another small breakthrough on our NERD technology.

[Rosemary Ainslie]

I have just noticed these posts - from AbbaRue and Gyula.  And, indeed, just looked at Professor Niknejad's file.  Very interesting.  Not that I can understand those equations.  Now.  Gyula.  I get it that you're an authority on those MOSFETS.  And I get it that you're refuting the relevance of Niknejad's work to our own based on AbbaRue's mistaken assumption that, as he puts it - 'Main difference is he used the drain  instead of the source. (Theoretically that shouldn't matter.)' which is referenced here.

...The lecture paper you linked shows the so called cross coupled oscillators with  devices like MOSFETs and BJTs (bipolar junction transistor) devices.  Such circuits exhibit negative resistance between the MOSFETs drains or between the BJTs collectors as a result of positive feedback. Positive feedback results when one device’s  control electrode (i.e. the gate of the MOSFET or base of the BJT) is connected to the other device’s drain or collector electrode and vice versa, as shown in Page 12 in your link. You wrote:  Main difference is he used the drain  instead of the source. (Theoretically that shouldn't matter.)

Unlike you I'm NOT an expert.  But it occurs to me that AbbaRue's conclusions are 'on the money'.  The fact is that those oscillations should persist IF one transposes  'P Channel MOSFET's appropriately on that circuit.  In which event the DRAIN would indeed be used instead of the source.  And I'm ENTIRELY satisfied that IF we we replaced our own with the P Channel type then we would, MOST CERTAINLY manage the same oscillations for the same extended durations.

Then with respect to this comment from a previous post on another thread.

...This means that if you have a 200V max drain-source voltage rated n-channel MOSFET, the body diode is reverse biased all the way up to 200V as long as the drain gets max 200V positive DC or peak AC with respect to the source electrode. The moment the drain voltage goes negative with respect to the source electrode, the body diode gets forward biased (if the voltage difference gets higher than 0.8-1V of course in the forward direction) so current can flow via the body diode, regardless of the control voltage between the source-gate electrodes.
Of course, if you exceed the max 200V drain-source voltage rating the manufacturer defined for a particular MOSFET type, then the device will gradually start conducting current between its drain-source path, either via the reverse biased body diode or via the drain-source electrodes, bringing the MOSFET gradually or suddenly into a destruction process.

I've highlighted the point at issue.

You are here describing our standard 'runaway' parasitic oscillation.  This is most certainly the result of paralleled transistors.  And you are also on the money.  BUT.  There is nothing STANDARD about the oscillation that our circuit manages.  In the first instance the paralleling of those MOSFETs at Q2 is not a required condition to generate that oscillation - as I've mentioned - everywhere.  And the relationship between Q1 and Q2 is absolutely NOT in parallel - as you yourself, realise.  Then the oscillation does not DEGRADE as is ALWAYS evident in a parasitic oscillation.  Then FINALLY - our MOSFETS are apparently conducting current way in excess of it's rated tolerance levels without any evident degradation at all.  And that they're still entirely in tact is really easy to demonstrate and prove.  So.  The question persists.  How is it that those transistors can conduct energy from the battery supply when the battery is NOT in a closed circuit condition?  You also seem to think that the negative terminal of the function generator can carry this relatively HIGH current flow from the gate of Q2 to the source - or as Poynty Point refers to it - to the Negative Terminal of the battery supply.  While there is INDEED a bridge between the function generator's probes - with an appropriate signal charge at Q2 - THERE IS NO EVIDENCE IN THE WAVEFORM TAKEN ACROSS THOSE TRANSISTORS THAT CAN ACCOUNT FOR THAT HIGH CURRENT FLOW.  Quite apart from which, I'm in the process of arguing WHY it is that the battery CANNOT conduct.  Are my comments that irrelevant that you can simply IGNORE them?  Please bear in mind that we have what looks like a normal parasitic oscillation PERSISTING over an extended period that in, no way, compromises the material of those transistors. 

With reference to this question to AbbaRue - if I may presume to answer it...

Can you agree with the connection of Q2’s source to the battery negative via the 50 Ohm inner resistance of the FG?

You are RIGHT.  BUT there is absolutely NO EVIDENCE OF THIS CONDUCTION IN OUR WAVEFORMS.  May I refer you to our previous waveform details.  I'll take the trouble to download this again when I've posted here.  I think that, on the whole, this entire subject is very easily dismissed if one also takes the trouble to IGNORE the evidence.  But I'm not sure that ignoring the evidence will entirely answer the questions. With respect.

Regards,
Rosemary
added

[Rosemary Ainslie]

Gyula,

Here - I hope - is the typical waveform detail.  Note the blue trace during the period when the oscillation occurs.  THAT is absolutely ALL the voltage that is evident during this period.  It is NOT reflecting the flow of high current from anywhere at all.  And ignore the file name.  It was first intended for MileHigh.  But he cannot understand the significance. Sadly.

Again, regards,
Rosemary

[TinselKoala]

Remember these, from hhoforvolts..... be sure to read the descriptions and hhoforvolts's comments.
http://www.youtube.com/watch?v=cZfkrsH_USo
http://www.youtube.com/watch?v=NfvEak7rV7g

[Rosemary Ainslie]

And here, I hope, is another example of that voltage across those transistors.  This sample is of particular interest because it results in a voltage swing at the batteries that actually take it below zero - for very brief durations.  And we all KNOW that for the battery to deliver that much energy in such brief moments - then we'd be dealing with something that could probably power entire household requirements.  So.  It's clearly NOT energy that is coming from the battery supply.

BUT please note this much.  The voltage at the switch is absolutely NOT consistent with those high voltage spikes.  Wherever it is that this current is moving from and to - it is NOT through the signal terminals - else it WOULD be evident on the voltage across those transistors.  (the blue trace)

Again,
Rosemary

added the most of this post.  Just needed to see that I'd uploaded the right file. 

[Rosemary Ainslie]

My Dear MileHigh,

The only way I 'errored'  - as Poynty puts it - is when I assumed that you had the wherewithall to understand the niceties of my argument.  I was in the process of explaining all this when I was broadsided by the unexpected though partial endorsement by Gravock for our thesis.  No sooner was that managed than AbbaRue proposed to take the circuit seriously.  Which shows a commendable impartiality which is a rare and 'wondrous' thing.  Especially given the excess of ego that usually litters this thread.  However, given a bit more time for the shock of all this this dissipate - then I'll resume that explanation.  But it will require that you read it.  Which may tax your patience.  I only assure you that you WILL be rewarded.

And, unlike others - whatever their level of credential - I am MOST happy to acknowledge when I've been on a learning curve.  The question is - are you? And your comments are valid.  I SHOULD indeed have pointed out what trace relates to what.  Abject apologies.  Here's that list.

IN ALL OUR DOWNLOADS
The blue trace is across Q1 which corresponds to the signal terminals (both) .
The Orange/yellow trace is across the shunt
The Red trace is the sum of the voltages across the shunt and the battery
The Lilac/Cerise trace is across the battery
Integrations - like the sum - usually refer to the orange and Cerise (battery and shunt - but given as voltages)

If I've left out any, then let me know.  And now, MilesUpThere - could you PLEASE take careful note of that BLUE TRACE.  Quite apart from the argument that will follow - where I will FINALLY PROVE that it is IMPOSSIBLE that the battery can be discharging any current during the 'OFF' period of those duty cycles - just NOTA BENE - as ever.  THERE IS ALSO NO EVIDENCE OF THE BATTERY DISCHARGING THROUGH THOSE SWITCHES OR ANYWHERE IN THE VICINITY OF THOSE PROBES. You will recall that this was the cornerstone, the lynchpin of your argument. 

ALSO.  I have NEVER disputed that there is a FLOW OF CURRENT from the battery during the ON PERIOD of the duty cycle - subject to the required setting of the function generator's offset switch.  WHAT I am ACTUALLY REFERRING TO - is the manifest miracle of so much current flowing in that oscillation - DURING the OFF period of the duty cycle - when there is NO EVIDENCE of a commensurate INCREASE in the voltage across the switch.  Without this increase - then one must think?  Surely?  Where does all that current come from?  Where does all that current go?  How does it sidestep that switch?  Well.  That's what I HOPE TO EXPLAIN.  Eventually.  It's just that there are way more distractions than I can comfortably deal with.  I am, after all, merely mortal.

Kindest regards MileHigh.  I value your excessive smugness.  The more so as I hope to put a dent in it.  One day.  LOL

Rosie Pose.


added
Highlighted the trace details