Hi all,

I've been on other things for the last week and have just caught up. I'm glad the other thread has finished because it was going nowhere.

As Poynt, MH and I have emphasised to Rosemary, build quality of the test circuit is of utmost importance as well as the correct equipment. At least Aaron had the sense to bail out in the early stages having publicly displayed his ridiculous mess of a test circuit!

It amazes me that Rosemary had the audacity to make a COP17 claim for a circuit that she clearly could not accurately draw because she had no record of the exact circuit and build details of the original setup testing by her 'experts'! As Paul said at the beginning, a claimant must be seen as genuine and sincere before time and effort is spent by others in investigating that claim. Her failure to produce full circuit and build details at the outset should have been ample justification to terminate her forum thread way back.

It now only leaves the AT to cobble together a submission to the real 'experts', which I really cannot see happening.

Hoppy

I'm not discounting the calorimetry method at all for these measurements, however it would be rather difficult to obtain one for the power being supplied to the circuit by the source.

The only method available to obtain a POS (power output from supply) measurement, is through the use of either a shunt resistor (filtered or unfiltered) or a current probe, and the source voltage (filtered or unfiltered).

This is the crux of the argument regarding the AT's results. Their POS measurements/calculations are not based on good data. Proof of the Ainslie claim only requires the POS and PIL (power into load resistor) measurements. The PIL measurement was obtained using a DC control test and came to about 4.5W. That leaves the POS measurement, and thus far the AT has not produced an accurate one.

The means to do so is not only well within reach, but has been repeatedly spelled out for them. Their refusal to utilize the suggested methods to obtain the accurate measurements required, is a sure path to not only rejection of their paper (if they in fact do submit one) but perhaps even public ridicule after the fact.

It's unfortunate that stubbornness, pride and delusion reign over common sense in this case. Since the AT refuses to listen to anyone that questions their path, perhaps the true wake-up call will only come when the "no thanks" response to their paper fully sinks in.

.99

Most Likely the AT is anticipating this post, so rather than disappoint...

Evidently they aren't going to admit it, but quite clearly the Ainslie Team (AT) has finally implemented part of what I have recommended as an approach to improve the integrity of their measurements. This is evident in two tests that Fuzzy has performed, namely #12 where all the probe grounds were relocated directly to the bottom on the shunt resistor, and in test #13 where the shunt probe tip was moved from the MOSFET source directly to the top of the shunt resistor.

Notice in comparing the three scope shots the marked decrease in the amplitude of the negative shunt voltage excursion and the related spikes. This is precisely what I had predicted would happen if the shunt inductance was reduced. Moving the probe tip and ground lead connection points directly across the shunt resistor (as I show in my diagram as a first approach improvement) has done just that.

So they are apparently heeding at least some of my advice, even though when first posted, rather than show some gratitude, specific effort was expended to denigrate and insult, and to trivialize my contribution. Nonetheless, making this small change is progress on their part.

If they were to implement the second part of the "First Approach" improvements by sampling the battery voltage through a separate sense lead (red sense lead in my diagram), they would find that the false elevated battery voltage indicated by the Green trace in Fuzzy's scope shot, would commensurately be reduced by about 40% (I estimate). The remaining voltage bump is present because the black return wire to the battery represents a minimum of 60% of the total inductance which is still a problem. If point P5 can be successfully utilized, the decrease could be much more significant, perhaps as much as 90%.

Moving forward to the "second approach" improvement would allow almost the complete elimination of the false battery voltage bump.

The result of minimizing these false battery voltage and shunt voltage levels, is to make the measurements more accurate, and bring the power calculations closer to reality.

There is much more detail which has been exchanged with my contact at Tektronix, but what is included in this short document is more than sufficient to explain the problems and correct solutions to get the right measurements.

http://www.overunity.com/index.php?action=downloads;sa=view;down=348
(click link and see download link at very top of page, or click here: http://www.overunity.com/index.php?action=downloads;sa=downfile&id=348)

.99