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

Re: The secondary behavior

Hi Ken,

	I ran a model of putting a scope probe in the middle of the secondary.
Even a 2.5pF probe has a dramatic effect on the pure transmission line
model.  I makes the waveform very choppy and all when driven from a square
pulse.  It is hard to say what it all means, but probe loading in this
situation (actually any situation) has to be considered.  Unfortunately, I
cannot predict what will happen with modeling since this behavior lies in a
gray area between the lumped and T-line models.  Since the T-line
parameters are not uniform or "transverse", the typical T-line models are
of limited use. 

However, since most of the energy storage in a secondary is near the top of
the coil, I would suspect that the current would race up most of the coil
fairly quickly.  Near the top is probably where the small transmission line
delays may start to have an effect.

In order to really model the secondary, perhaps one could use many Ls in
series shunted to ground by many Cs.  The series R could also be put in too
since it is important.  The conductance (G) appears to be negligible.  This
would be much like a lumped parameter model of a transmission line.
However, the advantage would be that you could magnetically couple the
inductances and use arbitrary capacitance values along the line.  I would
think figuring out the inductances and coupling in such a model would be
fairly easy.  However, the distribution of the capacitance along the
secondary may be much harder.  Perhaps E-Tesla5 could be modified to pick
out a series of capacitances along the secondary.  However, this
distribution varies with every coil configuration so there probably would
not be a nice universal model that would work on every coil.  But maybe a
single distribution would be "close enough".  I don't think such a model
would have any practical value, but it may be able to reproduce the subtle
secondary behavior very accurately to confirm that we are not nuts! ;-))

Also, if we could accurately predict the secondary behavior at this
microscopic level with a "lumped model", it would really be funny!  :-))  I
don't know if there are any pure transmission line models that could cover
this situation.  Antonio has mentioned waveguide theory but that is beyond
me.  I would think the equations would be overwhelming to all but a few
people.  But a piecewise lumped model could run on MicroSim and feed
information to a waveguide modeling effort.  

We are very lucky lumped models and equations work so well.  Simple
transmission line theory is badly hampered by the non-uniform capacitance
along a secondary.  It is interesting to note that my 9.3pF coil
capacitance needs to be pushed to 23pF (C parameter) for a T-line model.
The large energy storage at the top of the coil is smeared out along an
unrealistic larger uniform capacitance in a standard T-line model.  If
there was a T-line model that could us a linear equation as a function of
distance for the capacitance distribution and be able to uniformly couple
the inductance along the coils length, that may be perfect.

Cheers,

	Terry


At 12:59 PM 04/17/2000 -0700, you wrote:
>I haven't been fully following all the details of the various
>transmission-line postings on this subject, but I'll toss out the
>following that I just noticed:
>
>Using the gated Wavetek 115 setup about which I've written, I applied,
>this time, a pulse burst ~12 cycles long, at resonance of ~96 KHz.  
>
>I put my 2nd scope probe about 1/2 way up on the secondary coil (coil
>standing upright perhaps 18" above 3" concrete on (earth) grade.  
>
>I noticed this interesting thing:  (Equalizing the 1/2-way and top
>amplitudes as seen on the scope screen), the very first peak of the
>half-way-up wave occurs ~2 us ahead of that at the top.  But then, within
>2 cycles, it "catches up", and all subsequent cycles of mid vs. top are
>>in phase<, including those while the coil rings after excitation ceases.
>
>I also tried looking for any differences in the behavior of the 1st few
>cycles (of the top signal) with a) driving resistance from the signal
>generator varied between ~5 and 50 ohms and b) the top toroid in place or
>not.  Readjusting the frequency for resonance as regards b), I couldn't
>see any differences to speak of, with the 1st cycle spread all across the
>scope screen.
>
>But that half-way-up behavior intrigues me: could that be what's "slow"?
>
>Just thought I'd muddy the water a bit.  And you're right, Terry...this
>horse ain't dead!
>
>Ken Herrick
>________________________________________________________________
>YOU'RE PAYING TOO MUCH FOR THE INTERNET!
>Juno now offers FREE Internet Access!
>Try it today - there's no risk!  For your FREE software, visit:
>http://dl.www.juno-dot-com/get/tagj.
>