Rosemary Ainslie Quantum Magazine Circuit COP > 17 Claims

[poynt99]

For those that may be interested. A pic from Rose's latest test.

[TinselKoala]

Thanks, Poynt99, for providing that, along with your annotations.

Interesting. I am "assuming" that the color/channel assignments are the same as Ainsile typically uses:
Yellow: voltage drop across the CVR ( which? The "noninductive" arrangement on the demo board?) This channel is displayed at 1.19 Volts per division !! Aren't Digital Scopes wonderful! You can set them to make the traces as difficult to read as possible!
This is clearly obfuscatory. It is impossible to read the actual level of the Current from this trace, and even the Numbers In Boxes are useless, displaying the "mean mean" and the max and min "mean" or average values of this trace.
Blue: Gate signal from FG (or 555 or other source?) at 50 volts per division! Got something to HIDE, Ainslie?
Purple: Vbatt (from where? the noninductive battery connection shown in August, or at the board itself?) and at 100 V/div to capture the peaks.... and only the peaks are displayed in the parameters box, not the much more useful _average_ or true minimums and maximums. It's not even possible to tell how many batteries were used. The peak current of just under 2 amps would seem to indicate that only two batteries were used but since the mosfet is still in its linear operation region, not turning fully on, perhaps 3 batteries were used and the circuit resistance is just higher than normal.

Red: Math trace making the spurious calculation and displayed at 500 v/div. Useless. But it shows negative values! Miracle of Mismeasurement!

I note that the scope's timebase is set to 1 microsecond per division and that the operating frequency is about 187.2 kHz, as confirmed by the numbers in boxes. The Gate drive duty cycle appears to be about 25 or 30 percent HI. The mosfet itself is not fast enough to properly turn fully on and off at this frequency so it's not showing that much real "ON" time itself.

So once again we have a garbage scopeshot. The Current trace displayed at 1.19 V/Div, with "max and min means" displayed !! The Math is showing a spurious value due to the attempts at integrating the spikes.... from non-deskewed probes, most probably, without any filtration. The Gate trace is displayed at 50 v/division so that the total displayed amplitude is less than one full division -- not very informative. The math trace at 500 V/div is just silly. This is done to get the spurious peaks to stay on the screen. Good luck getting proper Math values if any portion of the trace is offscreen vertically.

So once again we have a scope screen that is more obfuscatory than informative, and the information that is actually needed must be extracted painfully by interpreting trace positions with odd scale values, rather than by using cursors or properly obtained parameter measurements.

I also note that the Gate drive signal is the triggering channel and the trigger is set to -24.0 V !! The Blue Gate drive signal is set to 50 Volts/division ! The Gate signal appears to swing about 20 volts p-p and has the extreme negative offset that her FG is capable of.  Curiously.... the gate signal risetime is a bit under 400 nanoseconds. This seems very slow to me for a "square" wave pulse. I can attain 10 ns or better risetimes at that frequency. Is Ainslie using the triangle ramp setting again, as she has tried to use before?

Will someone PLEASE teach these people how properly to display INFORMATION on an oscilloscope? I declare: You get more real and usable and valid information from my ANALOG SCOPES and my narrations, than you do from Ainslie's Etch-a-Sketch LeCroy abuse.

[TinselKoala]

Uh-oh.... now she's threatening to cut off Poynt99 from her "open-source" project.

Meanwhile... she still can't get her own story straight. Did Poynt99 imply anything by posting the scopeshot? Does she "stand by" the Math trace waveform, or does she SPECIFICALLY state it to be "irrelevant?"

You decide. And think about this: just who is _actually_ performing "Ainslie's" experiment, actually, anyway? Not her, certainly. Can you imagine the incompetent and ignorant Ainslie programming the scope, running a trial, saving a data file to the USB stick, then transferring that into a spreadsheet for analysis? Yeah... sure she did. Right.

[MarkE]

On the Secret of DPDT: It must be a secret of the most secret kind, since I've encountered several projects that would have been a lot easier had they used such a switch. The Steorn Waterways Demo was a great example. When they wanted to show that reversing the polarity of their coils didn't affect the functioning of their Orbo pulse motors, they spent 5 minutes with a screwdriver to swap the coil connections and then another 5 minutes swapping them back, and of course the motor had to be stopped while doing this. The Secret of DPDT was revealed to the lads by me in a video at that time. More applications have followed, such as the rather unorthodox but quite viable and easy application of the Secret to the Ainslie Q17 circuit.

But just for completeness, I'll cobble together an instantiation of your dual-555, or 556, circuit and have it available for testing alongside the Q17, FTC, GreyBox and other 555 circuits I've used to clock the basic mosfet switch.


Speaking of testing, does anyone know of any complete and _valid_ data set, anywhere, that includes a correct and complete schematic, the operating parameters, a waveform shot or two, and temperature/time data for the load heating, that shows any of Ainslie's (or even GMeast's) claimed OU heat effects? COP >17, COP INFINITY, COP 3, 300 percent OU, twice, or a teeny bit OU .... whatever? I don't think I've ever actually seen any comprehensive heat data that connects all the dots, except in my own work from 2009/2010 and the past few weeks.

I see.  Steorn were such goofballs that it didn't dawn on them to use a simple two pole toggle switch or a switch and a relay.  What dopes.

I started a thread that discussed Greg's tests and where I think he should go next.  He didn't think much of my suggestions.  It's too bad, because he really does seem like he is trying to set up decently controlled experiments.  I am unaware of anything better from Ms. Ainslie than the tests last summer.

[MarkE]

For those that may be interested. A pic from Rose's latest test.

At 500V*V for the math trace, and a 0.25 Ohm CSR the math trace is 2000W/division.  The power consumed is below the display resolution.