motor so that the mesh covers the
holes through which the brushes are visible. Be sure to ground this
mesh and secure it with a clamp of a plastic wire tie and you'll
have reduced your noise as much as is possible without spending a
fortune.

Other tips would be to use ferrite cores and put a single loop of
the positive motor wire through the loop. Here's a large file link
that describes the technical nature of what a ferrite core can do
for you.
http://www.fair-rite.com/newfair/pdf/15th_Edition_Fair-Rite_catalog_Cable_and_Co
nnector.pdf#page=4

To shorten that explanation to a single sentence, the wire you run
through that loop is an inductor and the ferrite core makes it a
minimum of 25 times as efficient as a noise reducer with just a
single loop. They're very light in weight as only a tiny one is
needed.

Disclaimer: I realize that this is a very technical issue but I
tried to simplify it for anyone to read. The engineers among us will
realize that I did not go into the level of explanation that would
confuse the hobbiest. I did that on purpose so if you want to
discuss theory in greater depth, do it without me.
--
Ray

between 10nF andf 100nF plate type capacitors are ideal. Don't use
tubular..they have significant series inductance. Normally ceramic is best.