Testing the TK Tar Baby

[TinselKoala]

Yes, MH, she's contradicting herself yet again, in the most blatant and transparent way. How many times has she denied that the FG can pass current? How can she not realize that the current to charge the GATE is not even measureable by her instruments it's so small? I showed a mosfet gate turning fully on by the charge in my fingertips. Not dealing with Yet Another example of her bogus math, her confounding of energy and power terms, and her lie about "not posting the video" when it's linked from her own blog-- it's the same diversionary tactic she always uses when a real issue comes up. Insult, lie, threaten, bloviate, anything at all but deal with the issues recently raised:

IsoTech GFG324 function generator.
Current through the FG or 555 or other bias supply.
Posting of the video and lying about it.
YET ANOTHER calculation that is grossly in error and a claim made based on that error... not corrected or retracted or even MENTIONED by her.
The "diagram" by which she expected PW to understand her meaning... and then he gets berated when he can't make out the gobbledegook and idiosyncratic handwriting, not designed to communicate anything but the writer's egotism.



Sorry, I thought I was being clear about the schematic I am using.  I use the same schematic, except for the substitution of a 555, a battery, an external power supply, or the charge pump inverter as "drop-in" substitutes for the FG, as Ainslie claims to have used for the experiments described in the papers. Which one? I use the one with the Q2 mosfets on the left, and the "direct connection" where the FG "negative" is on the transistor side of the CVR. As below.

Now... why aren't the common mosfet drains monitored in Ainslie's work? Here's why: it is because she still doesn't grasp that, with high-side switching, the DRAIN of the mosfet is HIGH... that is, at battery voltage... when the mosfet is OFF, and when the mosfet turns ON the drain voltage goes LOW, nearly to the negative rail or just above zero volts. Take a look at this blog post of hers, and note that she is confused and amazed by the drain voltage going to nearly zero when the mosfets turn on.
http://newlightondarkenergy.blogspot.com/2012/01/232-another-summation-based-on.html

ETA: I guess by now I should do a Revision E, since I am now using that tiny, much lower inductance, 0.2 ohm CVR instead of the ridiculous power resistors. If it gets hot when using a positive drive to actually turn a mosfet on fully, then I'll make a higher-power stack of them. But they cost 1.50 each at the surplus house... 3x the cost of an ordinary cement power resistor.
Oh well.

[TinselKoala]

By the way, I've done some testing with the IRF830a mosfets today, and it looks to me like they will make the same negative mean power if the battery isn't de-rippled by caps. The CVR oscillations have the same basic shape and distribution wrt the zero reference and the battery ripple looks just the same, but the frequency is about twice that seen with the PG50s.

I've also run today with the 555 timer making NEGATIVE pulses, powered by the charge pump inverter which in turn is powered by the lowest of the main batteries, no external bias source used.

So.. I've made purely positive gate drive pulses using the plain 555 circuit I've posted before, but without oscillations, just mosfet on fully. (This happens because there is no "negative offset" possible with just a simple 555 hookup. A negative voltage on Q2 sources is required for the oscillations to occur and the positive 555 can't provide it.)

I've used a straight 9v battery through a pot/resistor combo to make the _negative_ bias drive for the low heat oscillation mode.

I've done the same with an external regulated supply.

I've also done the same just using the charge pump inverter to act as a negative bias source, and this can be powered either by 12 volts from the regulated power supply or the main battery.

And I've used the 555 timer, run from the inverter, to make the negative-going pulsed bias supply.

All these methods work, some with more distortion than others. With  just the charge inverter running from an external supply and no 555 pulser, the oscillations become almost perfectly sinusoidal. Running the 555 "upside down" using the inverter from the main battery and making pulsations as well as oscillations, the waveshapes are quite distorted but still distinct and at the correct frequency. I think with appropriate filtering the system could work equally well using any of the bias sources I've tried.

[picowatt]

TK,

I read the blog link.

Based on that read, I think part of RA's misunderstanding is that she believes the gate of Q2 is being switched by the FG.  The gate of Q2 will always be very close to ground.  It is at whatever voltage the non-battery end of the CSR is at.  I think, from her drawing in that blog link, that she believes the FG is making the Q2 gate positive when the FG output is negative.  Her reference to current flow or switching regarding Q2's gate further supports this.

I do not think she realizes an FG has a bipolar power supply that allows its output to swing either positive or negative with respect to its signal ground, which will always remain at the potential to which it is attached.  When the FG output is positive, it charges the Q1 gate and turns Q1 on.  But when ithe FG output is negative, the Q2 gate continues to remain very close to ground and its source is brought negative by the FG, which biases Q2 on and current flows thru Q2 and the FG to the CSR.  I believe she thinks the FG is switching the Q2 gate when the FG is negative and this is clearly not so.   

Possiibly, if someone redrew a drawing like the first one GL posted with two batteries in series representing the FG bipolar supply with the centertap of the batteries connected to the CSR and a SPDT switch to select the polarity supplied via a 50R to the Q2 source and Q1 gate, it might help her understand.

Might...

PW

[fuzzytomcat]

Hi all,

Here are some nifty animations of a n-Channel Enhancement MOSFET ... some may find these useful or educational.

http://www-g.eng.cam.ac.uk/mmg/teaching/linearcircuits/mosfet.html           ( N-channel enhancement MOSFET ( adjustable ))

http://matec.org/animations/matec/M001FL02.swf      ( This is an animation of a n-Channel Enhancement MOSFET Characteristic Curves. )

http://wps.pearsoned.com/wps/media/objects/11427/11702033/Biasing%20of%20D%20Mosfet.swf   ( Biasing of E-MOSFET (enhanced type))

Fuzzy

[TinselKoala]

@PW: Yes, that is part of the problem.... her reaction to the video I made of the FG simply lighting the LEDs alternately was that I somehow had faked it or was employing "misdirection" of some kind. She can't seem to believe that the FG is a bipolar power supply, even though she's using it as one in her circuit.

But, from long experience with Ainslie ... I think there is another big conceptual error revealed in that post. You see that she is talking about putting a probe on the drains, in her garbled manner.

The oscilloscope probes are placed directly across the batteries that ground is at the source rail and the probe is at the drain.  Which is standard convention.  Then. During the period when the oscillation is greater than zero - in other words - when the battery is DISCHARGING - then it's voltage it falls.  And it SERIOUSLY falls.  It goes from + 12 volts to + 0.5.  Given a  6 battery bank, for example, then it goes from + 72 volts to + 3 volts.  At which point the oscillation reaches its peak positive voltage.  And this voltage increase is during the period when the applied signal at Q1, is negative.

So she's applying a signal to Q1 gate (and Q2 sources) that is negative... which means the signal at Q2 gates is positive and switching those mosfets on... making the voltage at the common drain pins fall. And she's describing the anti-phase relationship between the voltage on the mosfet _drains_ and some other signal from the circuit.

So she is seeing the normal drop in indicated voltage at the drain of a mosfet doing high-side switching when the transistor is turned ON. She has been confused about this many times in the past; in fact it is at the heart of the earlier COP>17 claimed circuit. Recall the duty cycle issue with the earlier COP>17 claim? Her timer produced a signal that caused the DRAIN ... which she sometimes thinks is equivalent to monitoring the load...  to be HIGH... at battery voltage.... for 3.7 percent of the time. So she thought she was making the LOAD be ON and carrying current 3.7 percent of the time. Which of course is exactly backwards.

This is the reason why I did the several demonstrations where I stress that when the DRAIN of the mosfet is LOW-- at or near the zero voltage level just as she describes in the blog-- the mosfet is ON, conducting, and this is absolutely normal and exactly what is expected, and when the Drain voltage is HIGH-- at battery voltage, the mosfet is OFF and not conducting and so the load is not carrying any current. She can't seem to wrap her mind around this concept.
She sees the drain pin oscillating and thinks that when the voltage indicates low during the oscillations that the battery has discharged its entire capacity or more in a "fraction of a moment", not realizing that the drop in voltage at the drain pin only means that the mosfet is on or partially so and nothing more.

I think this is why the common drains is never shown in her scopeshots except for that one time in the video. She doesn't understand what it is showing so she leaves it out and strenuously avoids discussion of it. Remember me commenting on the "Green Trace" and her denials that it was the drains... but her never saying what she believed  it was. In spite of the narration and the circuit diagram and the actual viewing of the scene she still denied that that trace on the TEk screen was the common drains, yet never has said what she believes it to be.