Rosemary Ainslie Quantum Magazine Circuit COP > 17 Claims

[picowatt]

PW that is true in the DC case and if they used short wires without the inductance that I mentioned then the resulting common source configuration does not oscillate.  That holds to 10uH.  At 100uH  the circuit oscillates immediately.  During start-up, peak drain voltages of 800V are present, settling to somewhat more pedestrian peak voltages of 600V at about 600kHz.   At 600kHz, 100uH is about 377 Ohms. 

Ms. Ainslie has never understood the operation of the function generator and its internal resistance as it pertains to gate bias as well a DC current path for Q2 MOSFET current.  They got their oscillations, made bad measurements, drew incorrect conclusions and were happy until under supervision they made half way decent measurements and found out that their over unity measurements were all the illusions that Poynt99, and TinselKoala had been telling them they were.

My understanding is that despite showing June 29 that Figure 3, Figure 6, and Figure 7 of Paper 1 were all the result of probing errors, Ms. Ainslie has withdrawn her retraction because she has some new strange ideas about how current can accumulate or deplete on a node in violation of KCL.  She already endorses those silly self-powered socket strip videos.

MarkE,

Are you discussing a scenario based on sim work?

This subject has been so beat to death, I am amazed anyone still bothers.

Even with a lot of inductance, if Q2 is switched on hard at DC, her load will see more current than her linearly biased Q2 oscillations ever produced (and more efficiently I might add).

I have always considered Q2 to be configured as common gate, do I understand that you consider Q2 to be configured as common source?

PW

[MarkE]

MarkE,

Are you discussing a scenario based on sim work?

This subject has been so beat to death, I am amazed anyone still bothers.

Even with a lot of inductance, if Q2 is switched on hard at DC, her load will see more current than her linearly biased Q2 oscillations ever produced (and more efficiently I might add).

I have always considered Q2 to be configured as common gate, do I understand that you consider Q2 to be configured as common source?

PW

PW yes I simulated to confirm that the circuit oscillates stably with 100uH and 20mOhm in series with the composite Q2 source.  I got an extensive configurable SPICE model of the circuit from Steve.

If the Q2 MOSFETs don't oscillate then four fully enhanced IRFPG50s have one fourth the series resistance of one and a tiny fraction of the average resistance seen when they oscillate.  In the low impedance battery configuration, the required inductance for oscillation is many times greater than with a 50 Ohm resistance as provided by the function generator.

The Q2 configuration used with the function generator is a common gate, because the gate is connected to a low impedance whereas the source and drain are not.  Replacing the function generator with something that has a low impedance like another lead acid battery arguably changes the circuit such that it behaves as either a common gate or common source configuration since both terminals are from an AC standpoint grounded.  Inserting sufficient inductance that the source is again decoupled by an impedance much higher than the gate puts us back in a common gate configuration at frequencies high enough for the inductive reactance to matter.  The simple diagram above does not show any impedance inserted between the source terminal and the Vee supply, so as drawn it works as either common gate or common source.

[picowatt]

PW yes I simulated to confirm that the circuit oscillates stably with 100uH and 20mOhm in series with the composite Q2 source.  I got an extensive configurable SPICE model of the circuit from Steve.

If the Q2 MOSFETs don't oscillate then four fully enhanced IRFPG50s have one fourth the series resistance of one and a tiny fraction of the average resistance seen when they oscillate.  In the low impedance battery configuration, the required inductance for oscillation is many times greater than with a 50 Ohm resistance as provided by the function generator.

The Q2 configuration used with the function generator is a common gate, because the gate is connected to a low impedance whereas the source and drain are not.  Replacing the function generator with something that has a low impedance like another lead acid battery arguably changes the circuit such that it behaves as either a common gate or common source configuration since both terminals are from an AC standpoint grounded.  Inserting sufficient inductance that the source is again decoupled by an impedance much higher than the gate puts us back in a common gate configuration at frequencies high enough for the inductive reactance to matter.  The simple diagram above does not show any impedance inserted between the source terminal and the Vee supply, so as drawn it works as either common gate or common source.

Yes, after rereading your previous post, I realized you were referring to the case where the Q2 source had a low impedance path as being common source.

As you say, the configuration type can at times become a bit gray, but when in doubt, I choose the least dynamic terminal as the "common".

I read your write up over at Mark D's website.  It was a very well written and concise summation of this whole affair.  In concert with what Mark D posted (and her earlier retraction), I would have thought all of this would have been put to bed by now.

PW

[MarkE]

Yes, after rereading your previous post, I realized you were referring to the case where the Q2 source had a low impedance path as being common source.

As you say, the configuration type can at times become a bit gray, but when in doubt, I choose the least dynamic terminal as the "common".

I read your write up over at Mark D's website.  It was a very well written and concise summation of this whole affair.  In concert with what Mark D posted (and her earlier retraction), I would have thought all of this would have been put to bed by now.

PW

I am a big fan of common gate / common base circuits.  A bit of trivia that you probably know is that the first transistor circuits were common base configurations. 

I wrote a detailed report on the Quantum Magazine article focusing entirely on the discrepancies of the signal generator portion of the circuit.  That is the report that is on Revolution-Green.  Ms. Ainslie objected strongly to my comments but never addressed any of the issues that I raised.

The demonstrations in June and August of last year should have put the Paper 1 and Paper 2 completely to rest.  Ms. Ainslie honorably withdrew both papers in light of those demonstrations and then rather inexplicably without new experiment data decided to "reinstate" them.  She came up with some wild ideas that there was actually far more power being circulated in the breadboard circuit heater than indicated by her own temperature rise measurements. 

She now states that she intends to reproduce the experiments from the Quantum Magazine article as well as the Paper 1 and Paper 2 experiments.  I think that is fine.  As much as I know what I expect from those experiments, I think it is good for people who have ideas to put those ideas to the test.  If Ms. Ainslie and her team do not make gross mistakes then they should observe physically reasonable results.

[TinselKoala]

I'm sorry, but some of the above discussion from sims doesn't seem to jive completely with my results in the hardware. I know watching my demonstration videos gets old, but could you please watch again, and comment specifically upon, this video here, particularly "part 4b", the latter part of the video. Schematics are shown at the front of the video but also see the 4b schematic below.
The series resistance of 100 ohms with the bias battery is a carbon resistor. The CSR is a one-ohm wirewound power resistor.
So what I'm getting from the discussion above and from my testing is that the output impedance of the bias source is rather critical. If it's too low, as when I hooked the circuit up without the 100R briefly near the end of the video, it doesn't oscillate, but with a bit higher impedance (through the 100R and the bulb) it does. At 7:09 in the video I touch the bulb to the positive battery terminal bypassing the 100R, the bulb lights up brightly and the tpA trace shows -4 volts but with no oscillation. By fiddling with the connection I briefly get a pulse of oscillation, then I place the 100R in series and obtain the continuous oscillation and a dim bulb.
I presume the bulb lights up in the no-oscillation case because the mosfet's body Zener is reverse-biased enough to conduct, and that the two batteries are now in series and the scope trace from tpA is indicating the voltage drop across the body diode and the load resistance. Is that right?
In any case, due to the action of the body diode, the voltage at tpA never seems able to go below -4 volts or so, oscillating or not, regardless of the bias source impedance. Of course when the oscillations are happening this -4 V is the mean of the oscillations. Also, during oscillations there should be some current flowing across the source-gate capacitance as well, right?
I'm still trying to understand fully the behaviour of the circuit myself, so I'm happy for any and all guidance, and if I'm completely missing something please let me know!

http://www.youtube.com/watch?v=kV2ePEbJ76I