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

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

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