Thanks Graham, OK, I think I see what is happening here
The current algorithm tries to 'learn' that this is your transponder using a set of heuristics about power, and relative height.
The data you collected is very good, I think I can use this to try to refine the algorithm.
I also need to add more logging, because it would appear that your transponder signal (received by PilotAware) is below the threshold
defined as identifying your own transponder.
In order to diagnose this I need additional logging
Thx
Lee