Testing the TK Tar Baby

[TinselKoala]

TK,

The dates indicate that this could be quite a different test. I would not rely on this being representative of the oscillations shown in SCRN235.

Look, I'm NOT saying I am right about the 50W, I am saying this is what the scope computes as a MEAN power over that entire display. It may be wrong, but is there any reason that displayed 179mA could be wrong? Possibly yes. Is 50W average an unreasonable power into the circuit? I'd say no, based on my findings.

I would agree that going strictly by the displayed traces, the average power appears to be closer to 183W or so.

I have never argued that what the scope computes is not what the scope computes.  I am arguing that what the scope SHOWS is in fact correct, under the conditions of the experiment. (73 volts and 14-15 Ohms TCR) for nearly half the time during the SCRN0235 trial. And that what the scope is computing for the "mean" values many times is incorrect, possibly due to sampling errors, since the scope is not being used correctly. Taking the "mean" of a burst envelope containing an undersampled high-frequency oscillation within part of it is not correct and I think you know this. What are the scope's computed mean voltage drop values for screenshots taken _during_ oscillations, not the burst enevelope? They are typically under 100 mV negative, usually more like 40-50 mV.. So well less than half an amp.

Every expanded scopeshot of the oscillations themselves, when grouped according to the battery voltage, shows virtually identical waveforms. But your argument requires that the SCRN0235 oscillation waveform, which we cannot see, to be radically different from all those we have seen.

For the average power in SCRN0235 to come out to 50 W, I think there would have to be over 2 amps average current in the negative direction during the oscillation portion. No expanded trace of the oscillations has ever shown anything like that. Peak currents yes, but immediately followed by a peak in the opposite direction, for essentially no _net_ current flow, much less 2 amps or more negative. And the scope's own data boxes bear this out when the oscillations themselves are examined at a proper sample rate.

In some _other_ trials, where 73 volts and a long ON time are not used, I have no problem with a 50 Watt average power level. In this trial, though..... unless I can see a full derivation for that value that does not depend on an unknown calculation from an undersampled waveform-- I'm gonna have to go with the higher average power figure, for that shot.

Look.... I'm arguing for an extremely high power level in a Free Energy experiment, higher even than the claimant has claimed, and my argument is supported by the claimant's own_actual_ data, and even using what I believe to be the incorrect mean value, the average power is STILL very high.

I am interpreting this high power level as mostly coming from Q1 being turned fully on and carrying 5 amperes of current during the time when there ARE NO OSCILLATIONS which are supposed to be producing the "unusual effects" like anomalous load heating.   And I assert that all the heating in the load on this trial is accounted for by the high power levels that result from the Gate HI time when Q1 is fully on and carrying 5 amperes at 73 volts, and that the oscillations play very little role in adding to the heat.... and they certainly cannot subtract from the heat. Even negative power has to wind up somewhere! Does your toaster's element cool off during the "negative" phase of the AC power from the wall? I doubt it.

I also assert that the Q1 mosfet will not survive long under this treatment. And further I assert that the immediately subsequent scope shots, and the forum posts describing them, illustrate a mosfet driven to failure WITHOUT BEING NOTICED and all the Gate HI _with flatline CVR_ traces after that were performed with a bad mosfet in the Q1 hole. And I further assert that the "high heat" in the load mode CANNOT be done with a 72 volt battery pack and substantial Q1 mosfet ON times and that tiny heatsink because the transistor will fail. These assertions are supported by the actual data in the scopeshots and the statements made in the forum posts.

And later today, they MIGHT be supported by an experiment. I intend to put 5 amperes through an IRFPG50 using a 73 volt source and 15 Ohms of total circuit resistance, and I will use a 45 percent ON duty cycle at 32 kHz. I will display the waveforms and a timer. My heatsink is a little larger than Ainslie's, it is a real TO-247 heatsink using thermal paste and it has fins, we can be sure that it has better heat transfer ability than the bit of U-channel that she was using at the time of this shot.

Of course.... if the inverted schematic, placing the Gang of 4 in the Q1 position, was used.... then of course they could carry the 5 amps without difficulty. And with that I have no problem.  Except that the subsequent shots do appear to show a failure, of course.

[TinselKoala]

@TK


I don't know if it will help to illustrate the point, but is there any mileage in explicitly duplicating the standard "DC" portion of the trace in question using a  power supply and a function generator to switch it with the same frequency and duty cycle as the smoking mosfet screenshot?

If I have it correctly, it would make it easier for whatever power averaging features your scope has to do the job correctly, and maybe give people greater confidence in the conclusion for the DC component at least.

With the DC portion replicated in this way you could regard that overall shot as the sum of two traces - standard pulsed DC who's power contribution you can measure without controversy, overlaid with bizarro mosfet spasms.

That is exactly what I do intend to do, later this afternoon if possible. But my scope doesn't have any power averaging features, it has to be done manually and this process takes a little while. However there is a much more easily demonstrated consequence of putting that much power through a poorly mounted mosfet, and I think my scope will illustrate that consequence quite well.

We shall see. Stay tuned to this channel for more exciting results from Moses and Festus, Boris and Natasha.

[TinselKoala]

@MH: Of course, if I punch a big hole in the side of my bucket (a cap to ground), pretty soon there won't be much left to slosh around in there higher than the hole, and if the hole is big enough, even the heavy rainfall won't keep my bucket full enough to slosh. And if I punch a hole in the Bottom (a resistor to ground) it gets even worse for my sloshes.


There was another test you suggested that appeared to involve three scope measurements at the same time. Unfortunately my scope only has the two channels. The next time I take Tar Baby out to play, I'll see what I can arrange, but I don't know when that will be at this point.

[TinselKoala]

Now...... Referring again to SCRN0235:

The "average" or "mean" power must be the "average" of the Gate HI time power and the Gate LO time power, right? It's not correct to average the "voltage" across the whole cycle consisting of oscillations on one part and DC on the other part. You have to compute the powers separately and then use them both in the computation of the average across the whole envelope. The oscs on the CVR are an AC signal. How do we find the "average" power in an AC signal? Usually this is taken to be the RMS power, is it not?

[Rosemary Ainslie]

Guys - yet another appeal to Stefan Hartman.  Let's see what comes of this one.   

Regards
Rosemary

My dear Stefan,

There is still an uninterrupted series of utterly incorrect allegations related to the objects of our tests that are fully described in our paper.  Together with this is the continual degradation and misrepresentation of our work.  If you do NOT act against this within the next 24 hours, then I must assume that you endorse this or that it conforms to your preferred editorial bias.

Let me know.

Regards,
Rosemary