Sonic Resonator Results and Findings, As Well As LTspice Models To Download

[fritz]

... The graphical approach:

Print your input power trace and your output power trace using the same scale.
Take a scissor and cut out your input power "block" with the length T.
Than cut out your output power block - which has the length of 0.64 T.
You will find out that if you cut the ouput power block to the heigth of the input power block - the remaining piece will fit to your input power block.
-> Both blocks have different shape but same area.
Well thats graphical integration.
You can even print out on mm paper and count the dots under both curves.

This method works for every signal. In the case you have real AC you can sum up
positive and negative dots.
So we don´t need your formula here which is kind of dumb-rule and works only within quite narrow circumstances.

So if you don´t know if you´re right - take a scissor - instead of describing what you think might be - over and over.

[D.R.Jackson]

... The graphical approach:

Print your input power trace and your output power trace using the same scale.
Take a scissor and cut out your input power "block" with the length T.
Than cut out your output power block - which has the length of 0.64 T.
You will find out that if you cut the ouput power block to the heigth of the input power block - the remaining piece will fit to your input power block.
-> Both blocks have different shape but same area.
Well thats graphical integration.
You can even print out on mm paper and count the dots under both curves.

This method works for every signal. In the case you have real AC you can sum up
positive and negative dots.
So we don´t need your formula here which is kind of dumb-rule and works only within quite narrow circumstances.

So if you don´t know if you´re right - take a scissor - instead of describing what you think might be - over and over.

Ok that sound interesting,  also I am downloading all the replies so I can read and make sure I know what things to reply too.  So,I will try to get to every post made here.

But your suggestion is an interesting idea I will work on today.

Quite frankly the replies here are pretty interesting to examine from every angle, and so, I need to answer as many of them as I can as we go along.  And if they make me discover something then thats good.

[D.R.Jackson]

Oh no, that awful thing called coil series resistance and series loss!

Here are the ways to reduce real world component losses, in such things as coils.

Well, I am sure that some of you can think in dimensional terms when it comes to fundamental driving frequencies and reactances.  If you want to work around coil series resistance losses.

Which means that as we increase our sonic resonator driving frequency from 1 kHz up to 10 kHz, the coil windings decrease 10 times too.  And so, the wire length decreases and so does the series resistance loss.  Thus this is one way we can over come the series resistance of the coil by moving the fundamental sonic resonator frequency to 10 kHz. Or as high as 100 kHz.

At 100 kHz the coil dimensions have decreased considerably to obtain the same reactance at 100 kHz and the series resistance has dropped too.  The ratio of series resistance to inductance decreases in inverse proportion as we go higher in frequency.

Since a good transformer core can make up for the number of coil turns, and so reduce the turns by the permeability factor of the core, we can make a transformer of a good core material to reduce losses.  And at 10 kHz and 100 kHz we can use a low loss toroid core.

So in transforming our fundamental frequency to 10 kHz we can choose a coil or transformer for the same reactance but yet, the series resistance has dropped 10 times.  At 20 kHz it will drop 20 times, and at 50 kHz it will drop 50 times.  And so at 100 kHz it will drop 100 times as compared to the resistance of the coils and transformer at 1 kHz.

So these are the work around's to the problem of internal series resistance in coils and transformers.  And thus is the means to over come the series resistance losses of the windings.

Also, we now have 60 Hz power supply transformers made of highly efficient toroid cores.  And we have such cores for audio also.  And so, from 10 to 100 kHz we can have us a fairly efficient coil or transformer with reduced series resistance losses via more efficient core materials that cut down on the number of coil winding turns which in turn reduces the length of wire needed for the windings.  And we can also opt to choose a wire diameter that will provide us with a low resistance that when modeled in the sonic resonator will not result in allot of loss.

Now as for capacitors, in my ARRL handbook it is stated that a capacitor is considered to be a lossless device in that in AC circuits the energy stored in the capacitor is all returned back into the system.  And as far as dielectric leakage current goes, the amount of leakage current is in the uA range and is so low that its effect is grossly masked by the much larger operating currents of the circuit and so, barely even effects the circuit in a noticeable way.

So, given the above considerations, and kinds of thinking we can work around such things as coil series resistance losses merely by upscaling the fundamental frequency of operation of the circuit.  If your an engineer or a technician you should have know this is how it works.

Anyways, given the right wire diameter for our coil, and a good transformer core, along with the right frequency we can get to a design where the transformer series resistance in around 1 ohm or less.

So, this is the kind of thinking that you can view things by if you work with radio frequencies over a spectrum of operation. 

We do not have to be limited to the losses that we might realize in a real circuit at 1 kHz, simply by upscaling the frequency higher.

And this makes the awful topic of coil series resistance somewhat tamable don't you think?

Things are not as bad as some might wants us to think.  Since there are more options than one.

Now to answer a few more things questioned in these post here.

The circuit if left unfiltered will allow AC current to flow into the power supply and LTspice will show you that the power supply power plot will have AC power wave forms in the power supply.  So the filter after the power supply filters out these power wave forms so that the power supply then will have a pure DC current flowing.  And there are no instantaneous pulses of wave form peaks that appear in the analysis given the right amount of filtering.

Next, as to the period of the out put wave forms.  The output is constant at the fundamental driving frequency in the circuit of 1 kHz.  And so, the period of the wave form is constant for every wave cycle produced at output.

So that is the conditions of the power supply.  If you do not filter the energy of L1 from entering into the power supply then you will have AC dissipation occurring in the power supply and that is not at all efficient.  All periodic information must be filtered out of the power supply so than no sine wave or pulsed information enters into the power supply.

Also, having explained the actions of L2 and Dx in a previous posted reply here.  We want to look at the effects of resonance and how it equates to enhancing the circuit concept once some losses have been returned back to the circuit from L2 and Dx where the returned losses via Dx are  = to, the Voltage at the far end of L2 times the current of Dx.  Which mean that we have some power added back to the circuit that was recaptured from L1.

The electromechanical resonance of the circuit is a momentum factor.  Since, in real circuits when you turn off the power.  The resonant circuits will power down in a slower way than the rest of the circuit, in Milli second terms, since the resonant circuit tends to try to oscillate on its own for a small period after the power is turned off.  What this equates to is a storage of energy like that of a fly wheel.  However as compared to a fly wheel the resonant circuit will loose its momentum fairly fast in Milli second terms.

In this way it can be seen that the resonant circuit is an inertial momentum component that tends to want to remain in motion following the laws of inertia for things in motion.  Now while the power is on, the inertia stays in motion and adds that energy to the circuit.  And will tend to obey the laws of inertia.  And so, since the coils of the resonant circuit are adding into the circuit an induced higher voltage, we can see some reasons for having an unusually high square wave output for an extended period of the wave cycle.  The period of the half wave cycle being extended to 0.64 wave cycles seems to be an indication of the momentum of the output also wanting to stay in motion and resist wanting to go to the negative half cycle by as much as +50 degrees past 180 degrees.  Now this does not mean that the momentum of the resonant circuit is extending the power wave cycles half cycle in terms of momentum.  But the timings of events that occur in the circuit effect the output period more than anything.

[TinselKoala]

I see you still don't address the points I and others have made. I will point out here that I have not called anybody names and I am addressing only the issues you have brought up. Nevertheless, stalkers want to flame ME, but they don't want to address the points I make in my posts. Am I wrong? If so, how? Other aliases that I may or may not have used have nothing to do with the points I make in this thread. Do they? So how am I wrong?

1) You are comparing instantaneous power when you should be comparing energy.
2) You do not seem to understand the mathematical concept of integration, or that energy is the time integral of an instantaneous power waveform.
3) You are ignoring the built-in capability of LTSpice to integrate the instantaneous power waveforms for you. Instead you complicate the issue by looking for "rms value of square wave" when your output waveform deviates significantly from squareness. As I showed with my own scope trace and your own numbers.
4) Other people are telling you the same thing.
5) You do not seem to realize that the area under the instantaneous power curve IS the energy, and integration gives you this area exactly.
6) I and others have challenged you to physically cut out your waveforms and weigh them or otherwise quantify the area, if you don't like LTSpice's integration. Why don't you do it?
7) Energy is Not Power. Your system does not output more ENERGY than it takes as input. It is NOT overunity, and it is evident from your long and complex posts that you don't understand the concepts of Power, Energy, Work, and so forth, as they are standardly defined, and as I asked you about in a much earlier post, to which you also never replied.
8 ) Your calculation of overunity power relies on your assumption of a square wave output, but your output waveform is not square at all. Compute, please, what a 62 percent duty cycle, instead of your assumed 64 percent, would do to your overunity claim.
9) You are the one who quoted long strings of impossibly precise numbers. That looks pretty silly to me, but if you want to call my insistence on correct numbers of significant digits an obsession, fine, I'm obsessed.  But remember: When you say something is "8.0943785" or whatever, you are saying it's NOT 8.0943786 and it's NOT 8.0943784, and so forth...so of course I will challenge you on it, and so should anybody with an understanding of real precision and a hankering for truth.
10) You've started multiple threads on the same topic and made incredibly long posts that say basically the same things, and make the same errors. I think one thread should be enough for anybody.

[TinselKoala]

... The graphical approach:

Print your input power trace and your output power trace using the same scale.
Take a scissor and cut out your input power "block" with the length T.
Than cut out your output power block - which has the length of 0.64 T.
You will find out that if you cut the ouput power block to the heigth of the input power block - the remaining piece will fit to your input power block.
-> Both blocks have different shape but same area.
Well thats graphical integration.
You can even print out on mm paper and count the dots under both curves.

This method works for every signal. In the case you have real AC you can sum up
positive and negative dots.
So we don´t need your formula here which is kind of dumb-rule and works only within quite narrow circumstances.

So if you don´t know if you´re right - take a scissor - instead of describing what you think might be - over and over.

It is necessary to use a sufficient horizontal scale, so that the deviations from squareness (the slow rise and fall times) can be seen, before the cut-outs are made. 300 microseconds per horizontal division is extremely coarse, but even on that scale one can see that the rise times of the "square" wave are not vertical--hence the wave is not square. But to be fair the cutouts should be made with a horizontal scale of 50 microseconds per division--then you can see what you've really got.
I suggested weighing the cutouts but I realize not everyone has a sufficiently accurate scale. Printing out on small graph paper and counting the squares is also possible.
I should point out that the rise times seem to be on the order of 20-50 microseconds, as I computed earlier when I showed a real square wave from my oscilloscope. That makes those waves not really square, and actually is a significant deviation that will affect the results. A true square wave should have a rise time several orders of magnitude faster than that. The square wave from my oscilloscope's calibrator has a rise time of under 50 nanoseconds, and that's not even that good.