Back .

Re: Designing High-Gain Triple Resonance Tesla-Transformers

Edited/Updated:  October 27, 2004

First Post

Subject: Designing High-Gain Triple Resonance Tesla-Transformers.  Date: Mon, 09 Aug 2004 17:17:28 -0600

Original poster: Terry Fritz

Hi All:  Jay Reed has sent me the following paper:

http://hot-streamer-dot-com/temp/DesigningtripleresonanceTeslatransformers.pdf

Designing High-Gain Triple Resonance Tesla-Transformers

He does not have Internet access, but I can give you a mailing address if you want to contact him.  Cheers, Terry.

 

Date : Mon, 09 Aug 2004 20:57:06 -0600.  Subject : Re: Designing High-Gain Triple Resonance Tesla-Transformers

Original poster: "Antonio Carlos M. de Queiroz"

Interesting. Is this a published paper?  I seems possible to adapt my design procedure to account for a capacitance across the third coil, but I wonder how to determine the value of this capacitance, given the dimensions of the coil.  I usually consider that the capacitances of a coil can be modeled as two capacitances to ground, similar to Medhurst's capacitances, one at each end of the coil. The "other plate" of a Medhurst capacitance is the ground, and not the other end of the coil.  Of course some of the electric field ends at the other end of the coil, but how to determine how to split the Medhurst capacitance in two parts, one to ground and another to the other end of the coil?

A possibility would be to model a coil with distributed capacitance more precisely, by splitting it into a large LC network, as Paul Nicholson did, and then looking for the first notch in the frequency response obtained by exciting the coil at one end and looking at the voltage at the other. A capacitance across the coil can then model this notch. The capacitance from the output to ground can be obtained from the first resonance observed, that would be the resonance frequency of the coil with the two capacitances added.  An identical capacitance, or a somewhat different one, if the actual geometry of the system is considered, can be obtained by exciting the network from the output side and observing the input.  There are other details and complications if this extra capacitor is added, but nothing that can't be treated.  Antonio Carlos M. de Queiroz.

  

Date : Mon, 09 Aug 2004 20:57:12 -0600.  Subject : Re: Designing High-Gain Triple Resonance Tesla-Transformers

Original poster: "Antonio Carlos M. de Queiroz"

Actually, I have how to design the triple resonance networks with that structure directly. See my last paper about multiple resonance networks, "Band-pass multiple resonance networks", presented in a conference two weeks ago: http://www.coe.ufrj.br/~acmq/papers/papers.html .  Look at the form with balanced output (discussed in this list some time ago too). Remove the "normal" high-pass output branch and you have the desired structure.  So, the procedure would be to design a band-pass m.r.n. and then to rearrange the last two capacitors and inductor. Insert a transformer at the input and any voltage gain with perfect energy transfer can be obtained.
I can make a complete example, if there is interest.  Antonio Carlos M. de Queiroz.

 

Date : Mon, 09 Aug 2004 21:06:55 -0600.  Subject : Re: Designing High-Gain Triple Resonance Tesla-Transformers

Original poster: Terry Fritz

Hi Antonio,

>Interesting. Is this a published paper?

He tried to get it published in the Journal of Scientific Instruments but they thought it was too long and such...

>I seems possible to adapt my design procedure to account for a capacitance across the third coil, >but I wonder how to determine the value of this capacitance, given the dimensions of the coil.
>I usually consider that the capacitances of a coil can be modeled as two capacitances to ground, >similar to Medhurst's capacitances, one at each end of the coil. The "other plate" of a Medhurst
>capacitance is the ground, and not the other end of the coil.  Of course some of the electric field >ends at the other end of the coil, but how to determine how to split the Medhurst capacitance
>in two parts, one to ground and another to the other end of the coil?

It also gets messy since there is a wire leading to the coil which may have a significant effect and whose position has wide variation.

>A possibility would be to model a coil with distributed capacitance
>more precisely, by splitting it into a large LC network, as Paul
>Nicholson did, and then looking for the first notch in the frequency
>response obtained by exciting the coil at one end and looking at
>the voltage at the other. A capacitance across the coil can then
>model this notch. The capacitance from the output to ground can
>be obtained from the first resonance observed, that would be the
>resonance frequency of the coil with the two capacitances added.
>An identical capacitance, or a somewhat different one, if the
>actual geometry of the system is considered, can be obtained by
>exciting the network from the output side and observing the input.
>There are other details and complications if this extra capacitor
>is added, but nothing that can't be treated.  Antonio Carlos M. de Queiroz.

Interesting.  I will print your latest paper and comments and send them to Jay.  Cheers, Terry.

 

Date : Tue, 10 Aug 2004 07:30:39 -0600.  Subject : Re: Designing High-Gain Triple Resonance Tesla-Transformers

 Original poster: "Harvey Norris" <harvich@yahoo.com>

--- Tesla list <tesla@pupman.com> wrote:

To each his own, about two (or more) years ago, I ran a tank circuit from source alternator frequency of 480 multi-turned into a pair of 11mh 14 gauge coils, each setting on high induction coils of 23 gauge that in turn by air core induction sent currents into a group of 20 of the same 14 gauge coils. The net result of this was that the ending group of 20 coils received more amperage then if the same group of coils were directly connected to the same source of emf!  They gained more amperage through a triple resonance then if they were directly connected to the source of emf, meaning that the air core action gave better results then the direct connection could give. Go figure? This was an example of a triple resonance device because:

1) The multi-turned secondary acted exactly as a tesla secondary acted, giving superior performance where the amp turns of the secondary exceeded the amp turns of the primary.

2) The primaries configured as tank to the source freq resonance exhibited resonant rise of amperage with respect to its source of amperage.

3) The second part (multi-turned coils) by air core induction exhibited resonant rise of voltage with respect to the primary voltages imposed on its action.

4) The third component (from secondaries resonant voltage rise) was another tank circuit that acted from the second components voltage rise.

5) Essentially a source frequency magnifier action, all done with NO Tesla high frequency arc gap involved.

6) Source frequency converted to bipolar tanks/ secondaries making inductive air core voltage rise/
line coupled tertiaries reconconverted
[sic] again for tank amperage rise on ending elements.

7) Parallel resonance as primaries/ Series resonance thru air core secondary matching/Parallel resonances as ending line coupled ending components.

8) Ending components exhibit more amperage through triple resonance, then if they were originally connected to the same wires connected to primaries.

9) Provable and demonstratable, if those words can be found in a dictionary: definition; source frequency resonance; a subject beyond most tesla coiler's expertise.  Sincerely HDN.

 

Date : Tue, 10 Aug 2004 07:31:02 -0600.  Subject : RE: Designing High-Gain Triple Resonance Tesla-Transformers

 Original poster: "Steve Conner"

>Jay Reed has sent me the following paper:

Very interesting.. But it seems from the voltage plots (figs. 3 and 4) that his "optimized" design does not have complete energy transfer, since V2 doesn't drop to zero when V1 peaks. Compare to the plot for Bienosiek's design, fig. 3, where v1 _does_ drop to zero.

Maybe Reed's design is optimised for voltage gain only, and not energy transfer. But I would have thought that improving the energy transfer would increase the voltage gain too, and hence a coil with optimal voltage gain should automatically have 100% transfer.  Steve C.

 

Date : Tue, 10 Aug 2004 15:29:08 -0600.  Subject : RE: Designing High-Gain Triple Resonance Tesla-Transformers

 Original poster: "Steve Conner"

>source frequency resonance; a subject beyond most tesla coiler's expertise.

Well not really... Anyone who has ever played around with VTTCs or SSTCs has used source frequency resonance (or forced response as Antonio would call it) since the vacuum tube/MOSFET/IGBT/whatever is just a source of high frequency AC.

The only significant difference is that it can produce far higher frequencies than a rotating alternator could, so we can use conveniently-sized air cored coils. I've seen pictures of Harvey Norris's alternator-driven resonance coils and they're _BIG_ :o

I believe Tesla himself used special HF alternators to drive his coils at one stage, and similar machines were used for medical electronics and radio transmission, until they were superceded by tube and transistor generators.

>The net result of this was that the ending group of 20 coils recieved more amperage
>then if the same group of coils were directly connected to the same source of emf!

This is a great demonstration of the powers of resonance, but not really unusual. You can easily induce enormous currents in a parallel resonant circuit by applying just a small current to it. Nobody's breaking any laws of physics, as the current is reactive- no actual work is done by it. The power associated with that current is just sloshing between stored electric and magnetic energy.

For instance, in an RF amplifier as used by radio hams etc. the resonant output circuit can see currents of a few amps, although the plate current of the tube driving it may only be ~0.2A. This may seem wasteful, but in fact a strong resonance is desirable in the output circuit- it makes sure only the wanted signal is amplified, and other unwanted frequencies (harmonics etc) are filtered out.  Steve C.

 

Date : Tue, 10 Aug 2004 18:29:10 -0600.  Subject : Re: Designing High-Gain Triple Resonance Tesla-Transformers

Original poster: "Antonio Carlos M. de Queiroz"
Tesla list wrote:
>
> Original poster: "Steve Conner"
>
>  >Jay Reed has sent me the following paper:
>
> Very interesting.. But it seems from the voltage plots (figs. 3 and 4) that
> his "optimized" design does not have complete energy transfer, since V2
> doesn't drop to zero when V1 peaks. Compare to the plot for Bienosiek's
> design, fig. 3, where v1 _does_ drop to zero.
>
> Maybe Reed's design is optimised for voltage gain only, and not energy
> transfer. But I would have thought that improving the energy transfer would
> increase the voltage gain too, and hence a coil with optimal voltage gain
> should automatically have 100% transfer.

Apparently he is trying to find a solution similar to what can be done in a Tesla coil, by increasing the input capacitor and readjusting the coupling coefficient (not changing the coils), while keeping the mode.  This can result in a small increase in the voltage gain, at the expense of incomplete energy transfer. Of course, if the coils are redesigned, with the same input capacitor complete energy transfer is again possible, with still higher gain.  I find strange that the author apparently doesn't realize that with a transformer any voltage gain can be obtained, in design at least.  So, comparisons with designs with other gains doesn't make sense.  I tried to work an analytical solution for this problem, but so far didn't obtain anything simple.  I tried too numerical optimization, but much to my surprise the optimizer finds a solution, changing the mode, that results in higher gain than the solution that keeps the mode.  (See the optmag program in http://www.coe.ufrj.br/~acmq/programs).  The 1:2:3 mode, for example: The coupling coefficient for the complete energy transfer case is 0.6742.  Multiplying C1 by 2.134 and changing the coupling coefficient to 0.7156, the voltage gain is multiplied by 1.21 at the second output peak.  It's possible to obtain even more, but several peaks away.  Antonio Carlos M. de Queiroz.

 

Date : Wed, 11 Aug 2004 07:39:58 -0600.  Subject : Re: Designing High-Gain Triple Resonance Tesla-Transformers

Original poster: "mercurus2000" <mercurus2000@cox.net>

Do you have links to Harvey's alternator driven coils? I would love to see them.
> conveniently-sized air cored coils. I've seen pictures of Harvey Norris's
> alternator-driven resonance coils and they're _BIG_ :o

 

Date : Wed, 11 Aug 2004 13:03:08 -0600.  Subject : RE: Designing High-Gain Triple Resonance Tesla-Transformers

Original poster: "Day, Michael"

It would appear that obtaining coupling coefficient of 0.6742, and 0.7156 may present a formable problem.  These coupling coefficients appear unusually high as compared to the coupling coefficients that I have seen quoted by the average coiler hobbyist.  Does anyone on this list have any experience(s) in obtaining comparable couplings that they would be interested in sharing?

To date, the only resonant transformers that I have seen having such high coupling coefficients have been of the spiral strip design that include inner and outer ring cages to shape the electric field in the margins of the transformer.  (See, for example, IEEE Transactions on Nuclear Science, Vol. NS-26, No. 23, June 1979, pp4211-4213.)  What is not clear to me in the spiral strip design is how the ring cages are connected.  That is, are the ring cages connected to their respective inner and outer winding?

Are there designs, other than the spiral strip design, that can provide such high coupling coefficients, and is there a preferred design?  Mike Day.

<snip>

(See the optmag program in http://www.coe.ufrj.br/~acmq/programs
The 1:2:3 mode, for example:  The coupling coefficient for the complete energy transfer case is 0.6742.  Multiplying C1 by 2.134 and changing the coupling coefficient to 0.7156, the voltage gain is multiplied by 1.21 at the second output peak.  It's possible to obtain even more, but several peaks away.  Antonio Carlos M. de Queiroz.

 

Date : Fri, 13 Aug 2004 07:42:58 -0600.  Subject : Re: Designing High-Gain Triple Resonance Tesla-Transformers

 Original poster: "Antonio Carlos M. de Queiroz"
Tesla list wrote:
>
> Original poster: "Day, Michael"
>
> It would appear that obtaining coupling coefficient of 0.6742, and 0.7156 may present a formable >problem.  These coupling coefficients appear unusually high as compared to the coupling >coefficients that I have seen quoted by the average coiler hobbyist.  Does anyone on this list have
> any experience(s) in obtaining comparable couplings that they would be interested in sharing?

Systems with these high couplings are not usually built as our open air systems. They must be encased in pressurized gas, oil, or solid dielectric for proper insulation.  It's possible, however, to make a magnifier driver with quite high coupling by using a flat primary coil and a short solenoidal coil. To increase the effective coupling, the bottom of the secondary coil can be connected to the top of the primary coil (ground the outer end of the primary coil, and connect the inner end to the bottom of the secondary coil, both wound in the same direction). Make the primary coil conical, and coupling coefficients around 0.7 can be easily obtained.

> To date, the only resonant transformers that I have seen having such high coupling coefficients >have been of the spiral strip design that include inner and outer ring cages to shape the
> electric field in the margins of the transformer.  (See, for example, IEEE Transactions on Nuclear >Science, Vol. NS-26, No. 23, June 1979, pp4211-4213.)  What is not clear to me in the spiral strip >design is how the ring cages are connected.  That is, are the ring cages connected to their >respective inner and outer winding?

Couldn't find the article. Too old to be on the IEEE site.  Antonio Carlos M. de Queiroz.

 

Date : Fri, 13 Aug 2004 17:24:25 -0600.  Subject : RE: Designing High-Gain Triple Resonance Tesla-Transformers

Original poster: "Day, Michael"

>It's possible, however, to make a magnifier driver with quite high coupling by using a flat primary >coil and a short solenoidal coil. To increase the effective coupling, the bottom of the secondary >coil can be connected to the top of the primary coil (ground the outer end of the primary coil, and >connect the inner end to the bottom of the secondary coil, both wound in the same direction). >Make the primary coil conical, and coupling coefficients around 0.7 can be easily obtained.

Most solenoid coils that I have seen having such high coupling coefficients appear to be vulnerable to HV breakdown at the final turns of the secondary windings.

The spiral strip transformer, on the other hand, produces uniform voltage grading through the thickness of the secondary winding.  The equal potential lines outside the windings, however, bend sharply around the edges of the thin winding conductors creating highly enhanced electric fields which results in electrical breakdown.  The function of the ring cages is to maintain the coaxial field distribution across the margin which is nearly parallel to the uniform field throughout the thickness of the winding.

The outer cage appears to be connected to the primary turn.  FIG. 7 of U.S. patent No. 5,079,482 provides a clear illustration of an outer cage.  It is not clear to me how an inner ring would be connected.

  >> To date, the only resonant transformers that I have seen having
  >> such high coupling coefficients have been of the spiral strip
  >> design that include inner and outer ring cages to shape the
  >> electric field in the margins of the transformer.  (See, for
  >> example, IEEE Transactions on Nuclear Science, Vol. NS-26, No.
  >> 23, June 1979, pp4211-4213.)  What is not clear to me in the
  >> spiral strip design is how the ring cages are connected.  That
  >> is, are the ring cages connected to their respective inner and outer winding?

>Couldn't find the article. Too old to be on the IEEE site.

The article gives the best description I could find.  The description would indicate that the inner cage rings have significant differences from the outer rings.  I have other references that illustrate ring cages in pulse transformers, however, they do not provide much of a description.

If the inner ring cage is merely to control the equipotential lines in the transformer margin, then might they not be left floating?  Mike Day.

 

Date : Sat, 14 Aug 2004 20:10:47 -0600.  Subject : Re: Designing High-Gain Triple Resonance Tesla-Transformers

 Original poster: "Antonio Carlos M. de Queiroz"
Tesla list wrote:

> Original poster: "Day, Michael" <Michael.Day@USPTO.GOV>

> Most solenoid coils that I have seen having such high coupling
> coefficients appear to be vulnerable to HV breakdown at the final
> turns of the secondary windings.
>
> The spiral strip transformer, on the other hand, produces
> uniform voltage grading through the thickness of the secondary
> winding.  The equal potential lines outside the windings, however,
> bend sharply around the edges of the thin winding conductors
> creating highly enhanced electric fields which results in electrical
> breakdown.  The function of the ring cages is to maintain the coaxial
> field distribution across the margin which is nearly parallel to the
> uniform field throughout the thickness of the winding.
>
> The outer cage appears to be connected to the primary turn.  FIG. 7 of
> U.S. patent No. 5,079,482 provides a clear illustration of an outer
> cage.  It is not clear to me how an inner ring would be connected.

This patent is very interesting. Has even a calculation method for the inductances and mutual inductances. The code that I have can reproduce the primary inductance correctly, but can't simulate the secondary coil.  Something to update.  There is an outer cage connected to the primary coil and an inner cage connected to the center of the secondary coil. (For who didn't see the patent, the transformer uses a single turn primary coil and a flat secondary coil inside it, both wound with flat strips.).  Antonio Carlos M. de Queiroz.

 

Date : Sat, 14 Aug 2004 20:12:50 -0600.  Subject : Re: Designing High-Gain Triple Resonance Tesla-Transformers

 Original poster: "Dr. Resonance"

Going to a solenoid coil will give you a lot of "insulation headaches".  You can wrap a lot of polypropylene around the lower part of sec, but high freq. currents can "capacitively couple" right thru insulation due to displacement currents.

Since you are at present only in the design stage, it would be better to consider a much larger coil radius thus providing better coupling and still using a flat spiral for the pri.

Use Antonio's program and run a lot of "what-ifs" until you get the result you need.  Dr. Resonance.

 

Date : Mon, 16 Aug 2004 12:49:27 -0600.  Subject : RE: Designing High-Gain Triple Resonance Tesla-Transformers

 Original poster: "Day, Michael"

Interesting.  U.S. patent 5,079,482 does appear to include some a novel expression for the mutual inductance of a spiral strip, as well as the self-inductance of the secondary spiral winding.  These expressions would facilitate the implementation of the spiral strip design without having to resorting to external tuning inductors.

Patent 5,079,482 cites US patent 4,931,700, also by the inventor Jay L. Reed, which provides details of the electron beam gun.  The electron gun generates an intense relativistic electron beam that does not require an evacuated envelope.

I wonder who this work was done for, and if these devices were ever built.  Mike Day.

>This patent is very interesting. Has even a calculation method for the inductances and mutual inductances. The code that I have can reproduce the primary inductance correctly, but can't simulate the secondary coil...Something to update.  There is an outer cage connected to the primary coil and an inner cage connected to >the center of the secondary coil. (For who didn't see the patent, the transformer uses a single turn primary coil and a flat secondary coil inside it, both wound with flat strips).  Antonio Carlos M. de Queiroz.

Top of Page