Experimental tesla coil using Radfio Shack Megacable Speaker wire as primary

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

Original poster: "Harvey Norris" <harvich-at-yahoo-dot-com> 

Normally a TC has a primary of relatively low
inductance, and relatively high capacitance, where the
capacity is a value close to , or above the ability of
the NST to provide that reactance current. ( Example;
the 30 ma NST can supply about 5 nf of capacity, and
in LTR designs we use 1.5 that value or about 7.5nf)
This necessitates designing the primary inductance
value around the capacity we intend to use. (Tapping
the primary at various points to reach tuning)
In this design however a relatively large inductance
would be tried, and since the inductance of the
primary is so large, the capacity is of course highly
reduced from the norm.  The inductance of 4 layers of
megacable speaker wire, that appear as pancake style
geometries is 2.3 mh.  However because the braided
wire is ~ 1/8th inch wide, and the winding layers are
packed tightly the primaries themselves have
significant internal capacity, and even the way we
route the 4 layers as interconnections seems to
influence that internal capacity, meaning that
significant internal capacity also exists between the
layers themselves. Because of these factors it is
projected that after we find the resonant value of
capacity to correspond to the resonant frequency of
the secondary, this may turn out to be the wrong
value, becuase those calculations do not account for
the internal capacity of the primaries themselves.
(people may argue that point, but in impedance tests
at 480 hz I found that this does not change, even
though we can widely vary the internal capacity of the
primary structure by layer routings, implying that the
internal capacity does not significantly change the
inductance)  It is thought that for the case of a
330,000 hz resonator,( my secondaries resonant
frequency) the needed capacity will be at least half
the value obtained by direct calculation, again
because the calculation does not talke into context
the internal capacity of the primary layers.

If one goes into Radio Shack and takes a look at the
MegaCable speaker wire package, it comes in a ready
formed pancake spiral.  The dimensions are as follows;
From; Influence of Internal Capacity on Spiral
Corridors
http://groups.yahoo-dot-com/group/teslafy/message/1238
"Granted when a braided flat cable is wound we know
that capacity exists between each layer, where the
dimensions of these megacables are about 9.5 inch OD,
2 inch ID, and about 37 turns on each layer, and 4
layers in magnetic agreement comes to about 2.3
millihenries. One might wonder that if we change the
relationship of how the internal capacity is made,
does this change the inductance? From my
experimentation at 480 hz, no it does not. Maybe at
higher frequencies it does. Some math folks might like
to know that experimental info, so it is mentioned.
And each copper braid is about 1/8 inch wide, ( if it
is up to standard specs!, watch out there!)
and when you stack 4 layers together they come to
about 1 inch. So the H/d ratio is about 1/9.5 or
.105."

A particular alternator resonant circuit was used to
scope out the resonant frequencies of the primary
layers, where the TEK scope has a stated internal
capacity of 30 pf;

Scoping of identical Return wind spirals in 4
layers/50 mv/div:1 us/div
http://groups.yahoo-dot-com/group/teslafy/files/DEL/Dsc00680.jpg
Cycles appear is 2.2 us. or ~ 454 khz.

In this method the spirals are stacked in identical
directions, and the inside spiral connection is routed
to the next layers outside connection.

Now we take the top set of spirals and turn them
upside down so they are in reverse direction to the
bottom set. But we also reverse the ordering
connections to the top set of spirals. We can call
this
arrangement an inverse bifilar. Magnetic unity is
still preserved, but we have enhanced the internal
capacity between layers, which means the resonant
frequency should go down. The first version of
this is to start from the outside of clockwise layer
1, connect its inside winding to the inside of
counterclockwise layer 3, outside of 3 to outside of
2, inside of 2 to inside of 4, and end at outside of
4. Given an example of 200 volts across total amount
of winds now we have more non linear complications
involved that between layers 1 and 3, layer 4 exists
in between, so instead of just 50 volts between
layers we have 100 volts in the analogy. But it gets
more complicated then that. Now the voltage difference
between winds, which was formerly linear, now becomes
non-linear just like the example of the non linear
voltage difference bewteen adjacent winds. In any case
we have increased the voltage between the first
corridor between layers 1 and 2, because an extra 50
ft is in between there. Between the next layer
difference, this is also higher then the first example
in TOTAL CV^2 difference because of the non-linearity
issue. And the remaining layer difference between 3
and 4 then resembles the first corridor example
because again 50 ft of spiral is between those
connections. In this example we have sandwiched an
extra 50 ft between two of the layer connections, but
not all three of them.

In this example we see that 3 cycles now
occur in 9.4 us. or ~ 319 khz. The resonant frequency
has been reduced 70% from the original winding method
of returned winds.
1st case inverse bifilar winds
http://groups.yahoo-dot-com/group/teslafy/files/DEL/Dsc00677.jpg


It is this winding method that is contemplated for
design usage as a tesla primary.  Going a bit further
in theory, if the scope data is accurate, we have
engineered a primary set up that actually already
resonates WITH THE STATED SCOPE INTERNAL CAPACITY OF
30 PF, so as to be ALREADY close to the secondaries
resonant frequency!  However 30 pf is indeed a small
capacity to be put in series with a primary, so
modifications of the primary arc gap are contemplated
to model that of the Marx voltage doubler, where the
source will see 4 times the capacity before arc gap
firing, and upon firing the capacity value for primary
resonance will appear as two capacities in series. To
replicate the 30 pf scoping conditions that enabled
the noted resonant frequency scoping observation, we
find that two 60 pf values in series meets that
requirement. And prior to arc gap firing we then have
a total of two 60 pf values being oppositely charged
up in parallel (Schematically inversely) for a total
capacity charging of 120 pf, which is slightly below
1/4 of the 30 ma NST's charging ability.  Because of
this we might suspect that we will be able to start an
arc gap at very low firing voltages, and in fact with
very large inductances 5 times these stated values, I
have investigated this type of MARXIST arc gap, and
found that it would fire at half the voltage that the
conventional arc gap would fire, this with about only
a 10 volt variac input into the NST, using a very
narrow arc gap.  So we are actually charging two
capacities in parallel in opposite polarity
connections, where each capacity is in series to the
return potential end of the supply transformer with
two of the four layers, but these layers have a very
high mutual inductance between them, being adjacent
layers of spirals.

Now if all of this were true, we ought to be able to
also add the secondary in series with the primary,
with the secondary placed on top of the primary on
this scoping technique, and since each component is
roughly equal in resonant frequency, the new resonant
frequency that we see should be half of the original
resonant frequency we first observed on just the
primary alone.  This is just what occured, after
making some modifications to the scoping technique.

Secondary- primary series connection to ambient
scoping./ no outside connection
http://groups.yahoo-dot-com/group/teslafy/files/DEL/Dsc00683.jpg
"just about what I was looking for, a cycle time ~
twice what I was getting from the primary there shown
at 7 us." ( Poor pic to be redone)

The following info above is just the bare bones of
what the research showed, and a slew of other
complications are involved, including the fact that
the resonant frequency scopings are obtained WITHOUT
ANY EXCITATION MADE ONTO THE TESTED COILS, WHICH IS
TERMED AN AMBIENT SCOPING.  This gives much cleaner
trace signals then what the actual test gives, which
is a multitude of multitraced signals, all in harmony
however. The first energized test that makes the coils
ring via an adjacent neon firing still shows the
ringing later on, even though that neon is not
energized.  The bottom metal plate connection that the
spirals rest on has been shown to be connected to
comparatively large capacities from the alternator 480
hz resonant system, and somehow, someway the
environmental ambient charging of larger capacities is
thought to make the sensor coil ring at its resonant
frequency. With no other power input then the scope
itself, you can turn the scope on and record a .4 volt
high frequency signal coming from the megacables being
tested, but once we remove the connection of the metal
plate to the capacities, the signal is reduced
considerably. In fact if a halide lamp, (I often use
this for taking shaded scope pictures) is connected to
the same utility strip that the scope is hooked to,
then the  ambient scoping becomes a larger 1 volt
signal.

This whole subject and how the initial fired scopings
were made is quite complex, so I have tried to stick
to the relevant points here, and maybe this info will
be helpful to others who wish to try using megacable
speaker wire spirals as experimental tesla primaries.
One might try firing a tesla coil at some distance
away from a scoped megacable spiral set, and see if
residual signals remain on the scope after the tesla
coil is shut off, or maybe also running a metal plate
under the spirals connected to the TC's capacity after
shutoff still produces the ambient ringing signals. Or
perhaps folks can go the conventional route using a
signal generator to confirm the same resonant
frequencies coming from these spirals as I have
obtained, but remember the internal capacity of the
scope you use will also change that signal, we are
actually only measuring how the spirals resonate to
the scopes internal capacity.  Also significant for
theorists such as Paul Nicholson is that fact that
this unusual geometry and large internal capacity
seems to indicate frequencies WELL BELOW what the
quarter wave value for 200 ft  of flattened speaker
spiral cable would yeild.

At Re: [teslafy] Influence of Internal Capacity on
Spiral Corridors
http://groups.yahoo-dot-com/group/teslafy/message/1243

I mention some mistakes made on the original post, and
give some reasoning why even a lower resonant
frequency is found by an additional inverse bifilar
routing method. Sorry for the inconvenience that non
yahoo members might not be able to access my site, but
it is at
http://groups.yahoo-dot-com/group/teslafy/
if I remember the correct spelling here...
Sincerely HDN