Is joule thief circuit gets overunity?

[TinselKoala]

Here is a scope shot of the input waveforms of my joule thief circuit with the super cap now at about 384mV.
The input current waveform still looks fairly similar to the way it looked at 408mV, but the frequency
slowly increases as the input voltage drops. The frequency is now around 294kHz at this input voltage.

Channel 2 is the current trace? Is your current trace display inverted, or are you using different probe positions than are given in the last complete diagram?

I see you have also overlapped the baselines of the two traces, putting them both on the center graticule marker. And you are using bandwidth limiting on the current trace.

(Edited, Channel 2 is the greenish one).

[TinselKoala]

Lawrence, it should be clear to you by now that the term "ringing" has two different meanings. First, the common meaning, an audible tone like a high-pitched bell ringing, that your students with young ears can hear. This is caused by loose windings on the toroid and perhaps the toroid itself, vibrating at the frequency of oscillation of the JT circuit, which as we know is a function of voltage as well as the toroid and transistor characteristics. When the oscillation frequency is in the range of human hearing you hear the parts vibrating.
The second, more technical meaning is the _inductive ringdown_ and this can be seen very clearly in your last, highest-speed scopeshot. This is the "bounce" in a signal that is seen when some energy is sloshing  back and forth between an inductor and a capacitance. As you can see, I hope, from careful examination of that last scope shot, this energy sloshes back and forth (the ringing) and dissipates (the decrease in amplitude of the cycles during each ring) but most importantly: IT AVERAGES TO ZERO, or slightly below zero since there is always a little bit of current flowing in addition to the ring. The frequency of this inductive ringdown is very high, as you can see, and is clearly out of the range of hearing. The inductances and capacitances participating in this high-frequency ringing are simply the wires and spacings in your circuit. This is why tight construction and short wires are used in high frequency circuit layout: to reduce this kind of ringing noise, which usually is considered detrimental to circuit performance.
Another feature to note from your series of shots is the difference between the first and second shots. You can clearly see that the first shot is missing a lot of the action, due to the "picket fence" effect of sampling and display rates and resolutions.  The first shot looks like random noise. The second shot is beginning to resolve the transistor's switching frequency (the potentially audible ringing). By the time you get to the last shot you are able to resolve the ringing of the inductances and capacitances of the wiring and layout themselves, far above the range of human hearing.... and acting as a purely dissipative mechanism for heating up the components concerned, as you can see from the fact that each cluster of ring oscillations decreases in amplitude over time.

(And you _still_ aren't putting your baselines directly on a graticule marker. See the little arrows on the left side of the display? Please use your vertical positioning knobs to put these little arrows _exactly_ on one of the horizontal graticule lines, unless there is a specific and clear reason for not doing so. You have no such reason, and the lack of accuracy in your scope setup is making it unduly difficult for observers to interpret your scope traces. I have mentioned this issue to you several times now. The reason for those wiggly lines on a scope display is so that they can be interpreted. The numbers in boxes that you seem to rely upon do not tell the whole, or even the most important story that an oscilloscope can tell. Proper management of the display of the traces is an essential part of scoposcopy. PLEASE PUT THE CHANNEL BASELINES EXACTLY ON HORIZONTAL GRATICULE LINES in all your future scopeshots unless you have a good, conscious and articulatable reason for not doing so. But congratulations on setting your trigger a bit better.)

[TinselKoala]

The triggering point on the Atten and the similar scope that Void is using is indicated by several features of the display. Of course the trigger voltage and channel and edge are displayed as a "number in box" at the lower right of the screen. But on the trace display itself, the trigger voltage is indicated by a "T" and arrow, in the color of the channel selected, on the left edge of the display along with the channel baseline markers. The trigger TIME is indicated by the white arrow at the top of the display, and this is normally set at the center vertical graticule marker. SO the intersection of the lines defined by the "T" and the white arrow is where the scopes are triggering (or should be).

In Void's scopeshot this can be clearly seen: the trace triggers the scope at the intersection, right in the center of the screen horizontally and at the set trigger voltage vertically.

In Lawrence's scopeshots... not so much.  In the last one particularly, the trigger point has been moved off the screen to the left with the horizontal positioning control and we are seeing a "delayed" display, and it's unclear just where the scope is triggering and on what signal, since the trigger point is off the screen. In many other scopeshots from Lawrence, there doesn't seem to be anything happening at the trigger point on the display, yet the scope is triggering anyway.

[ltseung888]

If the "ringing pulse" represents current crossing the 0 ref line, then we must have positive and negative power in the circuit.  (Input Volotage is all positive.  Output Voltage and Output Current are all positive.)

What does that mean?  How can current flow forwards and backwards in a DC environment?

*** Look at the theoretical explanation again.  If there is a "hidden" pulsing source, the above is possible.  That is Lead-out energy.  So long as there is crossing of the 0 reference line - either the main Current or the spikes, there is lead-out energy.  Amen.

*** It is a matter of producing the circuit to achieve this crossing the 0 ref line behavior.  The Zhou Board is only one example.

[Void]

Channel 1 is the current trace? Is your current trace display inverted, or are you using different probe positions than are given in the last complete diagram?

I see you have also overlapped the baselines of the two traces, putting them both on the center graticule marker. And you are using bandwidth limiting on the current trace.

Hi Tinselkoala. My probes are oriented the same as I have shown in the schematic I posted previously. Yellow(Ch.1) is the input voltage waveform, and blue(Ch.2) is the input current waveform. The input current waveform should not be inverted. Yes, I have both waveforms vertically set at the center line. I like to set them that way.

Regarding the bandwidth limiting on channel 2, my scope turns this on automatically at low volts per division settings, and I am not able to turn it off. There seems to be some auto routine there that flips this on maybe as the signal gets a certain percentage of noise on it. My input current waveform is starting to get close to the noise level at this input voltage. I don't know what frequency the scope limits the bandwidth to when the BW Limit feature is turned on. The user manual for my scope seems to have been written by a grade four student. The PC interface software that comes with my scope also seems to have been written by the same grade four student. The scope itself however seems at least passable, so far.