# Re: FW: secondary wavelength-matching/s.s. drive

John & all-

You know, I've not paid too much attention to those formulas, for these
reasons that I reiterate:

1.  I don't need to calculate Fr; my system tunes itself.

2.  I persist in viewing the optimum toroid as nothing more than the
means for allowing the secondary voltage to rise to near the maximum that
would otherwise be available (with a super-optimum toroid).  Not >to< the
maximum, since there would then be no spark, but sufficiently near it so
that a) near-maximum energy has become stored in the secondary and b) a
spark will break out with regularity.

The optimum toroid must have a) an o.d. sufficiently large so as to
adequately shield the top turns of the secondary against spark break-out
from them and b) a smallest effective cross-sectional diameter (taking
into account surface irregularities) only large enough so that break-out
from it occurs comfortably below the maximum voltage that could otherwise
appear.  It seems to me that the o.d. is pretty easy to choose; a fairly
wide variation is OK with respect to the coil o.d.  The surface
smoothness and c.s. diameter, however, are critical factors; lack of
smoothness will shoot all to heck any assumption one might make as to a
required c.s. diameter.  Hence the [expected and anticipated!] 32 u-inch
smoothness specification on those commercial toroids that are on order.

I just see a can of worms in regard to trying to calculate the optimum
toroid for use with the ordinary double-tuned, open-loop (i.e., not-
frequency-regulated) spark-gap Tesla coil, for these reasons:

1.  A well-constructed secondary coil will have a fairly high Q and thus
its resonant-frequency peak will be fairly sharp.  And the exact
frequency of that peak, and the Q of the coil, will be strongly
influenced by the presence of electrical conductors in the vicinity of
the coil.

2.  We depend critically on the secondary's self-resonance and Q to boost
the voltage.  But the voltage we might get is also going to be strongly
affected by a) the match between the primary's and the secondary's
resonant frequencies, which will vary in uncontrolled ways; b) the
coupling between primary and secondary which is ordinarily pretty
unquantifiable; c) the rate of decay of the primary's damped oscillation
which is strongly dependent on the resistance of the spark gap and its
variation over time; d) the actual voltage, and its variation over time,
at which the gap fires; e) the propensity of conductors adjacent to the
secondary to absorb energy and thus reduce the attainable voltage; and no
doubt a host of other factors I don't know about.

So there are a lot of factors not satisfactorily quantified.  In my
system, what remains not quantified are a) actual secondary Q at any
instant; b) actual secondary Fr at any instant but I don't care about
that; c) primary:secondary coupling; and d) energy absorption by adjacent
conductors.  So I would still have a problem in choosing an optimum
toroid.  What I have done is just cut the Gordian knot and order one of
the 6" x 24" toroids.  I'm pretty sure it's going to break out with a
satisfactorily-long spark, especially when I finish my present task of
doubling my primary current since with my present current I get 2 ft.
sparks from a corrugated toroid, with the voltage rising to only 3/5 or
so of the available maximum.  But that's fuzzy reasoning; I wish I could
do better.

The fundamental problem with toroids is, of course, that one can't tweak
their diameters to suit the voltage one happens to be able to attain.  I
thought once that I'd designed one where I could, but I failed to
understand the necessity for extreme surface-smoothness.

On the subject of efficiency--

What we all >really< mean by that is shown by the formula, 100 x
(ZAP!)/(input power), expressing percent efficiency.  But unfortunately,
ZAP! is hard to quantify.  The other way of expressing it is how I did
and that can be quantified.

Ken Herrick

On Sat, 30 Sep 2000 11:47:57 -0600 "Tesla list" <tesla-at-pupman-dot-com>
writes:
> Original poster: "John H. Couture" <couturejh-at-worldnet.att-dot-net>
>
>
> Ken -
>
> What did you think about Tesla's idea of using the 1/4 wavelength to
> find
> the optimum toroid size of any secondary coil? I thought it was
> clever. How
> would you do it?
>
> Thanks for the information on the MOSFET efficiency. I was thinking
> of the
> overall efficiency = output/input. The input is no problem (1320
> watts) but
> the output from the secondary coil has always been subject to
> varying
> opinions.
>
> John Couture
>
> ---------------------
>
>
> -----Original Message-----
> From: Tesla list [mailto:tesla-at-pupman-dot-com]
> Sent: Friday, September 29, 2000 3:43 PM
> To: tesla-at-pupman-dot-com
> Subject: Re: FW: secondary wavelength-matching/s.s. drive
>
>
> Original poster: "Kennan C Herrick" <kcha1-at-juno-dot-com>
>
> John (Couture) & all--
>
> The main point of my posting was that some of the considerations you
> mentioned in your original posting I don't find to be important with
> the
> apparatus I use.  I don't much care what the secondary's resonant
> frequency is since my primary apparatus adjusts itself to exactly
> that
> frequency--cycle by cycle, as a matter of fact.  Since I don't much
> care
> about the frequency, I don't much care about Ctor or Cself; I just
> build
> the secondary for good aspect ratio, reduced proximity effect, low
> loss,
> etc.
>
> As to efficiency--I've not measured it but I take it to be measured
> by
> input power from the line vs. power dissipated in the apparatus.
> The
> biggest power-wasters in the apparatus would be the MOSFETs via
> their
> Rd-s-on but I find that they barely get warm while the coil is
> running.
> I drive them pretty well so that they are well turned on, and they
> should
> exhibit close to their rated Rd-s-on of around 1/4 ohm each.  With
> 12 in
> parallel during each 1/2 cycle of the excitation and with the
> (indirectly-measured) 220A or so primary current (in my present
> setup),
> that yields ~1KW Pdiss in the MOSFETs.
>
> But that's just during the pulse-burst.  With a 5% duty cycle, I'd
> draw
> ~11A from the power line at 120V (that's 5% of 220A; all the line
> current
> goes into the electrolytic capacitors that drive the primary via the
> MOSFETs; their current in = their current out).  That would yield
> ~1300W
> continuous input while running.  With the same 5% duty cycle, the
> average
> Pdiss of the MOSFETs would be ~50W.  Thus my efficiency (neglecting
> other
> factors, of course) would be ~100 x (1300 - 50)/1300 = ~96%.
>
> It's not really as good as that, of course, due to other factors,
> but
> I'll bet it's better than 90%.
>