[Home][2014 Index] Re: [TCML] Yet another TC modeling program [Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [TCML] Yet another TC modeling program



Bart wrote:

> I assume we could load the spice into any arbitrary spice
> program (LTSpice, etc..).

That's the intention.  Maybe some flavours of Spice will cope
better than others with the large sub-circuits.  I've never put
that to the test.  Here I use ngspice under Linux which does OK.

You can test a Spice program if you like, using a sub-circuit
generated by tssp (the new program makes a very similar file).

A sub-circuit for Marco Denicolai's Thor system,

 http://abelian.org/tssp/thors.cir

produced by tssp.  It's a 4 terminal 'device' although it also
implicitly uses the ground node '0'.

This diagram demonstrates the connections to the sub-circuit,

 http://abelian.org/tssp/thors.model.gif

N1, N2, N3, N4 are the four pins of the sub-circuit and the
other numbers are node numbers in my spice model.  The C to
ground from the secondary includes all the Cext from the coil
plus the topload.

Cpri is 96.7 nF and you might have to experiment with Rgap.
VB is a zero volt source which exists only to measure the
base current.

I plotted (back in 2004) some base current waveform comparisons
for Thor, comparing transient simulations from tssp and Spice,
against the measured current for a normal firing with Vpri of
around 14kV.

 http://abelian.org/tssp/thors.compare.gif

Not bad over the first beat.  After that the small differences
in mode frequencies and Q factors lead to increasing divergence.

We use secondary base current because it is the most revealing
measurement - top volts (pin 3) would not show the overtones
anything like as well.

Here's a link to Thor

 http://www.saunalahti.fi/dncmrc1/

--
Paul Nicholson
--
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla