Testing the TK Tar Baby

[TinselKoala]

She's never going to do any tests that have the potential for falsifying her conjectures. She's had YEARS to do real tests and she's had plenty of advice on how to do them. All this current talk about PROVING her claims is just that: talk. What we will see, if we see anything at all, will be more bogus "demonstrations" like the video she uploaded, then repudiated and lied about.

It would take a matter of a couple of days to run the Ainslie circuit in high-heat, positive gate drive mode with oscillations, long on time pulses heating a load to 190 degrees C as in the video, then performing a simple dim bulb test. Yet she talks about "proving" all this other stuff that's unrelated to the simple definitive test that she PROMISED STEFAN SHE WOULD PERFORM.

There will be no cooperation, no real testing and nothing new from Ainslie. Does anybody really expect her to change her behaviour suddenly, after all these years? I certainly don't.

And still she lies about me. She ALMOST comes to the point of accusing me and PW and MH and the other non-sycophants here of engineering her break-in... she wants to, you can tell from her words, but even SHE knows that would just be too outrageous.  Yet still she blames us for it, and accuses me of "fabrication" of something.Yet... all my videos contain demos that anyone can do for themselves if they have a function generator, a scope, and a few mosfets and LEDs. Nothing is "fabricated" in my work. But there is definitely fabrication in Ainslie's bloviating accusations and claims of overunity and "tests" that will be performed.

I've explained several times why I'm doing this. It's because Ainslie cannot restrain herself from lying and insulting and distorting and misrepresenting and willfully ignoring evidence, and she refuses to correct her many egregious errors, and she misleads the innocent newbies with her lies and distortions. She pretends to "science" but gives science a bad name by misrepresenting what the scientific method is. She herself is woefully uneducated, a high school dropout, yet she denigrates the education and experience of others while at the same time refusing to learn even the simplest terms relating to her subject. 

All of this and more is demonstrated over and over in the various threads and forums she's posted in. Ainslie targeted me personally, has made physical threats, has annoyed people with her attempts to discern my identity, and no doubt would seek to harm me if she could. She's compared me to Hitler, Savonarola, and Satan himself, and she's called me a criminal sociopath. All because I require her to tell the simple truth and to respect her betters. Yes, Ainslie, we do know better than you. She constantly denigrates and insults others and lies about what those others say; she even argues with the people themselves telling them that they are wrong about what they said, even when the proof that she's wrong is presented in black and white.

Two words: Dunning-Kruger.


And.... the IsoTech GFG 324, Rosemary. 

Error, mistake, mendacity.... what is it? If it's a typo, then you are lying now, because you haven't corrected it and in fact you maintain its correctness.... but you cannot point to any reference anywhere outside your own mind for a IsoTech GFG 324 function generator. This is important because it shows how little regard you have, Ainslie, for correct facts and "open source" cooperation.

The issue of the function generator itself is not important... what is important is that you presented incorrect information and you cannot support the information you presented with even so much as a link to a store which carries the IsoTech GFG324 for sale.

And yet Ainslie has claimed that she corrects her errors.... when there are many of her errors still standing uncorrected and unretracted. In fact I can't think of a single error that she HAS corrected. Not even this one.

Bubba you're getting tedious in the extreme.  Correctly it is one Joule per second - but since 1 watt = 1 Joule and since 1 Joule = 1 watt per second - then AS I'VE EXPLAINED EARLIER - the terms are INTERCHANGEABLE.  Which is ALSO explained in WIKI.

Really? Maybe you should read the WIKI entry again Rosemary, with a dictionary by your side. You will find that the WIKI entry does NOT in fact support your absurd claim that 1 watt = 1 Joule and one Joule = one Watt per second,  and that the terms are interchangeable.

We are supposed to be discussing power measurements... with someone who confounds basic units of power and energy, who doesn't understand the difference between a RATE and a QUANTITY, and who can't even grasp basic algebraic relationships.. and not only that, she REFUSES to learn the correct formulations and mathematic manipulations.

Now we are told that she won't be posting any more until she has a bunch more insulting and misrepresenting "transcriptions" of my videos to post. Well, bring it on, Ainslie, make a fool of yourself even more than you have already... .the videos are up for all to see, and then everyone can take the Ains-lies and compare them for themselves with the actual videos.

But of course, just like the last time she said she wasn't going to post.... she will.

[TinselKoala]

In case anyone is still interested....

Let me review.

I have made a Tar Baby. This Tar Baby uses the schematic that Ainslie apparently has said is the correct one, out of five or six different versions that still have not been corrected or the incorrect ones retracted.
This Tar Baby uses the same component values to within component tolerances. (Power resistors for example usually have a 10 percent tolerance in values.) Tar Baby uses the same mosfet type, the IRFPG50,  but can also use others. Tar Baby can run, and has BEEN SHOWN TO RUN using either a FG, a 555 timer, a charge pump inverter, a DC battery or a regulated power supply for bias, making oscillations and negative mean power AND a negatively- accumulating (that is, DECREASING) energy integral while heating a load.... stronger evidence than Ainslie has shown for her negative power computation. This negative mean power has been obtained on BOTH an analog oscilloscope and a fancy DSO. The negative mean power can be produced on demand instantly, by taking Tar Baby out of the bag, hooking it up and turning it on. No sensitive tuning is necessary.
I've shown all these things that the NERD team has only claimed, never shown. Especially, I am the only one who has shown the unit oscillating and making negative mean power WITHOUT any external bias source, using only the main batteries and the charge pump inverter based on a 555 configured as an oscillator, to make a DC voltage that is more negative than the most negative pole of the running battery.

Today, Tar Baby and the Tek DSO demonstrated the dependence of the negative mean power calculations on the ripple of the battery voltage trace as seen in the system where lead inductance is a great factor. When the ripple is smoothed out by filtering and decoupling capacitors, the mean negative power goes away and the true, positive power can be read. The battery voltage does not actually fluctuate to the degree that the battery TRACE fluctuates, and filtering demonstrates this. Of course the special pleading will begin: the presence of the capacitors prevents the oscillations from affecting the battery ! Of course they do, even I can see this simple fact.

I made some still photos and a casual video demonstrating all this... but so what. Those that have the ability to understand it know already what the message is, and those who need to see and understand the message and the data... cannot and will not, because of willfull ignorance and overweening arrogance. Two words: Dunning-Kruger.

This is why only a battery draw-down test will really confront the main claim that the batteries don't discharge. A simple draw down test using the existing Ainslie apparatus and the still fully-charged batteries. A simple test, using positive long gate pulses, a 72 volt battery pack, and heating a load to high heat as in the demo video. Why is this simple test so hard for her to do? Answer: because she knows it will fail. A truly confident -- and capable -- claimant would long ago have settled the matter by proving me wrong in the simplest and most unequivocal manner imaginable: by passing the Dim Bulb Test.

(That Tek screen is hard to photograph. The dark red letters and numbers especially don't like to show up, particularly in the video. )

[picowatt]

TK,

I was begining to get a bit worried that you may have been having car isues again.

So, it seems that your empiracal data is in agreement with .99's analysis.  Excellent work, both of you.

PW

[TinselKoala]

@PW: Car is fixed, finally. I am now worried that the engine isn't getting hot enough !! My temp gauge goes from a low mark at 100, a middle mark at 180, and redline at 260, with the first red mark at 240. Before, I'd see "normal" cruise at just below the first red mark and if I pushed it to 70 it would get to the redline and the check engine light would come on and if I didn't slow way down it would start venting to the overflow and boiling. Now with the cleaned-out radiator and the wetting agent (11 dollars a bottle, used half a bottle)... now, I ran all the way to Austin, right at 95 miles, at 70 mph the whole way and it didn't even get to the 180 mark... less than half way up the gauge ! And that's with AC on full blast fulltime. It feels like a new 22-year old car. At last... I can get there without being totally exhausted and sweaty. And it's a lot more comfortable for Maggie, I'm sure. (She always comes with me.)

I just did the most casual Tar Baby testing imaginable. I didn't feel very motivated, sorry. I just took the Tar Baby out of the box and hooked it up and turned it on. I hadn't even tested it with the cap banks first, nor did I do any tracing or circuit probing other than just using the normal probe points at the board, and common references all at the negative power lead to the board, except that the CVR reference is right up at the body of the resistor usually. And I wanted to make sure that it would work using only the charge pump inverter and the lowest main battery of the stack for the bias supply along with the filter caps, so that's all I tried.

And it does. Not as stably as with an external supply but that is probably also a matter of suitable filtering of the inverter's input power, but still solidly negative average mean power with the caps disconnected, and solid positive with them connected.
I used 10 of the 1 uF blue RIFA ones, 10 of the 10nF silver rolls, and ten of the 100 nF BaTO monolithics all in parallel. So I guess that's closer to 11 uF than to 10. Within tolerances.   

I got more stable math with AC coupling of the CVR signal, but for the video and the still shots I used DC coupling. I tried to narrow the trace width by using the scope's HF filtering,  but...... interestingly...... when I used BWL at 20mHz for the CVR and Batt traces.... the mean power product stayed positive.
I didn't even bother to set up an integral but I did have the scope compute the areas of the math trace lobes, and they too reflect the negative power reading, even more solidly than the raw mean does. This is sort of equivalent to an integration. If the Areas of the positive and negative parts of the Math trace total to a Negative number, this will obviously result in a decreasing energy integral, just as before.

I'm not CLAIMING overunity performance, I'm only MEASURING it.   

But.... since we know about scope artefacts and measurement pitfalls and so forth.... I don't suspect zipons as the cause of the measurement because there are MANY more likely and possible explanations, and like a good little scientist I try to DISPROVE my hypotheses as hard as I can. When I fail to disprove one, no matter how hard I try... then perhaps, provisionally, until more information comes along.... I am justified in believing that hypothesis to be correct. But if experimental evidence comes along that does NOT support that hypothesis.... then it's lather rinse repeat again  until an even more robust and TESTABLE hypothesis can be formed.

An example of this is my early belief that probe skew might have been responsible for the errors. When I was looking at Steorn's eOrbo, I found that this was the likely cause of their overunity measurements, because I was able to duplicate their measurements, using expensive differential voltage probes and non-contact current probes (three thousand dollars for the LeCroy current probe, for example) NOT deskewed... and when the probes were properly deskewed the OU measurement went away. However in the Ainslie case, lower frequencies, less spiky signals, and the use of all the same, passive compensated voltage probes meant that probe skew wasn't supported by the experimental data... and so had to be dropped as a working hypothesis. No problem, scientists make "mistakes" by looking at "wrong" hypotheses all the time; that's what real scientific experimentation is about. Making mistakes, identifying them, correcting them, revising hypotheses, discarding theories completely if they do not fit the experimental facts.

Some scientists I know have been working for the past 5 YEARS on an experimental project that involves an extremely sensitive vacuum Cavendish balance. This instrument, designed and built totally in-house, is so sensitive that simply parking a truck in the parking lot of the building causes the beam to swing over from the gravitational attraction of the truck to the test masses. It is maintained at a vacuum of 10e-7 Torr. A data run takes a week, because the natural oscillation frequency is about 4 hours and a reasonable number of swings -- 20 or 30 or so --are needed for the Fourier analysis that results in data from the instrument.  The instrument is so sensitive that large areas of the parking lot near the building are cordoned off and even the cleaning staff isn't allowed in the room where the experiment is set up.

And.... it turns out that the hypothesis the experiment was designed to test is not supported by the data. In other words, it is in error, or the data is wrong, or the analysis is wrong, or any number of other "mistakes" have been made.... but what does seem definite is that the data as obtained don't support the hypothesis under test and therefore call the entire overarching theoretical framework of the research into question. A team of six researchers and technicians has been working on this particular experiment for over 5 years, theory and hardware. And it's a "mistake", and this is how real science progresses: by falsifying theories.

I'll post a video of last night's Play Date in a few minutes.

[TinselKoala]

Hmmm.... eight views even before I post a link.  I hope the MiBs are finding this all entertaining.

http://www.youtube.com/watch?v=fBjcOzm7pFE

I apologise for not doing all the testing and fiddling around that was suggested. I'm losing my enthusiasm for this project, because, just as the person pointed out in the Dunning-Kruger clip.... everything we are doing is falling on deliberately deafened ears.  Nevertheless, and in spite of my flagging interest and the attacks on me that are sure to come from YKW, I will lug the thing around with me several more times to go play with little Tex (Tar Baby's playmate).

Just as before, when I replicated the COP>17 claims -- the actual measurements and data -- and showed that the batteries didn't charge up and the heat to the load was less than with the same DC power.... she will claim that I ... along with everybody else.... "failed" somehow and instead of doing a solid demonstration or experiment to refute our "failures" she simply bloviates, goes to a new forum, and claims that ALL of the people she once worked WITH are now incompetent thieves and liars.

(ETA: The COP>17 circuit was more interesting than this one, because you could actually DO something with it. It made a fair PWM motor controller, for example, and you actually could siphon off charge and use it to charge a capacitor to much higher voltages than the battery could supply, and you could charge an _external_ battery either with destructive HV pulsecharging to "fluff" it to apparent high charge levels, or even properly with filtered current pulses at more reasonable voltages using caps and diodes. But of course this charging power came from the main batteries, so couldn't be used to charge themselves other than to convert some real charge to "fluffy charge" with substantial energy losses and potential (pun intended) battery damage. This circuit here is useless except pedagogically.)