I had been working on the XTB-232 firmware to confirm it works with HomeSeer, and started to see a log of M13 commands. Comparing the HomeSeer log against the XTBM, the commands were actually the periodic G16 ON and OFF commands to refresh our hot water loop.
The other anomaly was the last batch of XTB-IIRs I had tested did not repeat my weak test command when I injected background noise. However, checking with the scope showed the incoming command was only about half the normal amplitude, and was below the detection threshold when the noise was injected.
Today while investigating why the XTB-232 was decoding the G16 ON and OFF as M13 (which is indicative of noise dominating on alternate half cycles), I discovered the cause was the Acer LCD monitor that replaced an aging 14” CRT monitor. Instead of plugging the new monitor into filtered power going to the computer UPS, I had just swapped the original power cord into the new monitor. That was plugged into an unfiltered power strip.
The new Acer LCD monitor was loading down the signal on this circuit to less than half the normal amplitude (as seen on the scope), and was apparently generating enough noise to corrupt my weak test signal, even though the noise was not strong enough to be detected by the XTBM.
With the Acer monitor plugged into the XPPF feeding the UPS, the XTB-232 now detects the G16 ON and OFF just fine. (Those incoming commands are also weak because they are leaking onto the isolated circuit that I use for X10 testing).
Bottom line, even I can overlook an “unfriendly device”.
Jeff