Eric Dollard

[orgonaut314]

Yes I understand. The resonant frequency and the spectrum are two different things.

The spectrum produced by the coil(s) is the spectrum measured and shown. They are different when the coils are tuned differently and have different resonant frequencies.

I measured my spectrum by sweeping the frequency and recording the amplitude. So my spectrum consists of resonance peaks yours not?


This is my spectrum:

very small peak at:
0.925MHz
something bigger peak at:
1.390MHz
I measured the very large peak at
2.766MHz

There is no resonance above this. I cannot plot the vallues because my nanoamp meter goes over the max.


What I would like to compare is this resonance spectrum for the secondary coil. Perhaps you are measuring how the sine signal gets distorted by a not perfect working frequency generator that does not make a perfect sine? I would probable find those distortions if I could measure more precise or used more power but the distortion is not my interest at this point for simplicity I assume the sine is perfect. I was studying the resonant behavior of the secondary coil to see if it behaved like the longitudional network. Those peaks really stand out and I see how the harmonics are at lower frequencies than the main peak.

It is a very odd spectrum as normally I would find harmonic peaks at higher frequencies than the fundamental peak. If you think about it there is a large resonance at a certain frequency than you double that frequency and there is no resonance at all. So these are not normal waves that make the signal peak. They seem like inverted waves or something? Eric made a lot of counterspace comments on them.

[orgonaut314]

Ok I think you are studying the shape of the output signal and its frequency content.

When I simulate the signal from the Tesla coil I get this strange shape. it is certainly not a simple sine wave. Very interesting. I plotted only 10ms. The input is a simple sinewave!



This is how it looks when we look at the last 10ms in 10s. Still not perfect sine there is some modulation and it does stay even after 100 seconds.


It turns out the internal resistance of the amplifier is important in dampening these transients. With 50Ohm it behaves like a nice sine wave.

[dR-Green]

I measured my spectrum by sweeping the frequency and recording the amplitude. So my spectrum consists of resonance peaks yours not?

I see! The spectrum is the spectrum or range of frequencies produced when the coil is in operation on a single frequency.

What you have are frequency response curves.

You press A4 on a piano and you hear a note, generally 440 cycles per second. But if you do a spectral analysis then you will find harmonics produced at 880, 1760 and so on, showing each harmonic associated with the note. That's the frequency spectrum of note A4.

In the same way, a coil produces a spectrum of frequencies, which don't necessarily have any relation to the self-resonant frequency of the coil.

Anyway looks like I didn't do it with the secondary alone unless it's in the coils compendium somewhere, if I did it then it should be there. I might have a play later and see.

Interesting findings anyway, I'm not sure how far the coil simulations will apply to reality but certainly introduces some things to keep in mind. It's already obvious that just changing the amplifier output impedance has an effect on the resonant frequency, in that is moves.

[orgonaut314]

Ok glad we figured it out. I was thinking that for the coil to carry an AM signal it should not distort a sinus carrier wave to much.

But about the internal resistance of the amplifier. I made a number of simulations and the resonant peaks change very much if the internal resistance is close to the impedance of the network. When it is much higher or lower the highest peak is best but when it comes close that peak gets much lower.

So that is something to take in consideration when a maximum peak is wanted. A transistor amplifier with almost zero internal resistance is the easiest solution. Using the frequency generator with 50 Ohm might not work so good.

[orgonaut314]

Another two pictures. This time a simulation of the Tesla coil as a network. Eric calls it the shunt concatenated network. You see the mutual capacitances on top and the mutual inductance as the first coils left. The rest is the transversal transmission line. Don't mind the resistors they are there for the simulator only.



The voltage seems stronger with 4 elements and the same 0,7V input. It has the same type of resonance as the longitudional network.



Looks like this one:


Has someone measured the complete frequency spectre with his Tesla coil? DrGreen?

It is funny that when you look at the voltages and currents along the nodes of this network for the first peak to the left of the maximum you see the nodes have a max at the start and at the end so two maxima. So this seems indeed a harmonic of the biggest resonance that had one maximum at the end node. Normally waves with higher frequencies have smaller wavelengths and are harmonics. This time 'waves' with lower frequencies are harmonics.

Also when looking at the differential equations that are 4th order they do not have a simple wave solution but clearly there is some wave like pattern.

Could it be that the wavelength goes with the frequency while normally the wavelength goes with 1/frequency?

Waves in counterspace

[ldrancer]

...

What?



10,000 people trying to be in the same place at the same time = people going nowhere. Have you encountered the concept of a traffic jam?

It's a known fact that humans are unable to control themselves. A population who have been told that an earthquake is coming will clog up the roads, and furthermore they'll trample over each other to be the first out of the door.

what? no. thats the answer. but how about 10,000 people knowing eric dollard can! predict earthquakes but its not imporatnt to yu to know about collapsing buildings but for an oil company to protect their oil. < and you being the one pushing this propoganda.


maybe 10,000 people can find yuo and you can tell me what a rat race is like.

[dR-Green]

what? no. thats the answer. but how about 10,000 people knowing eric dollard can! predict earthquakes but its not imporatnt to yu to know about collapsing buildings but for an oil company to protect their oil. < and you being the one pushing this propoganda.


maybe 10,000 people can find yuo and you can tell me what a rat race is like.

You can't pick up a building and move it to a safe place so your house has had it either way. If that's a problem then you obviously built your house in the wrong place.

The idea is limited to the laws that govern the real-world I'm afraid. What you can do is turn off the gas and electricity and water and prevent a big explosion and the costs of repairing it all afterwards. The oil company doesn't care if the pipe breaks and oil gushes out everywhere. The company doesn't particularly care about the destruction because you are the one who's paying to repair it in the end anyway, otherwise you are the one who has no electricity and no heat as a result of the damaged infrastructure. So forget the conspiracy theory nonsense.

You should study some experiments dealing with human psychology and crowd behaviour. Rather than following the exit sign out of a burning building, they follow the panicking crowd deeper into the inferno purely on the basis that's the direction everyone else is running. People aren't always as smart as you may think.

[dR-Green]

This time 'waves' with lower frequencies are harmonics.

There are such things as "undertones" although acoustic instruments don't naturally tend to produce them.

The table doesn't show what happens at higher frequencies. Is there no resonant peak above the fundamental?

[edit] But again, harmonics in this context usually refer to harmonic frequencies produced by something oscillating, as in a harmonic spectrum of an oscillating string at a fixed frequency. Resonant peaks are different. Is this a harmonic spectrum, or a graph of the resonant peaks after doing a frequency sweep? I notice Eric refers to them as a "resonant series" in the diagram and not "harmonics", suggesting it's a series of resonant peaks found when doing a frequency sweep.

[orgonaut314]

Hi drGreen, I will try to explain simpler for everyone to understand what I am thinking and if I make an error please say so.

To start lets look at a simple transmission line resonance. With two parallel wires we can think of the wire as a succession of capacitors and inductors. When the end of the two wires is left open (not connected) we have in fact a one wire transmission line or an antenna.

Normally in a circuit the wires are not seen as transmission lines because they carry dc or the frequencies are so low that there are no waves. But even than the wire is seen as a capacity and an inductance but the fact that these vallues are distributed does not matter than.

Now we are studying the antenna. That is a transmission line open ended. If the line is not wound in a coil we can use the following simple schematic for the line.



To see what is really happening this simulation has to be refined into infinite many nodes. That is difficult to do in software but on paper you can use capacitors with length dx and inductors with length dx. If you write down the relations for voltage and current in dx vallues you end up with a differential equation that gives a relation for voltage and current depending on time and space. A solution to that equation is a sine wave both for voltage and current. To find the real vallues the end conditions have to be matched and with voltage that means that at the end point the voltage can swing and the current must be zero.

In the simulation there are only 4 nodes and this seems to lead to maximum 4 harmonics witch means that there is a certain frequency that produces a standing wave with maximum voltage swing on the open end. Than there is a wave with another frequency that has 2 maxima and the next frequency has 3 maxima.

The voltage on the end note is plotted and we see that there is one ground frequency and three harmonics.




Now if we change the straight wire by winding it into a coil we change two thing. First the magnetic field in the windings now couples to the other windings and the electric field couples too. We now get this schematic



That has two currents going.
Horizontal


Vertical


This adds up to


All the vertical mutual inductance K and M are part of the longitudional current and all the self capacitance (to earth) and self inductance of the bare wire without the coil are part of the transversal current.

The resonance now looks different and it resembles a longitudional archetype network showed before.

Important to note is that the solution to the differential equation that results when we take capacitors and inductors of length dx is not a wave function. So the whole story about harmonics does not apply. Yet when we study this solution that the simulator shows us we see some wave like behavior. But different. There are no harmonics in the wave sense but Eric still calls these peaks to the left harmonics for some reason. We have to ask him.

[orgonaut314]

Sorry I had to edit the last post because I suddenly understood it better.

[dR-Green]

I'll do the frequency sweep tests and the spectrum again using only the secondary in the next couple of days. I have 20% 20 turn secondary and 15% 17 turn secondary designed for the same frequency. I expect to see a standard 1/4 wave distribution with maximum voltage at the free end and maximum current on the ground end. The RF current probe I used before has been reconfigured/dismantled so this time I'll use a loop of wire and scope for that purpose, and may end up with some phase measurements too.

[orgonaut314]

I used a 6080 tube in my coil in stead of the 6sn7. The internal resistance of that tube is much lower and with the pi network it is very small.

I still measure the magnetic field going up towards the top. The magnetic measuring loop horizontal.

[dR-Green]

I used a 6080 tube in my coil in stead of the 6sn7. The internal resistance of that coil is much lower and with the pi network it is very small.

I still measure the magnetic field going up towards the top. The magnetic measuring loop horizontal.

I don't know, it should be the other way. Where is the highest potential measured? Obviously it can't be on the ground end if it's grounded.

How are you measuring the current distribution?

Also when possible the setup should be in the middle of the room or as far away from other objects as possible because they will affect it. Your audio amplifier is probably getting a nice bit of energy too!

[orgonaut314]

Here I am measuring with a small coil on the lower end of the secondary. The middle of the small coil is on the lowest ring. The voltage reads 4,77V.



This one is on the middle of the secondary. The voltage reads 7,60V.



Here I am measuring on the top of the secondary. The middle of the small coil is on the highest ring. The voltage reads 7,29V.



I think the middle reads max because we can not measure a single winding and in the middle the average distance to all windings is the least. But you do see clearly that on the top the field is stronger than it is on the bottom.

I do expect this too because this is not a normal transversal quarter wave resonance. The mutual inductance changes the whole behaviour of the wave.

[dR-Green]

I think the middle reads max because we can not measure a single winding and in the middle the average distance to all windings is the least. But you do see clearly that on the top the field is stronger than it is on the bottom.

I do expect this too because this is not a normal transversal quarter wave resonance. The mutual inductance changes the whole behaviour of the wave.

I think the problem there is that the test equipment is much too close to the coil. As you approach the free end it becomes all the more sensitive, so when you are measuring the top it gives you a lower reading relative to the middle because it's detuned more so the output is actually less.

Keep the probe in the same place and put your hands near the coil and you will see what happens on the meter. The probe itself also has the same effect when you put it near the coil to do measurements. So the coil should be kept clear. Ideally you don't even want measurement probes in the field at all, but that's the only way to get measurements. So keep MAXIMUM distance between probes and coil at all times. Only have it so close as to make measurement possible. The actual values are meaningless so you don't have to try to get the highest numbers.

For measuring the current try a SINGLE loop of wire about 1/3 the diameter of the secondary coil, and for measuring voltage use a wire point or such small and pointy surfaces.

Also if the analogue meter and tube are still connected take them off. Is the secondary grounded? If it's just a ground plane (capacitance) and the capacitance on the top end is bigger then your 1/4 wave distribution will be upside down.