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

Re: Lumped vs. T-line - You be the judge...

Hi all,

I suspect as previously posted that high frequencies travel with a velocity
calculated by the standard transmission line equations.  Very high go even
faster and at low frequencies when the inductive coupling is in opposition
to the wave travels slowly. So the harmonics of your signal get to the
end first. A standard non dispersive transmission line does not have this
property and hence the edges remain intact. It just occurred to me that
if the distributed L and Cs are not constant this may also disrupt the
 edges. I can not account for the initial pulse possibly interturn C
but the maths don't support it or If you use very low C probe or poor
scope you may get an initial pulse connected to anything in the room.
A good trick is connect the probe tip to the ground point but keep
everything else the same you may be surprised how large a signal you
get.


When one end is shorted and the other is open it does have similar
 behaviour to a lumped tuned circuit at least at one point . But unlike
a tuned circuit it resonates at 1,3,5,7..etc. frequencies so when you
apply a transient you get the sum of each of the resonances but the
dispersion makes it look different than a non dispersive line.

Using a switched sinewave other experimenter/s have confirmed a
1/4 cycle delay at the resonance frequency and 1/8 half a long the coil.
It has also been show how the amplitude builds with each reflection
from the end.

.Could somebody please measure the propagation delay with
sinewaves at the resonant frequency and at say x2 X3 X4 X5
and X10 and X100 so the dispersion constant can be determined.

You could also try the L coupled model I am sending to Terry.
The model assumes a constant distribution of C which could easily
be changed. The fact that most transmission line models have
constant parameters along there length does not invalidate the
concept but in any case they dont have the dispersion due to the
coupled L.

Well we atleast we concluded  the phase issue.

Regards Bob

-----Original Message-----
From: Tesla List <tesla-at-pupman-dot-com>
To: tesla-at-pupman-dot-com <tesla-at-pupman-dot-com>
Date: 19 April 2000 13:58
Subject: Re: Lumped vs. T-line - You be the judge...


>Original Poster: "Malcolm Watts" <malcolm.watts-at-wnp.ac.nz>
>
>Hi Terry,
>                Well, let me throw my hat into the ring:
>
>> Original Poster: Terry Fritz <twftesla-at-uswest-dot-net>
>>
>> Hi All,
>>
>> I reran my tests with a TEKP6009 probe that has a loading of 2.5pF/10Meg
>> ohm.  This probe has about 1/6 the loading of a regular scope probe.
>>
>> A MicroSim model is here:
>> http://www.peakpeak-dot-com/~terryf/tesla/misc/T-line01.gif
>>
>> With the expected outputs here:
>> http://www.peakpeak-dot-com/~terryf/tesla/misc/T-line02.gif
>>
>> The two top models are for an ideal and lossy transmission line,
>> MicroSim's lossy line model is probably a little more accurate but the
>> outputs are almost identical in this case.  I have also allowed for the
>> limited bandwidth of the scope at very high gain to get a bit better
>> accuracy of the measurement system (it makes little difference).
>>
>> The expected outputs are very different now with the much lower loading
of
>> the probe.  The lumped case is a very nice sine function and the T-line
>> case is a fairly good square wave with some increasing reflections.  they
>> are now both very distinct
>>
>> The actual measured output is at:
>> http://www.peakpeak-dot-com/~terryf/tesla/misc/T-line03.gif
>>
>> The cleaner Excel plot is at:
>> http://www.peakpeak-dot-com/~terryf/tesla/misc/T-line04.gif
>>
>> Looks like it is really trying to be the lumped case but there are
>> definitely some T-line qualities to it also.  Notice the ragged area of
the
>> output in the first 800nS.  That must be the current tearing it's way up
>> the coil by induction hitting all the little resonances and such along
the
>> way.  There is a definite initial spike and another about 600nS latter.
>> Those are very real and repeatable.
>>
>> I am sure a very short inductor would looked lumped and a straight wire
>> would look like a T-line.  It looks like a Tesla coils is somewhere in
>> between with the lumped case being quite dominant.  Antonio and Malcolm
>> have suspected this and it certainly looks like they are right.
>>
>> So it looks like all the equations and theory behind lumped modeling and
>> Tesla coil calculations is very solid.  However, those that like
>> transmission lines now have me convinced that there are a few little
T-line
>> effects too.  I can't think of a situation were T-line modeling would
have
>> any advantage but at least we know better what we are dealing with.  I
>> think Antonio has really summed the situation up as follows:
>>
>> "The differences from the ideal transmission line are due to the fact
>> that a coil is not a TEM transmission line (a line where the electric
>> and magnetic fields are perpendicular to the direction of the line).
>> The magnetic coupling along the coil, predicted in the lumped model,
>> causes that initial rise. The transmission lines incorporated in
>> simulators, and treated in usual books, are all TEM models. They model
>> very well cables, but don't work very well for coils. Maybe a waveguide
>> model can be a better approximation."
>>
>> Cheers,
>>
>> Terry
>
>I for one have never believed that a uniform transmission line is an
>accurate model for a TC resonator. The capacitance distribution is
>all wrong. It is not balanced either although that argument may be
>considered passe in some quarters.
>          Nor have I believed (for a very long time) that it is purely
>lumped. The current distribution in a bare resonator speaks for
>itself.
>        It is all very well doing experiments when the coil is run parallel
>to a conductive ground plane but that is not how it is used in
>practice.
>
>Regards,
>Malcolm
>
>
>