Rosemary Ainslie Quantum Magazine Circuit COP > 17 Claims

[MarkE]

Ms. Ainslie offered some explanation of what she thinks she is trying to do today.  She and her collaborators have ordered some unspecified data logging equipment for their new experiments.  With that Ms. Ainslie repeated her confusion as to what transpired as is clearly visible in the test script that she published and the video of her August 11th demonstration.  Ms. Ainslie now contends that the voltage and current measurements taken at the batteries did not reflect actual battery power drain.

What was measured was the evidence of a current flow and voltage from the battery during the 'on' period of the duty cycle - that exceeded the level of wattage dissipated at the resistor.

Notwithstanding the evidence of a current flow from the battery - there is NO corresponding evidence of that voltage over the element resistor during that 'on' period.

First:  During the August 11, demonstration, the function generator output voltage never rose above the 4V Vgs MOSFET gate threshold voltage necessary to start conduction through the Q1 MOSFET.  In the parlance of Ms. Ainslie's publications, there was no "On" period.  Neither measurements at the white breadboard nor at the battery indicated power drain from the battery, nor voltage applied across the heater resistor during the Q1 "On" periods.

Second:  During the demonstration oscillatory current flowed through the Q2 MOSFETs during the Q1 "Off" period.  Measurements at both the white breadboard and at the battery indicated power drain from the battery, oscillatory voltage applied across, and oscillatory current flow through the heater resistor during the Q1 "Off" period.

Third:  Real power of approximately 15W was shown drawn from the battery.

Fourth:  Thermal output of the heater resistor was shown to be about 3W based on Ms. Ainslie's prior calibration of heater resistor temperature rise over ambient versus DC power applied.

The circuit is quite simple:  Battery+ => heater resistor => "QArray" MOSFET's & function generator => breadboard current sense resistor => battery side current sense resistor => Battery-

Current through the heater resistor passed through the battery which the oscilloscope measurements showed.   So we have a very simple situation:  ~15W from the battery was dissipated by the other elements in the loop.  Of those only about 20%: ~3W was dissipated by the heater resistor, while the function generator actually contributed a small amount of additional net power:  < 0.4W.  The remaining ~12.4W was dissipated by the "QArray" MOSFETs, the wiring and the current sense resistors.

Ms. Ainslie claims an anomaly of battery drain during the Q1 "On" periods, when the oscilloscope readings show no such thing.

The captures below were taken from the video available on Ms. Ainslie's blog site.  The first capture was taken during Phase 1 of the test.  The channel 1 current sense is clearly flat lined during the periods when the channel 3 gate drive voltage is positive.  The gate drive positive voltage during this time is about +2V, which is not enough to cause Q1 to conduct, and it does not conduct.  However, when the gate drive voltage swings negative, oscillations occur.  The Q2 MOSFETs oscillate wildly, the battery voltage as seen on the white breadboard in channel 2 swings hundreds of volts, and the current sense voltage: channel 1 indicates many Amperes swing.  The second picture from test phase 3 shows a pattern of: zero current, zero voltage excursion followed by a burst of current oscillation an voltage oscillation that resembles the first capture in basic form. The periods of current flow at the breadboard measured in Phase 1 correspond to the periods of current flow at the battery measured in Phase 3.  Ms. Ainslie is confused as to what her own demonstration clearly shows.

[TinselKoala]

Ainslie's typical mental state may be described as "overconfident confusion". She is confused about many things, and doesn't know how to construct experiments that would resolve her confusion. Or rather... she avoids constructing and performing such experiments, since she likely knows full well what the results will be.

For example it is trivial to demonstrate that the Q1 "on" periods which only have a Gate drive of 2 volts or less contribute ZERO to the heating of the load resistor. Simply use the FG's duty cycle control to extend the "on" periods to their longest possible durations, which of course will reduce the Q2 oscillation periods to very short portions of the period. Or apply the +2 V gate signal using a continuous source like a power supply. Surprise surprise.... when the Q1 mosfet only gets a +2 volt gate signal it STAYS OFF and the load DOES NOT WARM UP unless the Q2s are oscillating. In this situation only the Q2 oscillations contribute to load heating through the partially ON Q2 mosfets.... if they are removed or prevented from oscillation, there will be no heating of the load, because the Q1 mosfet will not turn on at the +2 V gate level.

But after all.... Ainslie claims, in the Figure 3 scopeshot that has STILL NOT BEEN REMOVED OR CORRECTED, that the Q1 mosfet received 8 volts or more at its Gate without turning on and passing current. The fabricated, fraudulently obtained data still remains in her daft manuscripts and still makes up the major bit of "evidence" given in attempted support of her ridiculous, mendacious claims.

And furthermore, even after all this time and her own demonstrations and the instruction given by her own partner Donovan Martin, not to mention my own very clear and very basic demonstrations .... Ainslie still cannot comprehend the linear operation behaviour of a mosfet. She has always modelled the mosfet as a simple switch that is either ON with zero resistance or OFF with infinite resistance, and she has never given any indication that she comprehends this functioning at all. She has no clue what "Rdss = 2.0 Ohms" means, nor can she read or interpret any of the performance graphs that are given in the IRFPG50 power mosfet data sheet to note that the Drain-Source resistance increases drastically when the mosfet is not receiving a full Gate charge.

Really, it's futile to expect any kind of rational analysis from Ainslie, because as we have noted time and time again, her distortion of reality extends to the blatant contradiction of actual experimental results and data, even the results and data obtained by her own crew. She cannot even describe accurately events of last week, much less those from the summer of last year, and to expect her to give accurate descriptions of what happened  twelve or fourteen years ago is utterly futile. Look at the way she distorts, misrepresents and even lies about the demonstrations of June 29 and August 11, 2013. Nothing she says can be trusted, without objective proof.

I did a little research on inexpensive data logging. It turns out that a simple Arduino, with a tiny bit of peripheral circuitry and a little programming, can handle all the data logging requirements of this experimental paradigm. Comma-delimited text files containing data from 12 digital and 5 analog input lines can be written directly to the computer over the USB interface, and this data can be imported into the usual spreadsheets for analysis and display, or the program "processing" can be used to graph the Arduino data directly, and of course to control and automate the data logging parameters.  The addition of the Adafruit DataLogger Shield (20 dollars US) to the main Arduino Uno (30 dollars US) and a micro SD card enables the Arduino datalogging system to "stand alone" and operate without connection to a computer; the Adafruit shield contains a Real Time Clock and can take samples of the input variables at just about any sample interval desired. For under 100 dollars, a complete Arduino-based voltage, current, temperature, and time datalogging system can be constructed that will monitor all six or eight necessary parameters with sufficient accuracy to test all the experimental hypotheses (yet again) that arise. Do I think Ainslie will go this route? Of course not ... but it may be of use to some other researchers who might need to do long-term data logging on a budget. And when one is tired of chasing invisible talking rabbits across Ainslie's fairy-tale quagmire, one can use the Arduino for other interesting things.
http://learn.adafruit.com/adafruit-data-logger-shield/overview

Or one could spend thousands of dollars for a LabView ADC-DAQ system and a current LabView program installation, and then hire a technician to spend hours and weeks learning how to program it.....  Whatever system the Ainslie mob decides to use it will result in long delays before any data is obtained.... or released. Recall that we STILL do not have the screenshots from June and August that they agreed, publicly, to release. And this is just the most recent bit of data that Ainslie has promised, and then not provided!

[MarkE]

In the Figure 3 case,  I attribute the false statements to human error and not design.  I think that Ms. Ainslie and her collaborators first learned that they generated Figure 3, Figure 6, and Figure 7 of paper 1 by connecting the Channel 1 scope probe tip to the low side of the current sense resistors on June 29, 2013 when they were guided through measurements by Steve.  Watching the video they were very frustrated and anxious when Poynt99 asked them to set the function generator amplitude and duty cycle as in Figure 3, but increase the frequency so that 27 minutes would not be required to capture a scope trace.  Ms. Ainslie attempted to negotiate a lower level than 12V on the gate:  first 8V and then a lower value, before diverting off into discussion of the excess heat generation that she claimed for the apparatus.  This suggests to me that they set-up shortly before going live and discovered they could not reproduce Figure 3.

There are all kinds of datalogging kit available at all price ranges.  I don't care how much they spend.  I care if they put together reasonable tests capable of answering the questions they pose.  I believe that they want to do that.  We won't know until they show what they set-up whether they can obtain reliable and meaningful results or not.  I am convinced that Ms. Ainslie believes her ideas are correct and that she believes that experiments will bear out her ideas.

[TinselKoala]

In the Figure 3 case,  I attribute the false statements to human error and not design.

I attribute the initial collection of the Figure 3 data to human error, not design. However, it is clear from the record that _subsequent_ scopeshots were obtained that did not include the error in connections and which produced valid data. This means that the initial error was recognized and corrected, things that do not happen unconsciously. The conclusion that Ainslie and her collaborators knowingly included the false Figure 3 and other scopeshots is inescapable, no matter how charitably one looks at the issue.
What's the first thing that any experimentalist does when data is obtained that violates all known precepts of the applicable theory? She _REPEATS_ the work, with careful error checking along the way, and submits it for external examination to double-check for errors. Ainslie herself may be so gushingly excited about such a result as to accept it at face value. Is Donovan Martin also so naive and self-serving?
If one wants to posit that the Figure 3 and other similar shots are not deliberately included in spite of conscious knowledge that they are invalid... then one must accept the alternative explanation that everyone whose names appear on those daft manuscripts are utterly incompetent, naive and credulous tinkerers without even the most fundamental grasp of how properly to use their own test equipment. Or, as present events have illustrated.... BOTH.

I think that Ms. Ainslie and her collaborators first learned that they generated Figure 3, Figure 6, and Figure 7 of paper 1 by connecting the Channel 1 scope probe tip to the low side of the current sense resistors on June 29, 2013 when they were guided through measurements by Steve.

Again, subsequent scopeshots to the original Figure 3 and the others you mention did not include the wiring error, so it must have been noticed and corrected before those later shots were obtained, and this does not happen through the secret agency of menehune or other Little people in South Africa. Ainslie herself or some member of her team must have hooked things up more correctly _after_ the questionable shots were obtained.  I find it perfectly plausible that Donovan Martin didn't know -- because Ainslie may not have told him about the wiring error -- and I find it perfectly plausible that Ainslie didn't _understand_ the significance of the wiring error. However I find it implausible to the extreme that Ainslie didn't KNOW that the connections were as they were, and I find it a remarkable expression of incompetence that Martin did not CHECK but simply took Ainslie's word for it that this impossible result had been obtained.

Watching the video they were very frustrated and anxious when Poynt99 asked them to set the function generator amplitude and duty cycle as in Figure 3, but increase the frequency so that 27 minutes would not be required to capture a scope trace.  Ms. Ainslie attempted to negotiate a lower level than 12V on the gate:  first 8V and then a lower value, before diverting off into discussion of the excess heat generation that she claimed for the apparatus.  This suggests to me that they set-up shortly before going live and discovered they could not reproduce Figure 3.

I don't think they "knew" until the live demo itself. It's apparent to me that Martin didn't really even understand the issue until Steve made the point that it was the 8 volt or greater Gate drive, in combination with zero indicated current, that was the problem. Sure, zero current can be obtained easily enough, and more than 8 volts gate drive... but not both at once.

There are all kinds of datalogging kit available at all price ranges.  I don't care how much they spend.  I care if they put together reasonable tests capable of answering the questions they pose.  I believe that they want to do that.  We won't know until they show what they set-up whether they can obtain reliable and meaningful results or not.  I am convinced that Ms. Ainslie believes her ideas are correct and that she believes that experiments will bear out her ideas.

Indeed. The LeCroy 324 itself, being GPIB/TCP-IP/USB remotely controllable, can be used for long-term data logging of all its channels. This functionality requires knowledge and skill to use, though.
http://cdn.teledynelecroy.com/files/manuals/wj300a-rcm-e.pdf
As to the issue of what kinds of "reasonable" testing Ainslie wants to perform... she has had good advice about testing since 2009, at least. Nothing that has been proposed lately is new. Had she genuinely wanted to perform tests of real hypotheses concerning her "thesis", she would have done so long ago. In fact, she did send off her entire apparatus (perhaps not including the batteries!) for independent testing by a laboratory in the USA a couple of years ago, back when she was claiming that her batteries did not discharge. The most she has reported about these tests is that they did reproduce her claimed "negative power product"... which anyone can do, even using analog scopes, as I have shown... and that they found that her batteries DID discharge normally, and that they sent her some "special resistors" and gave her advice about running better tests of her own using these special resistors. And that is the last we've heard of it.
No, Ainslie has demonstrated amply, over and over, that she is not interested in performing tests that have the ability to falsify her claims by the collection of appropriate and valid data. If her attitude is suddenly different.... well, it occasionally does rain, in every desert, if one waits long enough, I suppose. But I'll believe it when I see it.
I am convinced that no amount of experimental data will shake Ainslie's belief in her claims and her silly "thesis". Data that doesn't support the claims is simply wrong. She may not be able to tell you coherently why or how it's wrong, but if it contradicts her.... it's wrong, because her "thesis" is right, QED.

[MarkE]

I find it baffling that if they found the probe connection problem and recognized it for what it is before the June 29th test that they were:  1) willing to perform that demonstration, and 2) so befuddled when the day came.  What we do know without question is that as of June 29, they knew that Paper 1 was based on bogus tests.  August 11 only served to fully reinforce that fact.