Testing the TK Tar Baby

[TinselKoala]

I'm surprised that no one caught this. I thought it was relatively obvious, so I didn't mention it. TK, the average current through the circuit is from the battery. When you install an inductor in parallel to the reversed LED, it lights up only because of the energy release of that inductor due to the flyback effect.

You could prove this by installing a forward-oriented 1N4007 in series with this arrangement, and the reversed LED should still illuminate.

Yep, like I said earlier:
"In the total circuit, though, the power flows from the battery and is dissipated in the load and the other circuit components; some of it sloshes around and depending on where your measurements are taken you can watch it doing all kinds of crazy stuff..."... like even lighting up a reversed LED if you put the inductor in the right place.

As long as one is allowed to pick and choose which measurements to take where, and to interpret them however one wishes, one may garner support for just about any contention. It is only when measurements of a particular quantity, taken and analyzed one way, are found to agree with other measurements of that quantity taken differently by known good methods and analyzed by known accepted methods .... that they may be relied upon to inform valid conclusions.

[TinselKoala]

Altoid: Reversed Current at Load?
http://www.youtube.com/watch?v=AGkZKXSZTzA

[MileHigh]

TK:

Just to be sure we are on the same page, that clip does not prove there is any reverse current.  It's the forward current charging up the inductor, and then when the forward current shuts off, the inductor discharges through the LED.  There may also be reverse current.  We don't know, but the test is inconclusive.

I am still really suspecting that most, if not all, of the negative potential across the CVR is a fake-out.  But I can't get my hands dirty sitting in my armchair.

MileHigh

[Magluvin]

MH is right. If you are measuring across where the led/inductor are plugged in, it is only because of the dump from the inductor that makes the led light. Seemingly anyways. ;]

As for the finding out if there is negative flow, then the test should be done with just the inductor, no led, to see if there is negative current flow.   Its not hard to understand. With the led forward biased, it lights, but negative bias and the circuit isnt working to possibly produce a negative swing.  And also with the led forward biased, how could there be a negative current? There cant be, its a diode. ;]

So showing it with the led forward biased, your A shot shows some( i say some you say strong) negative battery voltage, as you say and show in the vid. How does that happen if the led does not conduct current in reverse? Are the oscillations beating on the leds barrier to send that negative flow back to the battery? Is the led going through reverse breakdown?

Then with the inductor and reverse led, yes, the lighting of the led could be the inductor dumping into the led, but by having them connected this way, how can we say that there is not negative current flowing through the led from elsewhere? I see that the scope A is not showing any negative at the bat any longer, but that combo, the led/inductor, tries to keep that inductor in forward motion. That inductor is too busy dumping into the diode to be part of any negative flow.


Mags

[TinselKoala]

We have had lots of clues, from Lewin's lecture, through the TK JT measurement pitfalls demo, through to the present case, that should be telling us that isolated measurements within an oscillating circuit need to be taken with a grain of salt, at least. The big takehome message here is that casual electrical measurements can be misleading as to the true state of affairs. I maintain that the oscilloscope does not lie (at least not at these frequencies) and that what it is measuring is "in some sense" real circuit behaviour. The present measurements aren't telling us useful data wrt the real power, though. So what if a bit of current can be shown to reverse momentarily across a component.... this is really irrelevant to the main issue, which is this: Does all the power dissipated in the circuit come from the battery, or not? Does the battery discharge while running a load, or not? So I come back to the analogy of the toaster: it is a mistake to take the "mean current" in an AC situation as indicative of anything much, as both phases of the power curve are dissipated in the load and none is returned to the source. Similarly, if one provides a  minuscule DC offset, with a battery, to a purely sinusoidal and symmetric AC waveform.... that DC power is also going to be dissipated in the load, not sent back to the source battery, no matter the polarity of the added DC offset, I think.
But all that is a red herring.... because the net current in the circuit is DC, it has the oscillations superposed on top of it, and all the power comes from the battery and is dissipated in the circuit components, including the battery (heating) and the load.
The negative going excursion on the battery trace is another red herring. The battery voltage is certainly not doing this... but the circuit is, it's a real reading of a real phenomenon. It's just not "battery voltage reversal."