X10 module frequency response

[nobody@xxxxxxxxxxxx (Dave Houston)]

Several months back Dan Lanciani posted about a mystery signal in the 200kHz
neighborhood that was blocking X-10 operation but that did not show up on an
ESM1 meter.

I finally got around to running some tests. I used the PWM output of a
PIC12F683 running at 8MHz (internal oscillator) measuring the frequency
using my Digital Sampling Oscilloscope card. The signal level on the
powerline was about 300mV. The signal waveshape was a little more triangular
than sinusoidal.

LM14A and LM465 lamp modules were set to the same address and plugged into
the same powerstrip (where the signal was injected). My pre-ELK ESM1 was
also plugged into this powerstrip.

The PWM signal was continuous and I tested by sending ON/OFF commands via an
RR501. The results are below. A YES means the module still worked (i.e. the
injected continuous signal did not block operation) while a NO means the
module operation was blocked. The ESM1 indicated a signal (one bar which
would be ~2 bars on an ELK made ESM1) between 69.9kHz and 168.7kHz.

I would expect an X-10 made switch would perform similar to the LM465 but I
have no switches so could not verify this. The LM14A performed much better
than the LM465. I have no idea how switches and modules from other
manufacturers would perform.

FREQ (kHz)     LM14A      LM465
225.1           YES        YES
202.7           YES         NO
184.4           YES         NO
168.7            NO         NO        ESM1
-----            --         --        ----
101.4            NO         NO        ESM1
 96.5            NO         NO        ESM1
 92.5            NO         NO        ESM1
 88.1           YES         NO        ESM1
 84.5           YES         NO        ESM1
 81.1           YES         NO        ESM1
 77.9           YES         NO        ESM1
 75.0           YES         NO        ESM1
 72.4           YES         NO        ESM1
 69.9           YES        YES        ESM1

I ran the series of tests twice and they repeated. A higher signal level
might block over a wider range.