Rosemary Ainslie Circuit Demonstration June 29, 2013 Video Segments

[poynt99]

A: "I'm telling you, it doesn't matter. Nobody "out there" really cares, except that Ickle Pickle Bryan Little fellow, whoever he is.

Mossfeet, or misfoots or something, I don't remember. Expensive little slappers, anyhow, and she pops them like popcorn."

[TinselKoala]

I believe I have finally located the source of Ainslie's "academics".

[TinselKoala]

I am sorry, I really am. The more I look at the video demonstrations and the current apologetics, the more I see the Donovan Martin-Rosemary Ainslie team as objects of ridicule rather than anything seriously to be considered. The very concept of an experiment seems foreign to them: Vary a variable that the experimenter can control and observe the effect of its variation on a variable that the experimenter can observe, while holding constant, randomizing or otherwise controlling for effects of third variables. That's the process, and the object should be to select experiments and variables that have some chance of showing the experimenter's guesses are wrong, not that they are right. It took them at least four hours to get to the point of taking six proper experimental data points (the scope captures that S.Weir talked them through), and that bit was done in fifteen minutes and would have taken even less time if Ainslie and Martin would have just shut up and cooperated.

[MileHigh]

An amusing observation about their hit rate.  I am not an Internet nerd but it looks to me that the legacy of the "bot swarm" that logged into her site way back when is the continuous pinging of her web page.  It's like there is an army of bots out there trolling the Internet looking for email addresses and stuff like that.  The bot swarm is aware of her site so they will be knocking on her door from now to kingdom come.

They tell me you can get 10 million email addresses for 20 bucks!  lol

[TinselKoala]

She's posted her actual email address publicly several times. I can just imagine what her inbox must look like.



Anyhow.... I just did a Test-to-Failure of the Q1 mosfet. Briefly, I set up a Gate input drive as slow as I could get with the F43 and a bit under 20 percent duty cycle, as close as I could get to the proportions in the Fig 3 shot, but of course not quite as slow. Her period was 160 seconds and mine was about 50 seconds, I think, with about 10 seconds HI per period.

ETA: I just checked and my duty cycle was a bit longer than hers. Sorry.... but my heatsink was better !

I started at ambient temperature and allowed the system to run, recording temperatures during the non-oscillating Q1 ON times. The temps recorded during oscillations are not reliable. Vbatt was 75 volts, about, no load and dropped to around 70 volts when the Q1 was fully on, and the inline ammeter  showed around 6 amps during those times. The gate drive was as shown in Fig 3, +12 V for the Q1 ON times and -4 V with fuzz on the bottom for the Q2 oscillation times. The mosfet lasted a little over 11 minutes and failed _shorted_, not open; the current trace went to 5 amps or so and stayed there and the mosfet began to +cool off+ even though it was carrying all that current. Its Rdss actually dropped, due to the internal short!  I'm sure that it would fail open if I kept running for a few more minutes but my load cell was already pegging its analog thermometer!

Anyway... I wasn't able to get the flatline of Fig 3 probably because I stopped too soon. But it only took 11 minutes for my "waveform to resolve itself" and produce a blown mosfet.

I captured all of this on video of course, including the post-mortem test showing the mosfet was indeed fully shorted. Processing and uploading now, should be ready in a couple of hours max.