Claimed OU circuit of Rosemary Ainslie

[MileHigh]

Peter,

The Naysayers are now shown to be the real amateurs who have no insight into this circuit. Sorry Poynt99, but SPICE just isn't going to show this level of complexity. Aaron has now published a circuit that has all of the flexibility necessary to produce all of the necessary effects. And, by placing the scope probe directly on the battery, you can see when the oscillations are happening, both when the FET is ON or OFF.

Don't be too hasty.  You have to take a step back from the "jumping the gun" strategy.

For example, if Aaron has the 4-channel version of the DSO you are in heaven.  You can scope the following signals relative to the physical battery ground terminal:  The 555 signal at the 555 source pin (not at the gate input because you don't want to disturb it for the first go round), the shunt resistor, the MOSFET drain pin, and the battery positive terminal.  That will give you a complete picture of what the ringings look like.

The DSO has a resolution of 2 nSec so you should be able to see propagation delays in the signals at the physical nodes so that would be interesting.  There may be a few nanoseconds delay between the MOSFET drain spiking high (i.e.; the coil-resistor is the same node) and the battery positive terminal spiking high, which is showing you the reflection (If that is the case TBD).  You can put a shunt resistor on the battery positive terminal to see if there is any charging current when the spike hits.  I know Aaron mentioned the voltage potential only mechanism for charging the battery.  That doesn't mean you don't want to look anyways.

If you can see how the transients work and generate the oscillation on demand (assuming that there is no bug in the 555 circuit TBD) then you should be able to follow the energy trail.  It would be nice to have a full sense of the whole orchestration of the signals when the circuit is oscillating right down to the transient analysis level.  This is not trivial, and I doubt that Aaron has ever done it.  It is basically about generating a timing diagram for what is happening down to the nanosecond resolution level.  You are reverse-engineering what the circuit is doing in time.

If you can do that, you should be able to do the full audit of the energy trail.

If .99 was interested, he could try to insert some inductive and capacitive values for the wire interconnects to generate what looks like the same ringing that you see in the real life circuit.

If the DSO can't cut it looking for propagation delays, then I assume Aaron has a very fast analog scope that could do the job.

Spice, the navigation through the heavens of the circuit, totally blue eyes.

If you do this properly, you should be able to crunch some numbers to see the energy going out of the battery being sliced up like a pie.  One slice is the "recharge" comeback slice.  Another slice of the pie is the resistive element of the coil-resistor, another slice the MOSFET, and another slice is called "other."

The comeback slice of the pie can be further subdivided into two slices, the true recharge slice and the energy lost in recharging slice.

Now that's one pie.  Then there is the hope that there are 16 extra delicious heat pies that fall out of the sky and sit next to the main pie for every cycle.

MileHigh

[TinselKoala]

OK, I think I've got the circuit behaving as Aaron wants. It is indeed pretty tricky to tune it to get this behaviour, but Hoppy's clues, the kluged 555 circuit, and Aaron's final stable scope shots have given me enough information, finally.
I'll post a photo in a few minutes, in this post, as an edit. (I have to upload it from another computer.)
The only problem is that, as Hoppy noted, this mode turns the mosfet on most of the time. SO things heat up quick and the mosfet will likely fail if it isn't on a heatsink.

EDIT: here: and look I even used the Flukeoscope. "A" trace is the timer pin 3, "B" trace is across the battery; note the 200 volt spikes; load is 12.5 ohms, 200 uH or so ("catfood2"). Mucho heat and something over an amp average input current.

[MileHigh]

Great work TK.  In your setup the 555's not functioning properly.  You can see how you captured two cycles of a repeating irregular pattern.  The 555 is simply never supposed to do this.  Did you try the magic laying on of hands?

MH

Touché Wilby, that was a high quality posting!

[WilbyInebriated]

Spice, the navigation through the heavens of the circuit, totally blue eyes.

the spice melange, an awareness drug from frank herbert's 'dune'. the ironic thing is you using it in such context, as the book was largely about what could happen through dependence on such a substance.
but then i'm sure you were busy saying things like "nice stillsuit!" or "cool blue eyes!" instead of being able to comprehend the many interweaving plot layers...
may your knife chip and shatter.

"most discipline is hidden discipline, designed not to liberate but to limit. do not ask why? be cautious with how? why? leads inexorably to paradox. how? traps you in a universe of cause and effect. both deny the infinite."  - the apocrypha of arrakis

[TinselKoala]

@MH: you're telling me! It is totally unstably stable. But regularly unstable. You can see the same fine features repeating from cycle to cycle.
And sure, fingers make the frequency vary, and can kick it over the threshold into nice clean pulses. The pot settings to get it here are, as hoppy said, incredibly sensitive (and in my case I do think they are related to that bad spot). I see that Aaron has gone to 3 ten-turn precision pots.
But I think we really need, now, to see Aaron's 4 traces like you suggested, or even just any two including the current trace. At least in my case, the average input current is high.

But still, my traces aren't exactly like Aaron's, and that bothers me. I need to do some more fiddling.