[Prev][Next][Index][Thread]

Re: Dump the Lump

Hi Robert,

	I quoted both your posts today in this single post.

>Hi all,
>
>
>Thanks for the responses. My posting was just a vote. I have already written
>perhaps four pages or more on the subject to the list and Terry.  I was  gob
>smacked to find one of my posting pop up in a search for Corums papers. How
>does that happen Terry?

A miracle of web search engines.  The pupman site has been around for years
and has been well "drilled" by all the popular search engines.

>
>
snip
>
>
>First I should restate my views. The theory is not mine even Tesla reported
>it.

I should also point out that the views I express are certainly not of my
invention either.

>
>
>In its usually operation a Tesla coil secondary resonates due to the
>reflections of signals from what is normally the high voltage end in a
>similar way to a 1/4 wave antenna or any quarter wave stub (as it is
>referred to when not used as an antenna). A good analogy is how an organ
>pipe resonates. The coils behaviours as a classical transmission line due to
>its distributed capacitance and inductance which result in the comparatively
>slow propagation of ALL SIGNALS along the axis of the coil.  Typically a
>thousands of time slower than the velocity along a wire or coax cable.  The
>resonance occurs if the reflected signal returns to the driving end with the
>opposite polarity to that of the driving signal.  For an open circuit
>transmission line the minim length at which this can occur is 1/4 of a
>wavelength. If the line is terminated with capacitance, inductance or a
>short circuit the phase of the reflected signal is modified and hence the
>line will resonate at a different length. It can also resonate at odd
>multiplies of a 1/4 wave length again due to the phase of the reflection.
>The fundamental requirement for this to occur in the average Tesla coil at
>say 100KHz is the slow propagation of signal along the axis of the coil.

In the organ pipe, the resonant frequency is based on the time it takes for
the sound to travel from one end to the other.  The frequency is simply a
function of the length of the pipe and the speed of sound in the air.  If
sound travels 1000 feet per second and the pipe is 10 feet long the 1/4
wave length is 10 feet giving 250Hz.  I triump of resonant theory.  I would
be hard pressed to come up with a lumped model for that! :-O

In a straight wire whip antenna, the same holds true.  A CB radio antenna
102 inches long (plus about 5 inches for the spring and mounts) represents
almost exactly the 1/4 wavelength of the radio's ~27.2MHz signal.

In a coaxial cable, the length is all important given the velocity factor
of the cable.  If I want to set up a 1/4 wave standing wave in RG-8 with a
velocity factor of 66% at 13.56MHz, I get a cable 12 feet long based on
good o'l transmission line theory.

So, the question is, what "length" do I use for a Tesla coil and what
"velocity factor"?  I can certainly find the wire length and the coil's
dimensions.  But how do I calculate the velocity factor so I can calculate
the Fo frequency???

As a simple example:
My big coil is 10.25 inches in diameter, with 1000.5 turns of #24 wire that
is space wound to give a length of 30.0 inches.  The wire is 2684.8 inches
long.  What is the resonant frequency???

I could use E-Tesla5 to calculate the distributed capacitance and Wheeler's
formula find Fo to about 2%.  Or, I could use Wheeler's formula and the
Medhurst formula to find Fo to about 1%, but those methods do not rely on
antenna theory, VSWR, Q, Smith charts, or 1/4 of anything.  What
transmission line formula gives these "provable" results and what equations
are used to find them??  I don't have the Corum's vacuum tube book any more
(I sent it to Malcolm ;-)) or I would try and figure it out.  In the past,
it has not worked...  Maybe Malcolm could give it a try if he hasen't
burned it already ;-))

I have 100 feet of RG-6 TV coax here with a printed velocity factor of 78%.
 It does indeed tune a standing wave at the predicted 1.92MHz although the
signal is plagued by the odd harmonics that are messing with my signal
generator a bit.  The top and bottom phase appears to be nicely 90 degrees
out of phase.  I have a scope probe hooked to the signal generator end of
the cable and the other probe just lying near the other end of the cable
enough to pick up the signal without apparently loading it.  So all that
transmission line stuff seems to work just fine when applied to a
transmission line.  I just don't see what it all has to do with my Tesla
coil... 

>
>
>A Tesla coil has all the theoretical requirement  of distributed capacitance
>and inductance to support  slow propagation.  I have personally measured the
>slow propagation in a coil that I also used in a Tesla coil including the
>90deg phase shift (terminated coil), more of this later.  There  are also
>some devices that use the effect such as the luminance delay line in a
>television and the delay line used in some scopes to  allow observation of
>the trigger pulse (the display channels are delayed).

Delay lines are long and thin to reduce end to end magnetic coupling.  They
also use many isolated capacitors to ground along their length.  I would
submit that Tesla coil secondaries would make very poor delay lines since
they are typically short and fat and the distributed capacitance from one
point to the next is far from isolated.

>
>
>One could assume two modes of signal probation for the coil.  One mode
>would have a slow propagation time and be used  to explain standing waves,
>signal delays, transmission line effects etc.. and the above effect.  The
>second mode would have a fast propagation time where the input signal is
>able to influence the whole coil  relatively instantly (fast compared to the
>period of the operating frequency of a coil).  This mode is used to explain
>how the inductance and capacitance of the coil behave collectively as
>lumped components at resonance.
>
>
>I far as I am aware the only direct evidence to support the view that a fast
>propagation mode exists at the fundamental resonance frequency of a coil is
>the approximately zero phase shift measurements between the top and bottom
>of
>a resonating coil. As Terry can testify, initially this appeared to me as
>strong evidence to
>support a fast propagation mode.  However this was due to my lack of
>understanding of the properties of a standing wave. The reason that it is
>called a standing wave is just that it does not move only its amplitude
>varies i.e. no phase shift. This is because it is the sum of a forward
>travelling wave and the reflected or backward travelling wave.  I have
>already posted this point so I am surprised that it is still believed there
>would be any measurable phase shift at resonance. I guess this point may not
>be understood or believed. It can be verified from theory and  modelling.
>If you want to measure the phase shift on a real system you must use an
>edge signal or terminate the coil so the reflection does not confuse the
>measurement.

The standing wave on my coax seems to have a nice 90 degree phase shift.  I
should make it clear that my Tesla coil phase shifts were done under real
disruptive operation.  Would that make a difference?  Perhaps there would
not be enough time for standing wave to be set up in a real operating
disruptive coil?

snip...
>
>For those individual that are not into transmission line theory and are
>convinced that a signal can propagate instantly down the coil. I suggest
>that if they have access to a signal generator, scope and a coil check it
>out. I have and it just doesn't.  You can readily observe the slow
>propagation, reflections and resonance  up to several times the fundamental.
>With the small coil I used (12in by 1.25in) it was even possible at
>resonance to light a small florescent tube and observe the multiple voltage
>peaks at the higher resonance modes.  Although I have not tried it I am
>certain the same would be confirmed with a Spice model and again  no instant
>signal.

The graph at:

http://users.better-dot-org/tfritz/VoltDistBare.jpg

Shows these higher harmonics for a Tesla coil.  Note that as the harmonic
number is increased, the voltage level decreases rather dramatically.
Would transmission line theory predict this??  I have not thought about the
implcations of this and the "lumped" models...

>
>
snip...
>
>
>I agree the lumped models can predict the majority of voltages a round the
>secondary and even the secondary top voltage.  I had believed that we
>discussing how the secondary works not how it may be modelled.
>
>
  
I think the Corum's paper suggests that lumped element modeling will not
work for anything...  Thus, there "Test II" to show how bad lumped modeling
fails.  But perhaps you are looking beyond that paper...

>
>
>It appears that several people accept that multiply resonances or harmonics
>occur but are unable to give a  explanation with a lumped model. Terry and
>his artist are on the verge of  grasping the  transmission line theory.
>They recognise the problem with the harmonics and have given a technical
>justification for how such harmonics may be supported and they still cling
>to the erroneous phase issue. I believe the problem at least for Terry is
>that it is difficult to accept that signals can not instantly propagated
>down the coil. This is difficult to accept.  For me it was helped by seeing
>it on the scope so check it out Terry you have all the equipment.  The
>artist and or Terry have suggested that coil is a series of parallel tuned
>circuit that can resonate at different frequencies I don't see how it can do
>that.  But you are getting close just add the capacitance to ground and a
>bit of coupling between the coils and you will have a good model . Plug it
>in to Microsim or spice and see how it works, instant transmission no,
>standing waves yes. If you want I will try and convert one of your coils in
>to a 10 element model.  It should be good for the fundamental and third
>harmonic. May also be able to prove that the turn to turn capacitance is
>insignificant which will allow you to use a standard L and C model  Then
>hopefully I want have to right an other long post on this subject.
>

I have really never thought about the possibility that the low phase
difference may be due to the fact that standing wave may not have time to
get setup.  I have always tried to find explanations that would support the
tiny phase shift observation.  I may now have a new possible explanation to
explore ;-) 

>
>Hi all,
>
>This is just a follow on from my previous posting.
>
>Reading it again I have probably clouded the explanation with a host of side
>issues. So I will start again.
>
>
>1) As in most transmission line all signals travel along the Tesla coil with
>the same definite velocity.  This is true of all signals with frequencies
>from DC up to perhaps 100MHz. So this includes the resonance frequency of
>the coil and all its harmonics. In the case of a Tesla coil the velocity can
>be a thousand time slower than the velocity in a an open wire. This has been
>verified mathematically, modellingly  and by experiment.  There is no
>evidence to suggest that this is not true or to suggest that any signal can
>instantly travel down the coil.  . Probably the most convincing evidence is
>obtained by performing an experiment with a signal generator, scope and
>coil.

How does one find or calculate this velocity??  
Of course, dividing the speed of light by 4 times the fundamental frequency
found by lumped element theory would not be very convincing proof of the
transmission line theories. ;-))  The CB antenna, the pipe organ, the coax
cable,... don't need lumped theory to find the resonances, so why do Tesla
coils? 

>You will be able to verify directly that no instant transmission occurs up
>to the limit of your generator and scope assuming that is no more than say a
>100MHz. If you require advice on how to set up such an experiment I am
>willing to help.

The inductance and capacitance form a low pass filter that would attenuate
instant changes. 

>
>I challenge anybody who believes different and wants to argue the point with
>me, to do the experiment, as I have done, and show that instant or fast
>transmission occurs. It should be possible to perform the experiment in less
>than an hour.

And the experiment is??

>
>2) Given that all signals (up to a limit) including those with a frequency
>of that of the fundamental resonance of a Tesla coil travel along the coil
>with a slow velocity, the simplest explanation of the fundamental resonance
>and higher order resonances is that they are standing wave resonances
>caused by reflection from the end. This also explains how a coil that is
>shorted at both ends can resonate.
>
>Note.  The velocity is not strictly constant it can vary with frequency
>caused by frequency dependent losses and dielectric constants. This is
>called dispersion.  In addition at high frequencies when the individual
>turns become resonate new effects occur.  For the average Tesla  air cored
>secondary coil wound with copper wire I would not expect the dispersion to
>produce any more than a few percent change in velocity up to say 10Mhz and
>the high frequency resonances will not occur until the wavelength (in air)
>approaches the length of one turn.
>The above assumes that the secondary is wound as a single helical coil.

I would agree that is very true.

>
>Do the experiment Terry,  I know you love  them. and put the results on your
>site. You will have the first definitive paper proving it one way or the
>other.  Although I suspect it may just  produce an interpretation debate.
>We can then take the parameters of the coil you have used and convert them
>to a transmission line model and get some one to plug them into microsim or
>spice. For my part I can use them in a theoretical calculation.

Please explain how to do the experiment.  I have pleanty of nice toys
standing ready...  I have already given a nice test coil secondary
specifications.

>
>
>Down with the lumped theory, long live the transmission line theory. Am I
>biased, yes, but then I've done the experiment and more. Come on guys get
>with the programme Dump the Lump.

We need to get the transmission line theory to predict something useful
first.  Most of our design equations are based solely on lumped theory and
they work just fine.  We need to get new equations based solely on
transmission line theory working.

snip...

Thanks for taking the time to write this long post.  I am looking forward
to doing any tests or exploring any ideas to a conclusion.  When we stop
looking for ideas, we stop having them....

Cheers,

	Terry