another small breakthrough on our NERD technology.

[Rosemary Ainslie]

And there's more?  Golly.

The MOSFETs don't flip-flop. The tests seem to indicate they operate the device in two slightly different modes; one where Q1 is always OFF and Q2 alternates between ON and OFF,

Yes indeed.  I SEE this now.  This means that somehow the applied signal at the Q2 would be positive and then negative and then positive and so on.  That would INDEED explain EVERYTHING.  Which also would mean that our signal generator is simply NOT FUNCTIONING.  I'll need to take this up with the manufacturers.  Not a good thing.  Not a good thing at all.  What ever next?

... and a mode where the opposite occurs (lower battery voltage and offset setting).

Which would mean what?  That the signal at Q1 would then be positive and then negative and then positive and so on?  While Q2 just sits in the sidelines and sulks?  Good thinking.  It's about as reasonable an explanation for this anomaly as ANY rather frantic disclaimer would need.  It's not however, STRICTLY in line with the evidence.  Unless, of course there's anything more than deliberate ambivalence in that reference to 'offset setting'.  Does he mean that the offset is then also correspondingly lower?  Or does he mean that the offset is then 'higher'?  Either way.  It's wonderfully confusing.  Tell him from me that this is very well done indeed.

Regards
Rosemary

took out the balance of this post as it's falling into a black hole.  I don't want to be sucked in.   

[Rosemary Ainslie]

Right,

Please post a link to the user's manual or glossy, or web page advertisement for the GFG 324. Or better yet, post a pic of your test apparatus used for the data etc. in the paper.

The unit you used in your video demo was the one I mentioned, the Instek GFG-8216A. Isotech does however make the same model as Instek, i.e. in the 8200 series.
http://www.iso-techonline.com/products/iso-tech-oscilloscopes-function-generators.html

If you've changed FG's since that time, then my mistake, however I've not found a model 324. I believe 324 is the model number for your LeCroy scope.

First pic is from your video, second is from ad.

.99

My question here is WHO EXACTLY is Poynty talking to?  I'm confused by his rather uncharacteristic use of the word 'PLEASE'?  That's SURELY NOT our Poynty Point?  Good heavens.  In any event - I certainly don't have the wherewith all to post any pictures.  I have no camera at the moment.  More's the pity.  And even if I did - there's no need.  The usual practice is for the collaborators to SIGN OFF on the details of the paper when they're all satisfied that the facts are clearly and correctly presented.  And we've all signed off.  But as a rule, those readers of those papers usually take the representations at face value.  There is a presumption that there is no deliberate effort to misrepresent the facts.  And it would be a rather trivial FACT to distort - when that distortion may negate the entire paper.  I would have thought?

Regards,
Rosemary

[Rosemary Ainslie]

Now guys - just to keep the argument in full focus.  There are more confusions.  If there WERE a 'flip flop' condition where the battery simply used either Q1 OR Q2 - as was made available - then - we must also acknowledge that the battery would then ALSO be delivering it's current with very little interruption ALL THE TIME.  At worst there would be a 'spike' as it moved from the path of the one transistor and then to the other.  On the whole though the waveform would show a voltage that is CONTINUOUSLY greater than zero.

Which is why we are now hearing the argument that there's a mystical 'on one moment' 'off the next' and so on - so that it has a CHANCE for the circuit to construct that waveform - that oscillation. I see the problem now.  And it would have helped if I'd both seen it and mentioned it before.  This is when our 'naysayers' for want of a more polite term - actually serve science well.  They hone into the problems of their own counter proposals.  And gradually the questions are THRASHED OUT.  Never a bad thing.  Just an enormous pity that it cannot all be done more courteously.  It would, on the whole, encourage a greater participation. 

I only say this because I am in receipt of an inordinate amount of mail from those who do not participate on these forums.  I suspect that - given a less fraught environment - then they would be more ready to engage.  Much needed.  As there of many really excellent arguments that I hear from them.  And indeed, proposals for different kinds of tests.  It's  something that - perhaps - we should all work on.  Since I see that Poynty is trying to converse - albeit through TK and not myself - then perhaps too - it would be as well that I also desist with my sarcasm.  But then Poynty Point - you need to reign in your appalling manners.  They're shameful.

Regards,
Rosemary

edited
had to change 'there's' to 'there are'.  It was irritating me.

[Rosemary Ainslie]

This is the argument that I was trying to reference earlier...

The MOSFETs don't flip-flop. The tests seem to indicate they operate the device in two slightly different modes; one where Q1 is always OFF and Q2 alternates between ON and OFF, and a mode where the opposite occurs (lower battery voltage and offset setting). I've analysed the mode where Q2 is active and Q1 not. In this case, the FG offset is set to the NEGATIVE side (offset knob pulled and turned ccw), such that the FG output is never positive enough to turn Q1 ON, but because Q2-G is connected to the FG negative, this does turn Q2 ON (two negatives make a positive wrt Q2's V_GS).

Note this part of the concluding sentence.  '...but because the Q2-G (presumably G stands for GATE) is connected to the FG negative....'  Just that.  We were given to understand that it was NOT.  I proposed that it WAS.  My proposal was blasted with a blistering reminder that I KNEW NOT WHEREOF I SPOKE.  Why then are YOU now proposing this?  And IF INDEED it IS connected to the negative - then HOW DO YOU PROPOSE TO ARGUE THAT THE BATTERY IS ABLE TO DELIVER ANY ENERGY? 

I'm arguing that Rosemary does not know how to read a diagram (she can't see the common connections I listed), nor does she know how MOSFETs operate.

ON THE CONTRARY.  I have a fair and working knowledge of how MOSFETS work and I most CERTAINLY CAN SEE THOSE COMMON CONNECTIONS. 

Regards,
Rosemary

[TinselKoala]

If the core would not saturate, the circuit would have a Q of 7900  in unloaded state !!!


Yeah, it's so exciting now ! 

Yes, exactly. So a non-saturable core material is preferred, as each "slosh" as the energy goes between inductance and capacitance is an opportunity for energy to enter the system from outside and reinforce the resonance. Careful selection of your resonant frequency is also important here.... if you push the "swing" at too fast or too slow a rate you won't get optimal coupling of your input power to your resonant storage, so if you're looking to pick up energy from outside the system you need to have some idea of how to match its frequency. (My little contribution to the general theoretical BS around Tesla and MEGs and so on.)
Air (vacuum) works pretty good for a core material at the energies we are using. I hope you've had a chance to look at my TinselKoil videos on YT. I am using a similar switched-mosfet circuit as Rosemary does (except that I use a full bridge -- 4 mosfets -- instead of essentially 2), but because I know a bit about what I'm doing, I've gotten much better results.

@Rosemary, you seem to have trouble accepting that circuits like these can have current peaks in the multi-kiloAmpere range. Let me assure you this is not only very possible but common. POWER, as you have finally figured out, is the rate of energy dissipation. As a rate, it incorporates a time dimension. If the time duration of a high-current spike is small, there will be little POWER in it, hence little heating of conductors, and so on.

@.99-- yes, I can see that now-- the mosfet behaviour will be sensitive to the relationship between the battery voltage and the FG's output voltage level, and the mosfets will interact through the circuit's capacitances. It would be interesting to apply the FG's signal through an appropriately chosen series capacitor, to assure only AC coupling.
It's clear from the blather above that Rosemary really still doesn't understand her circuit, nor the basics of power measurement, and most especially artifacts induced by measurement probes and other wiring. Still--- isn't it relatively easy to build this circuit, or sim it, and show how it actually behaves?