Hmm
This is tricky .. not what Bojan has said (totally agree with him), but the mish mash of information before us.
There are two circuit diagrams in two separate papers.
The first one (Radiation Properties of Pulsar Magnetospheres, 1995) shows a transmission line model (Fig #6). This is the crude model used by Healy & Peratt to 'emulate' the signals emitted by a pulsar.
The second circuit diagram (called Figure #7), in the 'Advances' paper, 1998 (I believe), shows a generic model of a space plasma problem .. in this case, the flow of Birkeland currents in the Earth's magnetosphere/ionosphere. I can see a variable resistor in this one .. and a big EMF (potential difference source) ... 10e4 to 10e5 volts.
Neither is what you'd call a relaxation oscillator (for reasons Bojan has made clear).
Scott's lecture on Youtube at the 8:30 mark refers to paper #1 (1995) above. In his lecture he mentions the reflected wave on a transmission line and then talks about how easy it is to recreate a relaxation oscillator in the lab. He highly recommends the 1995 paper. He says the paper provides simulations of "17 of the known properties of Pulsars". He flips between using the term "Relaxation Oscillator", and the transmission line model, with pulses injected into it.
The implication being that this is THE Relaxation Oscillator model.
If it is THE Relaxtion Model for Pulsars, then it does not demonstrate Relaxation Model characteristics, so I still don't understand the analogy, in modelling terms.
Just wanting to get this clear ..in spite of the poor written and lecture communications by Peratt, Healy and Scott.
Cheers & Rgds.
PS: In Scott's lecture, his slides say: "Stable frequencies in trapped-ion clocks are equal to the observed outputs from pulsars" .. and .. "Plasma transmission lines from one location on a star to another - or between members of a closely spaced binary pair can easily produce these oscillations".