Jon,
Tom's experience with plasma TV's supports my theory of the coupling
mechanism. He found that common mode impedance was not very effective.
Theory: A possible equivalent circuit for the plasma TV EMI is an RF source,
a capacitor to ground at one end, and a wire to ground at the other end. We
have a loop with 30 pF in series. The CM (Common Mode) current along the wire
radiates 1.8 MHz energy. The 'wire' is the AC power ground wire and/or the TV
signal cable. The capacitor is the plasma TV screen capacitance to ground.
Let's say it is 30 pF to ground. That is -j3000k ohms at 1.8 MHz. And let's
say the wire to ground is 0 ohms. The loop impedance is -j3000 ohms. Now let's
say we add a 100 uH CM inductor in the wire-to-ground end. The CM inductor is
+j1000 ohms. The loop impedance is now -j2000 ohms. We have just increased the
CM current and the EMI. Before we see any improvement the CM inductor
impedance must be over 6 k ohms. We must go to 9 k ohms before we see a 6 dB
reduction in the CM current. To obtain +j9 k ohms we need 800 uH. Using a
ferrite
toroid having a mu of 500 and a length of 1" we need to loop the 'wire' through
it
about 9 times. That seems doable. With a +9 k ohm inductor we are probably
around parallel resonance, so that may help. But 6 dB has reduced the EMI by
just one S-unit.
But it is worth a try. I would start by disconnecting all cables, except of
course the power cable, from the TV. Connect a battery operated source of video
if that makes the TV happy. If the problem still exists then attack the power
cord with a big, high-mu ferrite core. Using multiple cores taped together
will help. Two cores give 2X the inductance of one core. When the EMI is
lowered
to an acceptable level connect the TV signal/video cable. If the EMI comes
up, attack it the same way. Looping the power cable and the video cable through
the same ferrite is the best from an EMI standpoint but might be difficult to
do.
If this helps but does not lower the EMI to an acceptable level then we need
to resort to unusual methods. At that point I can build you an active CM
"choke." I think we can shoot for a reduction in CM current of 20 dB using
this.
Maybe more. I can build it up and get it working if you'd like to try it.
Dave WX7G
_______________________________________________
Topband mailing list
Topband@contesting.com
http://lists.contesting.com/mailman/listinfo/topband
|