Claimed OU circuit of Rosemary Ainslie

[fuzzytomcat]

Thanks Fuzzy.

What's the plan to analyse the data? Is the MOSFET power significant that it needs to be subtracted out? A control test with the same load for comparison?

.99

Hi .99

Right now I'm in testing mode to get as much data as possible .... if data is missing or needed I'm leaving it up to other members to comment on if its reasonable request it will be done in time. Possibly ... some members at Energetic or OverUnity can crunch some data until testing is completed and I have some time prior to returning the TDS 3054C on loan to me from Lisa at Tektronix Corporation.

I hope you may have seen the earlyer postiing of the Mosfet timing of the source - drain - 555 timing  for comparison "with" the great 220 volt spike from the Mosfet drain ..... and also the 555 duty cycle - frequency

http://www.energeticforum.com/69850-post2875.html

Fuzzy   

[MileHigh]

Glen:

Congratulations!  You did a great job with your setup and your data logging.  Thanks for hanging the big resistor in mid air, now I can say that I helped contribute something and I am not a troll.  lol

I just spent the past 1 1/2 hours contemplating the waveforms for the switch-off ringdown.  Thank's to Harvey for asking for this.  Those are the most interesting waveforms seen so far in this experiment.

I think that I figured it out, I will probably make a posting later today.  The explanation that I worked out in my poor taxed brain is not trivial.

Rosemary:

Keep hoping for that COP!  My eighth impression is that everything looks normal.  In other words, it was startling to see the big reverse current spike lined up with the big positive voltage spike in Glen's waveforms, it looks like that should not be happening.  So I had to iterate over and over in my head to finally figure it out.  That's all part of the fun!

Paul:

I ask you to contemplate the oil bath method some more.  Have you ever actually done it?  I can see oodles of issues that would turn anyone into a frustrated oily mess trying to pull it off.  Did you see Glen's amazing hanging "gallows" coil-resistor setup?  KISS!

MileHigh

[MileHigh]

Well, about Glen's first waveform posting:

http://www.energeticforum.com/69850-post2875.html

I read Harvey's explanation, and I think that he is more or less on the right track at the beginning, but then I loose him in his last five or six sentences.  I can't quite understand what he is saying, so I can't say I necessarily agree or disagree with him.

Let me just share my thoughts, I am reasonably confident I am on track.

In the second scope capture, you can clearly see a more or less exponential decay envelope for the switch-off.  Just ignore all the ringing for a second and you can clearly see it.  The coil is nearly fully charged after four microseconds, it's a dumb resistor for two microseconds, and then it has a wild-ride discharge over the next four microseconds.  Note that you can see the little positive and negative "dv/dt" glitches, which are due to the gate-souce capacitance transferring small current spikes through the SHUNT resistor when the 555 signal switches.

By seeing these current spikes I can make an inference about the affect the gate resistor had on the 555 output rise-time.  It looks like it has essentially no affect on the 555 signal (or you could say that it did it's job and preserved the integrity of the 555 signal) such that the MOSFET gate input saw a nice clean square wave.

If what I said above is true than you can see that the MOSFET has about a one microsecond delay before it really starts to switch off.  I am not up on my MOSFET timing stats, but I am surprised a bit, it sounds really slow to me, with the caveat that I am not familiar with typical MOSFET switch-off times in general.  I can say that gate-level logic nowadays has switching delay times on the order of a few nanoseconds or less!  (I think! lol)

So the "big picture" is as follows:  When the MOSFET switches off the coil-resistor smacks the MOSFET, and then the MOSFET source-drain and related stray capacitance forms an LC tank circuit with the coil where the initial conditions find the total capacitance charged to a very high voltage.  It is very important to state that the "ground" for this LC tank circuit is the MOSFET source pin, and not the battery ground.  This rings down right through the MOSFET in both directions.  In the counterclockwise direction (initial conditions) the body diode of the MOSFET conducts.  In the clockwise direction, it is the really slow switch-off time of the body diode that acts more or less like an ordinary resistance.

The ringdown is happening too fast for the body diode to shut off after the counterclockwise current has stopped and the current goes to reverse direction.  So the clockwise current just plows through the body diode in a "reverse-biased" sense, the body diode simply being too slow on the draw.

What does this all mean:  It means the MOSFET does actually switch off, but then gets shaken like a rag doll by the LC tank circuit that is initially charged to about 230 volts.  The LC tank circuit (a.k.a.  LC resonator (Hey!  I said "resonance!" lol)) simply plows through the MOSFET, and the MOSFET is unable to effectively switch off.   The ringdown is not pretty in this case, because depending on which direction the current is flowing in, that determines which one of two nonlinear loads is acting as a resistive element dissipating energy.

Another way of putting it is that after the initial shut off of current done buy the MOSFET, it is all body diode action after this.  The body diode alternates between being forward-biased and conductng, and being reverse-biased and failing to not conduct, letting the current pass through it in the "wrong" direction.

How about them apples!  We just did some MOSFET switching transient analysis!  I could be wrong or missing something here, but that is what I think we are seeing.

This posting was not about investigating the COP claim, just understanding what is going on during the wild-ride ringdown.  However, it clearly explains what the source is for the big negaitve (a.k.a. counterclockwise) current spike that we are seeing across the current sensing resistor.

The source is...  the battery bone connects to the coil bone, and then the coil bone connects to the drain-source capacitance bone.  The battery bone is the souce for the big negative current spike.

MileHigh

[poynt99]

Glen,

Just playin' around a bit. I wanted to see if I could replicate your first group of scope shots (post #EF2875) with the 37kHz, 22% duty cycle drive. Think I got pretty close.



.99

[poynt99]

Nice analysis MH.



.99