[RFI] Touch light RFI Update - Final

Jon Ogden jon.ogden@cain-forlaw.com
Mon, 19 Nov 2001 10:12:20 -0600

Hi all,

Well, it's been several months since I've brought up the subject of the RFI
situation with my touch lighting system that is
part of my new kitchen cabinets.

I learned a bit in this about how those buggers work and what you can do to
help prevent transmissions from messing with them.

Here is a summary:

Touchlights:  Manufactured by Hera Lighting  (www.heralighting.com)

1.) I do NOT have noise problems on my receiver due to the lighting system.
No buzzes, squawks, etc.

2.) The problems I was having was due to RF getting into the touch dimmer
circuitry and stepping the lights through various levels as my signal
modulated.  This is with full legal limit on a tri-band yagi mounted nearly
directly above the kitchen.  Height above the kitchen is probably about 25
feet.  So it's a tough environment!

3.) Solutions:

This was the hard part.  The lighting system consists of a dimmer control
unit which houses the sensing circuitry.  The output of that goes to a low
voltage transformer unit and then from there to the lights.  Lots of cord in
all this which could be a problem.  On most bands, the solution that worked
best was to wrap an Amidon FT-77 large torroid around the power cord going
into the dimmer unit.  This fixed a lot of the problem.  However, on 10
Meters, that did not do the trick.  I still had problems.  Then I went to
focusing on the touch sensor line and getting the RF off that.  You cannot
just put bypass capacitors on the sensor line.  The sensor works by sensing
a change in capacitance due to touch.  Putting bypass capacitors on it
swamps it out.  Several people suggested putting an inductor and resistor in
the line.  This helps.  But too much inductance actually makes the problem
worse since the windings of the inductor act as an antenna.  Too much
resistance while eliminating the RFI prevents proper operation.  So it's a
real balancing act.

No the most important part of all this was my discovery of the fact that
what the touch line is connected to is probably the most important factor in
eliminating RFI.  The worst case was the lighting that was installed in the
fan hood.  The fan hood is made of stainless steel.  The orginal
configuration had the entire fan hood as a "touch" surface.  The touch line
was attached to it and you could touch anywhere on the hood to change the
lighting.  This acted as a huge antenna, obviously.  But the interesting
thing is this:  The large metal surface also begins to look more and more
like a ground to the touch system.  Therefore, the less sensitive it becomes
to capacitance changes from touching it.  If I added enough resistance to
the sense line to get rid of the RFI, the touch dimming wouldn't work at
all.  In fact, even adding a couple hundred Ohms resistance would cause the
touch sensing to screw up and still would not prevent the RFI.  So you have
two oppposite forces acting here.  One is that the large metal surface acts
as a good RF antenna.  However, that surface also begins to look more and
more like ground to the VCO circuitry attached to the sensing line and it
becomes less sensitive to pertubations in capacitance.

The fan hood has a glass shield which has a couple of stainless steel trim
pieces.  I ended up isolating those trim pieces from the rest of the hood
with clear plastic tape.  I then attached the sense wires to these trim
pieces (cannot solder to good stainless steel!  So it's a mechanical bond
only).  No I have a much smaller metal area.  Less size for an antenna and I
could then load the sense line down with resistance to prevent the RFI.  

I duplicated this in other areas too.  Under the cabinets the dimming is
done by touching a piece of aluminum trim.  Same story.  Having too much
trim connected to the touch line made it more susceptible to the RFI
interference, but made it less sensitive to human touch when loaded
sufficiently with resistance.

So it was a real balancing game to find the right size of area that can be
touched and the right resistance to put in the sense line.  Inductance
didn't seem to do much, but I left some in there anyhow.  No idea how big
the chokes were since they were from a bulk pack of various chokes I bought
at Radio Shack.

The interesting thing is that the sensitivity of the sense lines would
change depending on how dry your skin was, whether or not you were grounded,
etc.  Under one cabinet that had the largest amount of metal originally
connected to the sense line, I had it resistively loaded not to cause RFI.
However, sometimes touching it would change the lights, sometimes it
wouldn't.  Sometimes you'd have to have your shoes off and standing on the
bare floor, other times it wouldn't matter.  It was very interesting.  If I
lowered the resistance at that point, then the RFI would affect it.   The
trim on this set of cabinets consisted of 2 pieces that were electrically
connected with a short wire.  My solution was to electrically isolate the
trim pieces and just use one of them as a sensing area.  Then the problem
was solved.

In most cases I used somewhere between 5K Ohms and 22 KOhms depending on the
size of the metal I connected to.  The larger metal pieces needed lower
values of resistance.  The smaller ones could use a greater resistance on
the sense line.

Bottom line:  Things work pretty well.  Occasionally when transmitting, I
get a light come on, but that's the exception - not the norm.  Of course,
when we have a good thunderstorm the same thing happens.  One day I came
home from work after a huge thunderstorm and all the lights were on!  So
it's at a level that can certainly be lived with!

Once I get my tower put up with the antenna farther away from the kitchen,
that should help immensely as well.

OK, so there's my diatribe on the subject.  It's been an interesting



Jon Ogden
Sales Engineer
Cain-Forlaw Company
847-202-9898 (Voice)
847-202-9896 (Fax)