> "Those pointy corners look more typical trochoid than "typical lissajous". What calculation artifact corresponds to a line wheeled-over to form a trochoid loop? Not sure "hysteresis" explains this best."
I'm creating that phase-space waveform myself through a solution to a differential equation so know exactly what is causing the shape, which is a sample-and-hold lag integrator with a delta spike that is incommensurate with the clock frequency.
This is not difficult, especially considering that this is the kind of stuff that shouldn't be a problem to reverse engineer -- since the time variable is missing from the phase plot.
Consider this guy asking what's causing a similar pattern on stack exchange
Everyone tries to be so helpful, but the questioner should realize that he could figure it out himself if he left the time-scale in his data acquisition and plotted y(t) vs t (i.e. potential) and dy/dt vs t (i.e. current) side-by-side. Then he would see it is just a square wave, with likely some noise on the top.