> -----Original Message-----
> From: Jon Ogden
> I have found the entire system to be very layout dependent with respect to
> susceptibility to my transmissions. If I position things one way, my RF
> doesn't trigger the switching circuitry. Yet, move things slightly and
> then
> it does. I haven't been able to figure out any rhyme or reason to this
> either.
>
When trying to cure some TCLs that were generating HF noise, I also
found that wiring configuration played a part in the resonance, and
therefore, strength of the noise. I could pick up these lamps a block away
from the owners house, and when I plugged them in down in my basement they
also made a lot of noise there. But oddly, when plugged them into one
outlet in my kitchen, placing it even closer to my antenna than ever, it had
absolutely NO NOISE!
The AC house wiring does indeed provide resonant transmission lines
and act as antennas for these devices since they do have an AC circuitry
component. I found one lamp that was making noise a block away, and when I
located an extension cord in their garage emiitting the noise, unplugging
the cord made it go away completely. So it's not only the AC house wiring
but anything plugged into the outlets as well that sets up these resonances.
> Secondly, I tried once again experimenting wrapping the sensing line
> around
> one of my torroids. This time I used a larger (about 1.5" dia) 77
> material
> torroid and wrapped about 16 turns of the sensing wire around it. It
> helped
> with eliminating the RFI, but again, it was very dependent on where the
> sensing wire was in relation to the switching unit, etc. But here is the
> interesting part. With the torroid now wrapped, it is now acting as a
> pickup for the sensing wire. I can now touch the torroid and the
> switching
> unit will do its thing. All the wire on the sensing line is insulated so
> I
> am not touching bare wire. The touch wire seems very sensitive.
>
Even coils have what is called distributed capacitance between all
the wires, but it is usually too small to matter; however, in this case even
the smallest changes in capacitance are "supposed to" trigger the lamp.
That's the design. Try to go with very small coils with thin wires to
reduce the amount of capacitance. This might be why the recommendations are
usually for 1-10 microHenries and not 100 milliH coils like Radio Shack
sells. Think "thin wires" & "fewer turns" to reduce the distributed
capacitances of the coil.
> I am
> wondering if using a small diameter piece of coax instead of the sensing
> line will help things. I could ground the shield and attach the center
> conductor to whatever is the touch surface.
>
Couldn't hurt to try, but be aware that coaxial cable has
capacitance and might introduce new problems. But put the choke as close to
the control circuit box as possible. Then you still have the problem of the
"unshielded" hood. Maybe a small metallic plate somewhere else would make
better sense and reduce the "pick-up factor"?
I also just got some advice from a ham friend who is also an
oscillator
expert. He suggested putting an LC parallel resonant circuit at 14
MHz in
series with the sense line.
That would reduce the currents in the line, and increase the
impedance. Could be tricky since you'll need to keep the C low so as not to
interfere with the sensing lines function (change of C equals change of
state). I think it'll also need to be a high Q circuit (Q>10) with a low DC
resistance (you mentioned it quits working when a resistor is added).
So far I've not had any luck with resistors and a large size RF
choke (the
one sold by Radio Shack I forget the size).
I think it's 100 milliH. Way too big as discussed earlier. Go with
something in the microH range.
My friend says that one reason
why a choke may not work is that the choke has capacitance between
the
windings. This capacitance then acts as a pickup for the touch
sensing line
and actually counteracts the effect of the choke. Perhaps this is
why a big
size choke hasn't worked for me but maybe a smaller one
will.............
I think it has more to do with a change in capacitance. The
distributed C is quite small so anything that changes the C even a tiny bit
(like your hand being nearby) can have a large delta effect on C.
73,
de ed -K0iL
--
FAQ on WWW: http://www.contesting.com/FAQ/rfi
Submissions: rfi@contesting.com
Administrative requests: rfi-REQUEST@contesting.com
Questions: owner-rfi@contesting.com
|