Testing the TK Tar Baby

[TinselKoala]

(snip) they are taking the peaks of the oscillations on the mosfet drains, times the current at those times, and taking the time periods and multiplying these together to get a quasi-time-integration of a power signal, I think. Am I right about this?

So they are looking the area under the curve, formed by the peaks of the oscillations on the drain trace wrt the baseline of that trace. Am I right about this?


If so..... there is a minor problem.   


And this illustrates just what the problem is with doing it that way:

http://www.youtube.com/watch?v=_CXWWupl0MU

So if you are integrating the area "under" a drain trace oscillation signal.... You are actually integrating an "off" signal, over the time periods when the mosfet is OFF and not conducting current to the load.

At least I think that's how it works.

[TinselKoala]

Their OWN raw data don't even support their claims of load heating efficiency and battery recharging. The scope data as I have shown is bogus. Look at the channel means on the battery and the current. They are both positive. Yet the scope calculates a negative mean for the multiplication of the two channels. This is the NUMBER ONE sign that the scope is doing GIGO.

The NUMBER TWO sign is that the scope is not performing Ohm's Law on the voltage drop across the CVR, it is simply multiplying the raw voltage data together. Not only is the sign of the resultant math wrong, but so is the magnitude, because the voltage numbers that are given by the CVR trace need to be multiplied by 4 to give the current in Amperes, because of that 0.25 ohm CVR.

Now you know why I like to use a 1 Ohm CVR when possible.

All of this leaves me feeling very strangely. What does it say about the researchers making the claims? What does it say about the community of people looking at and BELIEVING her claims without critical analysis? The raw data that are available are sufficient enough to quash the claims altogether, just in those two scope shots that I have analyzed. And they have been available for a YEAR. I think .99 and MH have taken a look and pointed out some of the inconsistencies, but nobody listened to them.

COME ON PEOPLE, if you are going to Save the World from the Tyranny of Big Oil...... you really need to polish your skills and knowledge a little bit, to avoid getting suckered into fraud hoax lying mendacity like the Rosemary Ainslie claims of COP INFINITY.

[TinselKoala]

@Fuzzy:
Is Donovan Martin a blood relative of Rosemary Ainslie? I str that her "son" or maybe "grandson" was involved in posting or working with her at one time some years past. Is Donovan Martin Ainslie's son? Whatever happened to that son's involvement, I wonder.
I suppose Martin has still not deigned to answer your questions.

May I suggest the following: Send him a certified, registered letter, containing a reprint of the papers bearing his name, a link to the NERD thread and to this one, and a cover letter explaining what is up, also mentioning the "legal action" that Rosemary is threatening you with. Sign your letter with the honorific "Esquire".... that will get his attention for sure.

[MileHigh]

TK:

I read through your analysis and followed the annotated scope shot and it looks great.  Having two computer monitors really helps.  You can't go back when you add a second black monitor to your setup.

It's funny because it's raining on Rosie's parade so it may as well pour.  I can't tell you how many times I have felt frustration when Ms. Posie would "throw up" a screen shot of a DSO capture in a thread posting to advance her argument with no indication of what each colour meant, no summary of what each trace was doing, etc.  When I read her original report I had to make the mental effort to flip back and forth through the pages to remember what everything meant.  But after that whenever she would throw up a DSO capture my brain would simply turn into mush because of the excessive amount of work it would take to figure out what the image actually meant.  Rosie, your DSO scope shot presentation skills were nil, and I bet you the vast majority of readers ignored your scope shots just like me.  A mini rant!

I just will add some colour commentary to what you said referencing some of the recent things that have been discussed.

Look at the top yellow trace across the current sensing resistor.  You will notice that the oscillation is perfectly symmetrical about the ground reference as shown by TK's item #2.  It's perfectly symmetrical about ground because it's AC-coupled and coming from the Q2 MOSFET oscillator.  That portion of the trace is not even an indicator of the battery current.  That's because during negative offset oscillation mode, the battery current is flowing through the function generator.

Look at the mauve trace below the yellow trace which represents the battery voltage.  The massive AC voltage waveform that you see on the battery voltage is not real, it's a mirage.  We simply know that the batteries would not do this with 100% confidence.  One possible scenario previously discussed was that the Q2 MOSFET oscillator is responsible for this.  We know that there is an AC-coupled path between the Q2 MOSFET oscillator and the battery positive terminal.

In addition, because of the very high speed switching, the inductive energy stored in the interconnect wires themselves can cause AC voltage to be superimposed on the battery voltage.  Notice how the AC voltage is not symmetrical about the nominal battery voltage shown by TK's item #3 in the annotated scope capture.  Whenever the Q2 MOSFET array is in the process of switching off while in oscillation mode, that switching off will result in the magnetic field collapsing in the battery interconnect wires.  That will generate a positive voltage bump or spike.  That may explain why we see that the AC waveform superimposed on the battery voltage is positively offset.

Note how in Rosemary's report they just throw a bunch of DSO captures at you and make no attempt whatsoever to explain them.  The only thing that they want you to look at is the final (V x V) average voltage product calculated by the DSO for the math waveform.  That is simply ridiculous.  You do not capture scope traces without explaining all aspects of what they mean if you are submitting a report that purports to demonstrate over unity.  Like I said before, we are not mindless guppies swimming up against the glass in a fishbowl.  We have to use logic and reason and evidence and the analysis of that evidence to support our arguments.  We also have to demonstrate that we understand our data and clearly explain our captured scope traces and relate them back to the operation of the circuit.  Almost none of this was done for the NERD RAT claim and you are seeing the fallout from that right now.

MileHigh

[MileHigh]

I am now just going to do an abbreviated analysis of the yellow current trace and the mauve battery voltage trace and relate that back to the average pseudo power calculation,  i.e.; the (V xV) average, item #7 in TK's annotated DSO capture.

Let's look at the first negative oscillation component of the vaveform.  For the yellow current trace, we know it's completely bogus, it's not even the battery current.   For the mauve battery voltage trace, we know it's completely bogus, because the battery voltage is not doing that.   What you are looking at is likely external high-frequency AC and inductive collapse spikes superimposed on the normal steady DC battery voltage.

So, for the average pseudo power calculation for the negative oscillation component of the waveform, you have (garbage x garbage) = garbage-squared.

Let's look at the non-oscillating component of the waveform.  TK did that analysis a few postings back and we know that it represents about 120 watts of instantaneous power being dissipated here.  Q1 is on and real current is flowing in this case through the CSR.  Since this component represents about 50% of the duty cycle, the average power contribution during this phase is about 60 watts.

So, to keep life simple, let's forget about the (V x V) business and just talk in terms of "watts."  (I am cheating.)

If the average from the math trace indicates negative "wattage" then that means you have something like this:

(negative oscillation watts) + (non-oscillation watts) = negative "wattage"

(garbage-squared) + (60 watts) = negative "wattage"

Therefore, the "garbage-squared" portion of the waveform must average to even less than negative 60 watts!!!

The implication if you are operating in NERD RAT fantasy mode is quite startling.  While the circuit is running in negative oscillation mode the circuit is pumping more than 60 watts of power back into the batteries!!!

Incredible!

MileHigh