Ms. Ainslie states on her blog that she does not understand some of my comments:

" I am to be credited with the 'public' reproduction of private tests at August 10 2013 where I apparently 'showed that the heat output from the heating element was only 20% of the energy drawn from the battery'?  Not at all sure what he's saying here. "

For Ms. Ainslie's benefit: 20% means 20 parts in 100, two parts in ten, or one fifth of a whole.  I direct Ms. Ainslie's attention to a portion of her own retraction statement for the source of the 20% figure:

"Reference measurements taken at new sense points directly at the battery bank indicated average net positive battery drain of 14W to 15W. Maximum heater temperature rise during these experiments was 21C. From our electrical DC power to temperature rise tests conducted in 2011 and appear as Table II in this paper, a 21C heater temperature rise corresponds to an equivalent power of between 2.4W and 3.4W. We therefore obtained heat output that was only a fraction of the input power." 

2.4W to 3.4W heat output for 14W to 15W input draw from the batteries amounts to 16% to 24% output/input efficiency.  IE the figures yield a median 20% value.  That low efficiency compares very unfavorably against a direct wire connection from the battery to the heating element which would have yielded very close to 100% output heating power versus input electrical power efficiency.   The August 11, demonstration participants: Ms. Ainslie, Donovan Martin, and Steve Weir can be heard discussing those results near the end of the August 11 demonstration.  Ms. Ainslie declared "absolutely" that the output heat power was a much smaller than the measured battery input power.  The fraction was so small that Ms. Ainslie canceled phase four of the demonstration as unnecessary.

The "benefit" that Ms. Ainslie claims, she and her collaborators have failed to demonstrate.  They have instead shown that the majority of energy drawn from the battery is lost in other parts of her circuit other than the target heating element.  In other words, Ms. Ainslie and her collaborators have both shown and admitted that her Q-Array operated as she chose requires about five times as much input energy to evolve a given amount of output heat as an ordinary resistance heater.

I hope these expanded comments help Ms. Ainslie better understand my earlier post, and refresh her memory as to her own conclusions.

For those who may have come in late to the revelry, here is a clip from the long demonstration.

The first, short part of the clip illustrates the Ainslie team being surprised -- to the point of suspecting their apparatus -- by a perfectly normal and ordinary behaviour of their circuit, which Steve Weir figured out in seconds, once they had it together enough to present him with coherent data.

The second, longer part of the clip is the final portion of the demonstration, which Mark E. refers to above.

I have done absolutely no editing except to clip these out from the longer 4-hour stultifying recording of what has got to be the most amazing demonstration of incompetence in measurement that I have ever seen. (At one point in another clip the engineer Donovan Martin fumbles about for many minutes trying to determine the frequency of a simple near-sinusoidal signal on the digital oscilloscope, and is unable to do it.) The bad video is because these people decided to perform their important demonstration using a _cellphone_. The bad audio is because of the cellphone's connection and because the operator keeps putting his fingers over the microphone!



http://www.youtube.com/watch?v=JAz1Snh75HY
(Sorry, I didn't realize that the two clips are together in one segment. Edited away the first, redundant one.)

TinselKoala, from what I can tell Ms. Ainslie has confused some of your comments regarding the function generator.  This is a bit surprising because you have talked about it many times. 

Paper 1 shows the black lead of the function generator connected to the Q1 source pin side of the current sense resistors.  As you have said a number of times: because their function generator black lead connects to the function generator mains power green safety earth lead, and because their oscilloscope channel input ground leads all connect to its mains power green safety earth lead, had they connected their equipment as shown in Paper 1, Figure 1 as they said they did, they would have shorted out the current sense resistors through a large ground loop.  The function generator black lead on the Q1 side of the resistors would connect through the function generator green earth wire to the oscilloscope green earth wire and back to the channel input ground wires both connected to the battery side of the current sense resistors. 

Donovan Martin said during the June 29 video that they had at least then connected the function generator black lead instead to the battery negative connection on the white breadboard.  His comments may be verified by observing the location of the black function generator lead.  Moving the lead from the position documented in Paper 1 eliminated the ground loop, but also bypassed function generator current around the current sense resistor.  Your comments that they could not and did not measure the current contribution from the function generator is exactly correct.  Ms. Ainslie vociferously argued that there was no such contribution until the demonstration August 11 proved her wrong.

With the help of outside advisers, including poynt99, Ms. Ainslie and her collaborators set-up the August 10, August 11 tests as per Paper 1, but broke the ground loop by disconnecting the green safety lead at the function generator AC plug connection.  At poynt99's suggestion, for one test, they inserted an additional current sense resistor between the function generator black lead and the Q1 source side of the white breadboard CSR bank.  That resistor allowed them to measure the function generator contributed current isolated from Q1 and Q2 currents. During the Q1 OFF / Q2 oscillation period, they found 250mW to 400mW power supplied by the function generator. 

Another important observation that resulted only because Ms. Ainslie and her collaborators cooperated with poynt99 was their finding that where they had previously thought that their circuit was recharging the battery during the Q1 OFF / Q2 oscillation period, the reality is exactly the opposite:  During the Q1 OFF / Q2 oscillation phase, the battery supplied power to the circuit.  Presently, Ms. Ainslie declares:

"But the loss of energy from the battery during the 'OFF PERIOD' has NOT BEEN PROVED.  ONLY PROPOSED."

If Ms. Ainslie would review the video of her August 11 demonstration, then she can see for herself the battery power draw she and her collaborators measured during the Q1 "OFF PERIOD".

Again, Mark E., thanks for that expanded summary. The demo where the ground pin was removed was of course not quite public; I was specifically excluded from watching it. Ainslie clearly doesn't understand the groundloop issue, even if Martin might. It's not possible for me to know just where or how hookups were made or what the results were, but reports are that the current from the FG was indeed easily measured during this part of the demonstration. Ainslie remains muddled about a lot of things. For example, it was shown BY ME, long ago, that a negative bias voltage supplied in lieu of the Function Generator would produce continuous Q2 oscillations. The ACTUAL process by which Q2 is turned on and made to oscillate with the supply of a negative bias voltage is detailed in the video series that I posted a couple of days ago. In addition, the current from the FG was measured, long ago, and the issue was fully explored by me and by .99 in 2011 and 2012, long before Ainslie could even admit that the FG could act as a current path or source for the circuit. Continuous Q2 oscillations are very easy to produce by using a bias battery or power supply. No FG is needed... all that is really needed is a proper understanding of how the circuit operates.

http://www.youtube.com/watch?v=s8_ZTBtyTvo (uploaded to YT on July 14, 2013)
http://www.youtube.com/watch?v=8PitNm44_bE (uploaded to YT on June 30, 2013)

Note again that Ainslie is claiming that the circuit that is in the Quantum "box" NOW, which Steve Weir diagrammed and which I have tested, has something to do with the 2002 report in Quantum magazine (hardly a "peer reviewed" journal). She still has not confronted the issue of the 555 CHIP manufactured in 2007 that could not have been in the box for the Quantum article, nor the issue that the present "box" circuit operates at a much higher frequency than the article claimed, AND it uses a completely different mosfet, AND it only has one variable potentiometer where the Quantum circuit has two. It is simply another of Ainslie's trademark mendacities to pretend that the "box" circuit has anything to do with the earlier Quantum article. In fact the "box" circuit was built sometime AFTER May of 2007, to make the frequency and dutycycle combo that Glen Lettenmeier found in his work, not the other way around.

Ainslie is in a rather peculiar position in that she pretends to be able to criticise my demonstrations... yet she does not bother to watch them, or ask her "team" to watch them and "debunk" them. At one time she promised to take them each and tell us specifically what she finds wrong with them. I would love it if she would fulfil that particular promise. It would be high comedy indeed.

(Go ahead... ask her to provide a link to the "replication" that she claims was performed by "joit". But don't hold your breath waiting for it. )

TinselKoala you are welcome.  Video from the August 11 demonstration is available among other places on Rosemary Ainslie's blog.  They made a point of showing that they had disconnected the function generator green wire.

As far as I gather, Ms. Ainslie does not believe that it is important that she did or ever will publish the exact circuit used for the experiments she reported on in the 2002 magazine article.  Were reasonably accurate timings known, then the details of the timing circuit would not be very important to the truth or falsity of the claimed discovery.   Unfortunately, to the best of my knowledge Ms. Ainslie never published oscilloscope shots of what they measured when preparing that article.  This leaves everyone guessing as to what Ms. Ainslie and her collaborators actually observed.

Of late Ms. Ainslie expresses the view that only the duty cycle is important, and not the pulse frequency.  Seeing as that people have tested at the duty-cycle and frequency combination reported in the magazine article and did not see the gains that she reported, then either the both frequency and duty cycle really are important, or as with the Paper 1 and Paper 2, the gains that Ms. Ainslie and her collaborators reported were the result of measurement errors.  Ms. Ainslie insists that cannot be the case because she says that independent experts repeated her reported results.  Unfortunately, none of these independent experts have ever stepped forward.  The recently retrieved test fixture does not help much because without making permanent wiring and component changes, it cannot be configured to match either the schematic in the magazine article or to produce the timing stated in the magazine article.

The argument for measurement error is very strong.  You have shown that the combination of pulse frequency, but inverted duty cycle is attainable using the circuit depicted in the magazine article.  As you have shown that would account for much of what was reported in the magazine article.