Preliminary Kraken DOA software reaction to phase of two counter-conducting plasma antennas

In this experiment, the PCB with 2 CCFLs conducting plasma electrons in opposite directions is fed to Kraken Direction Of Arrival software via 2 rtl-sdr V3 dongles wired to drive the clock of one from the clock of the other.

Each frame of the animation is one second real time. The only changes directly to the experiment during the run were to restart the Kraken DOA system when it lost synchronization between the dongles and was unable to recover synchronization.

Other changes in the environment involved me walking around the house preparing lunch and periodically checking to see if the Kraken DOA system needed to be restarted so data collection could continue.

NB: The Kraken DOA software stops logging data when it loses sync between the signals -- or at least that's what it's supposed to do. I don't know if there is time averaging or other accumulation of data beyond samples but I don't believe so: When used as a DOA system (as opposed to qualitative phase difference change detection system) there is a phone app that allows one to be driving around hunting for the source of the signal broadcasting at the configured frequency.

The large variations inferred for both magnitude and direction of arrival of signal are a mystery and clearly point to a problem with using this system, as presently configured, to detect changes in phase between plasma and metal antenna states of the PCB housing the CCFLs.