Re: The PING Test

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Paul Nicholson by way of Terry Fritz <twftesla-at-qwest-dot-net>" <paul-at-abelian.demon.co.uk>

John Couture wrote:
> It does make sense and I showed in my post how any coiler
> can make the test and do the calcs.

John's ringdown test if fine, as an indicator of Q.  Many of us
use it.  By giving you the secondary Q, you thereby have the
energy storage efficiency of the secondary.  But it says nothing
about the system efficiency or output voltage, except in base-
driven CW mode, in which case Vtop/Vbase = Q applies (approximately).

In any system CW or impulsed, any definition of efficiency boils
down to asking what fraction of that initial energy ends up in
the useful load (whatever you define that to be).  

Therefore, in order to determine the fraction of energy leakage
which is defined as 'useful', you must take *two* Q measurements
(by ringdown, bandwidth, whatever you like) and compare them in
the usual way, ie 

  Efficiency = 100% times (Q_unloaded - Q_loaded)/Q_unloaded.

Any formula claiming to show any sort of transfer efficiency 
based on just a single Q value is a nonsense.  By definition, it
must evolve from the difference between the loaded and unloaded
behaviour of the system, regardless of what is considered to be
the 'loading' and what is considered to be the 'loss'. 

The lamp tests that John has described in the past will at best
give the efficiency of coupling to the lamp.  As such, it might
be a useful setup aid, but no better in principle than simply
monitoring the secondary base current.

As regards John's question:
> Should the  secondary turns be increased or decreased to
> increase the secondary voltage and the spark length with
> the same H/D coil?

The glib answer is that it depends whether you've got too 
many or too few turns to begin with.  The question needs to
be rephrased to make sense.  It should say: is there an optimum
number of turns for a secondary, given some constraints such
as a particular power supply, BPS, firing voltage etc, and 
allowing Fres, Cpri, Lpri, k, turns, wiresize, all to be free
variables?  And if so, how do we calculate it?

We have all the pieces with which to answer that question, bar
two.  The least important of the two missing links is to
be able to calculate the Q factor of the primary and secondary
coils.   Much more serious is that we don't understand how the
secondary couples to streamer loading - the dynamics of the 
coupling determine how well a particular impedance of secondary
will drive streamers from a given top terminal.  If we could
determine this, even just empirically, it would complete the
theoretical package necessary to answer the simple question:
"How many turns shall I put on the secondary?"  

So John is asking a question that we can't answer at the moment.
--
Paul Nicholson
--