@Conradelektro,

Now your experiments are regaining interest and relevency. Let me just point out that multiple output coils split the output along with the acceleration effect! This is the bane of the Thane Heins generator. Nevertheless, it's a mistake to hide from the effect and pretend it's meaningless! I believe you'll return to the "Synchro Coil" research with renewed insights. It would be very simple to slip a stack of coupled radial magnets into the coil core, in place of that metal bolt, after you're through licking your wounds sufficiently.

You notice how the measured inductance drops down a lot at 10 KHz.  I am not sure why and I am a bit surprised, especially for the air core.  So you should avoid your testing at the 10 KHz self-resonant frequency and stick to lower frequencies.

The good news is that if you use just one of the wires in your coil (the thicker wire) you will have a comparable, but not identical coil.  The geometry will be the same, but the inductance will measure about 1/4 the series bifilar version.   So you can still observe how that monofilar coil responds to it's own resonant condition and it will be comparable with the bifilar with the understanding that the inductance and resonant frequency has changed.  You just have to dial up the correct angular frequency on your magnet spinner.

@MileHigh:

I just measured the inductance of one wire of my big Bedini coil,

it is 42 mH (air core) and 132 mH (steel bolt core) same for both wires.

Great, you predicted the 1/4 inductance.

Inductance of one wire also drops at higher frequencies:

air core: 43 mH at 1 KHz and 11 mH at 10 KHz

steel core: 80 mH at 1 KHz and 10 mH at 10 KHz.


This calculator at http://www.1728.org/resfreq.htm then computes a resonance frequency of 138 Hz with a 10 µF cap (steel bolt core). And I hope that I can also do 138 Hz (8280 rpm).

But I will not be able to do 248 Hz (14880 rpm) with one wire air core.


The possible tests I envision (10 µF in LC circuit):

70 Hz --- bifilar, steel bolt core

120 Hz --- bifilar, air core

138 Hz -- monofilar (half coil), steel bolt core

248 Hz -- monofilar (half coil), air core not possible


My new magnet spinner should do up to 150 Hz (about 10.000 rpm). The 12 V motor should do up to 12300 rpm at 18 Volt (even 14800 with little load), but I expect some vibrations, why I do not dare to hope for more than 10.000 rpm).

Self resonance is at 4600 Hz (bifilar air core) and 4000 Hz (bifilar steel bolt core). I did not yet measure self resonance for one wire (half coil).

I did impedance measurements at 10 KHz (one of the settings of my LCR-meter) several times and also the LC-circuit resonance tests with several capacitors showed that the impedance drops rapidly with frequencies higher than 1 KHz. But I will only need the very low frequencies (70 Hz to 132 Hz).

Is it correct that high impedance coils do strange things at high frequencies (specially beyond their self resonance frequency)? 10 KHz (where the LCR-meter shows low impedance) is twice the self resonance frequency, the coil should do funny things?

Greetings, Conrad

Hi Conrad,

When I have some more time this weekend I may add some more comments, now I notice only the eddy current losses of the steel bolt, influencing even the measurements you do when using the LCR meter but the thing is that several experimenters when doing the so called delayed Lenz effect tests often use bolts as a core, so you can test it too, of course.
Nevertheless, I advise to have at hand a normal ferrite rod piece too (OD=6mm length=40mm) from RS components (link: http://at.rs-online.com/web/p/ferrit-stabkern/4674065/ )  with one such core eddy losses will be negligibly small even at 10 kHz or higher.
Your link to the multilayer-multiturn coil calculator is okay and I would like to refer to another one I have recently found a bit more flexible and precise, works also on-line but freely downloadable too and have some additional plug-in softwares to make it more versatile (I have no any business interest with its creator  ),  see here: http://coil32.narod.ru/calc/multi_layer-en.html  and download it here, the plug-ins are also here: http://coil32.narod.ru/download-en.html

Regarding the drop in inductance at the 10 kHz LCR meter's measuring frequency for the air core series bifilar, I think the high value self capacitance of this multiturn bifilar coil causes it, see below the explanation. You may have also noticed that the air core bifilar has a bit higher inductance at 1 kHz (176 mH) than at 100 Hz (168 mH), this is because the 1 kHz is closer to the 4.6 kHz self resonant frequency, here I assume this latter frequency is the first parallel resonant frequency of the series bifilar coil caused by the about 170 mH inductance and the self capacitance.

The interesting thing is that such bifilar coils (connected as series bifilar like it is now) do have multiple resonances as the frequency increases, what is more: parallel and series resonances should alternate each other as you sweep the frequency due to the distributed nature of the self capacitance and inductance.
Here is a good site (watchable only via the internet wayback machine since last December)  on some measurements where the series bifilar coil is made of a coaxial cable and in Figure 10 the magnitude of the impedance is shown in the function of the frequency (the high impedance peaks are the parallel and the low impedance valleys are the series resonance points:  https://web.archive.org/web/20130321175122/http://vk1od.net/antenna/coaxtrap/index.htm 

When you have some more time next week, you may wish to check your multiturn bifilar coil in a measuring circuit and experience (by sweeping through the frequencies) the impedance peaks or the valleys manifested by the changing voltage values across the series bifilar, driving it from your function generator. (A good measuring circuit is to be figured out for that.)

Of course a coaxial cable is a transmission line where the outer conductor encircles the inner conductor and these two are inherently guided in parallel with each other of course. When you make a bifilar coil, the two pieces of wires can also be considered as a (long) transmission line where they are guided next to each other in parallel too (though quite often these are twisted to increase the capacitance between them).
Now you or MileHigh can understand why the inductance for the air core series bifilar is so small at 10 kHz: at this frequency the impedance is surely near to a series resonant point (near to an impedance valley) but it is still inductive. At a series resonant frequency (especially at the first series frequency just after the first parallel one) the impedance is nearly the DC resistance of the two series coils (plus some dielectric losses of the enamel and bobbin or core losses when you use any ferromagnetic core etc).

Greetings,  Gyula

So to better explain what I'm envisioning, with a regular coil and tuning capacitor when the applied voltage declines and therefore the magnetic field declines to maintain the current, that current keeps flowing through the wire and it's resistance to the ends of the coils wire before it gets to the capacitor. But with the bifilar coil there is no capacitor at the ends of the coil. So what happens to the current ? My guess is the coils distributed capacitance takes current from along the entire length of the coils wire via the dielectric medium being the insulation on the wire. Just a guess that if so, the ESR of this capacitor is very good.

Any thoughts.     

P.S. As an example just say I had a coil of both kinds which has about 720 mH and about 22 Ohms DC resistance with 220 nF excited at 400 Hz.

A generator capable of 400 Hz and about 220 volts would be nice to experiment with. I wonder what the frequency of a car alternator would be at 1200 RPM, I'll have to check.
I could bypass the rectifier and use one phase of that maybe for a lower voltage. 

..

Farmhand:

You are right about the current being lower for a self-resonating coil but your reasoning is incorrect.  The current is lower because the amount of energy stored in the self-resonating coil is lower, whether that be monofiler or series bifilar.

But firstly and formostly, you are forgetting the basics.  The current flow in the coil is synonymous with the energy stored in the coil, period.  You simply have to have current flow and deal with the associated i-squared-r losses no matter what.  There is no possible bypass or work-around for the current flow in a self-resonating coil, because the coil stores the energy in the current flow itself.

But here is a simple thought experiment:  If you have a regular capacitor in the LC circuit and start to decrease the value of the capacitor, the frequency goes up.  What is mentioned less often but is also true is that the AC voltage has to increase while the capacitor decreases in size.  If you read 10 volts AC when the cap is 100 uF, and you assume that the resonator energy remains constant, what is the AC voltage when the capacitance is only 37 pF?

The self-resonating coil with no capacitance would have to resonate at some crazy high voltage.  The voltage would be so high that the air would break down and the energy would be nearly instantly drained out of the coil.  And that's what happens all the time in real life.  If you connect a coil across a 1.5 volt battery, when you go to disconnect, the the coil discharges through a spark gap.  The spark gap may be so tiny that you don't hear it or see it but it is still there.

So, it's essentially impossible to have a self-resonating coil store a lot of energy because it "shorts itself out" the moment you go to disconnect from the battery.  Unless you are in a vacuum or do some experiments with coils submerged in insulating oil, a self-resonating coil in air will always be a low energy affair compared to the same coil connected to a capacitor.

MileHigh

In the theoretical case I refer to above the distributed capacitance of the bifilar coil is equal to the added capacitor capacitance and the distributed capacitance of the monofilar coil (in theory). So a bifilar coil with a distributed capacitance of 220 nF and a monofilar coil with whatever capacitance is needed added to it to get it to the same 220 nF. Probably need a lot of wire.

If someone wants to test it and that amount of distributed capacitance is not possible then the values could simply be changed to what is possible. Thing is both coils would have the same capacitance and inductance one has an external capacitor and one doesn't.

.

 On a side note it is surprising how many different configurations this transformer with the 16 windings on it can be configured. I can make four bifilar coils and connect them in different ways to get different measurements. Interesting...

..

For the sake of it, each wire loop in a bifi that is between the lead out first and last loops are little caps in series. lol it would be interesting to see if a bifi resonates at the regular coil freq because of it. Its got to be a bit different than where a normal coil, the plates of the cap are shunted somewhat by each turn, but the bifi are only shunted by half of the whole coil. Much easier for the bifi capacitance to have a high voltage. And maybe even if it can ring at normal coil freq also.


Weve got to try all kinds of ways to test. The amount of voltage difference working in these 2 different coil has to make a big difference in some way. I have some ideas that I want to try. Like dumping fields collapse current from another coil into the bifi capacitor. Where a normal coil would impede the spike, the bifi should take it in.  Might be a way of making good use of collapse spikes. Especially if those spikes are timed to res freq. I also really want to try the multi core inductor experiments. Ive got some time starting tomorrow to fiddle.

Mags

Gyula:

The interesting thing is that such bifilar coils (connected as series bifilar like it is now) do have multiple resonances as the frequency increases, what is more: parallel and series resonances should alternate each other as you sweep the frequency due to the distributed nature of the self capacitance and inductance.

Thank you for that information, that's probably the most important posting around here in a while.  I suppose that could be simulated with some kind of lumped transmission line or something like that.  Your link looks like the real thing.  Ham radio guys are the real thing.

To add to that, when you pump some sort of signal into that complex filter, it's the way the individual frequency components in the signal react with the complex filter.   There are alternating poles and zeroes (peaks and nulls in the frequency response.)  The resultant response is addition of all the responses to the separate frequencies.  It's like a frequency "shake and bake" process that takes place.  I still think of a real physical spring in self-resonance.  There is a certain randomness and strangeness to the physical spring in self resonance because if you think about it there are actually two or more springs springs in one.  There is the spring that stretches along the axis, and the spring associated with the bending of the shaft of the spring.  The shaft can bend along two orthoganal axes, so that's like two springs.  The important point is that an ideal spring is massless, and the mass of the metal of the spring is modeled as a distributed capacitance.  It's a very good analogy for what we are talking about.

Also, I saw something once when I was a kid that was so cool.  It was a big disk that stood upright.  There was a big tuning dial on it.  You put it next to your AM radio and you could pull in far away stations and make them sound clear.  You moved the big dial to find the tuning spot.  Even as a kid I thought it must be some kind of tunable resonator, and now as an old kid I agree.  I should try looking it up online.  A passive LC resonator that resonates on the AM band.

MileHigh