Rosemary Ainslie Circuit Demonstration, June 1 2013

[fritz]

Riding a dead horse.
Too much fun.

[TinselKoala]

TK,

Although recently FIG 3 has been referenced to with regard to Q1 not turning on when it should, as discussed in the past, it is actually FIG3, FIG6, and FIG7 in the first paper that demonstrate a Q1 that is not turning on when it should.

PW

Fig 6 is SCRN0354, made on April 30, and Fig 7 is SCRN0355, made just under an hour later. I'm not sure about these shots the way I am about Fig 3, SCRN0253; I haven't blown them up for careful analysis. These might be showing just barely threshold activation of Q1. The shot SCRN0342, taken about an hour and a half before Fig 6, definitely does show current in the Q1.
Then the record skips, but I have two more shots, SCRN0361 and SCRN0362 from the 8th of May, neither of these show Q1 current but they also omit the gate drive trace altogether so they are not interpretable.

ETA: In any case I am not very concerned about Fig6 and Fig7 because they are taken at a much faster timebase setting than is Fig 3 and are not likely to have been the settings that actually blew the mosfet, especially since they only used 5 batteries. But Figure3, with its 160 second period and 16-17 second ON times, and the six batteries, just might do it. So this is the one I am focussing on and the one I am challenging Ainslie with.

[TinselKoala]

I'm posting this video link again because I think I buried it under the "timeline" reconstruction posts. This vid shows the reproduction of the Fig3 scopetraces much better than the prior one.
http://www.youtube.com/watch?v=nbkpQQvuP2I

I'm tempted to make yet another video where I use all six batteries, the long on time, and monitor the temp of the mosfet until it fails. My supplier charges six dollars for the silly things though.

[poynt99]

Good job on the video TK. It certainly seems to support the hypothesis that Q1 was blown open or disconnected for the Fig.3 scope shot.

The placement of the FG neg lead does have an effect on the CSR trace, which is why I noted the change on my markup of the setup (attached). I noticed that on the PESN photo there is no gnd lead on the Vbat probe, so I did not show one. I have posed a question to Rose about that and will update it if necessary.

Anyway, this is what I have thus far, awaiting Rose's approval as to its accuracy.

[poynt99]

Once again, here are two sim scope shots with colors adjusted to match those from Rose's scope shots. The first is with Q1 intact and operating properly, the second is with Q1 removed from the circuit. This is with +/- 14Vp FG drive.

Notice the absence of the inductive kickback spike on the pink trace when Q1 is removed.