STEORN DEMO LIVE & STREAM in Dublin, December 15th, 10 AM

[Omnibus]

Now, here's the story with the experimental Vr (absence of theoretical Vr is already established):

The screen shot I posted earlier (cited by @blueplanet) is taken with the current probe and the shunt measurement being done simultaneously. This has never been the case in the OU measurements but was done here just for comparison. It turns out, however, that the presence of the passive probe measuring the voltage drop across the 10Ohm resistor (there's also a 0.001uF capacitor in series with the 10Ohm resistor; I'm mentioning this as just a reminder) affects the current in the circuit. This is confirmed by the simultaneous measurement with the current probe:

Data when passive probe is measuring current across shunt:

Current probe: 30.4ppV 0.106ppA -> 286.8Ohm
Passive probe: 30.4ppV 0.126ppA -> 241.3Ohm

Now, here are the data when current is measured with the current probe without the passive probe and when current is measured with the passive probe without the current probe:

Current probe: 30.4ppV 0.126ppA -> 241.3Ohms
Passive probe: 30.4ppV 0.127ppA ->  239.4Ohms

So, current measured over the shunt is unaffected by the presence of the current probe but the current measured with the current probe is greatly affected by the presence of the passive probe. Of course, as I said, when actual measurements are done there's never a passive probe attached to measure current simultaneously with the current probe. So, this is a non-issue regarding the present studies. However, this has to to had in mind in the future had such necessity (to have a passive probe simultaneously measuring current with a current probe).

Also, as you can see, the Vin is unaffected either way, so we have to be pretty confident at this point that the voltage measurements are correct and are not the cause of the observed OU effect. So, back to the phase shift issue. What is causing the additional phase shift (additional to what a capacitance induces) which causes the experimental OU to be greater than the theoretical OU?

[blueplanet]

The operating frequency shown in your graph is in the neighborhood of 700 Khz (T=3.5*400ns), which is the frequency at which your "OU" occurs.

As I mentioned in one of my previous post, one of the criteria for OU to occur is abs(Vin) < abs(Vr). But this conditiion did not occur at 700 Khz. What went wrong?

And obviously, according to your graph, Ein, which is the input energy, does not appear to be constant or negative. Have I missed anything?

Now, here's the story with the experimental Vr (absence of theoretical Vr is already established):

The screen shot I posted earlier (cited by @blueplanet) is taken with the current probe and the shunt measurement being done simultaneously. This has never been the case in the OU measurements but was done here just for comparison. It turns out, however, that the presence of the passive probe measuring the voltage drop across the 10Ohm resistor (there's also a 0.001uF capacitor in series with the 10Ohm resistor; I'm mentioning this as just a reminder) affects the current in the circuit. This is confirmed by the simultaneous measurement with the current probe:
...

[Omnibus]

The operating frequency shown in your graph is in the neighborhood of 700 Khz (T=3.5*400ns), which is the frequency at which your "OU" occurs.

As I mentioned in one of my previous post, one of the criteria for OU to occur is abs(Vin) < abs(Vr). But this conditiion did not occur at 700 Khz. What went wrong?

And obviously, according to your graph, Ein, which is the input energy, does not appear to be constant or negative. Have I missed anything?

The criterion which you give for OU to appear is incorrect. The correct criterion is slope(int_0^T IVdt vs. t) < slope(int_0^T I*2Vdt vs. t). This correct criterion is implemented in the numerous Excel files and figures therefrom.

Please, don't dilute the discussion with incorrect assumptions and talk about problems that have already been solved.

[Omnibus]

As we've see, the energies properly integrated yield OU. I emphasize, proper integration (provided all the data is correct). Proper integration and not using Vm, rms and the like.

[Omnibus]

What we need to do now is find the expressions for the integrals int_0^T IVdt and int_0^T I*2Vdt and then find the first derivatives over time of these expressions. Many of us, myself included, are so spoiled by the numerical methods that the real good old math is put slightly on the side (it's true that numerical methods are very powerful and can dwell in territories inaccessible by the 'good old math' as far as engineering is concerned).