This is for Poynty
Thanks for the efforts here. But you've not given us a replication. I don't think that PSpice can manage it - quite frankly - unless and until it can trace the battery voltage as a direct product of the current through the shunt resistor. It really needs to show that second order phase shift as you term it. Also. See if you can do something where the resonance ramps up rather than down as it progresses.
Poynty. It bothers me that you're so easily satisfied with an explanation. Your reasons are wrong. Transistor Zener very much in tact. And there's no variation if and when we put the probes directly on the battery terminals. And you're rather keen on using your standard DMM. If you put the setting to AC it will show that negative battery voltage. But it never exceeds zero to the extent that you've shown it. Look again at our traces. The battery voltage 'flirts' with zero at best - and only now and then does it actually breach that level - and then only by a fraction.
I keep trying to post the link. It's not taking it. Here's the best I can manage.
PS - it seems that I've again exposed my 'amateurish' status. LOL. MH - take it as read that this is the fact. However, to clarify that AC setting of your standard DMM - here's what's meant. Poynty insists that we do not need those sophisticated DSO's. I needed to cater to his preference. (another edit) As well as putting the DSO probes closer to the battery terminals - I also used a standard DMM on the battery, putting these probes directly on the battery terminals. And on a DC setting it's rock solid at a given value. On an AC setting, interestingly, it moves all over the place - even into negative - but never to the extent of those voltages that PSpice shows.