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

Re: primary coil



    [The following text is in the "ISO-8859-1" character set]
    [Your display is set for the "US-ASCII" character set]
    [Some characters may be displayed incorrectly]



----------
> From: Tesla List <tesla-at-poodle.pupman-dot-com>
> To: Tesla-list-subscribers-at-poodle.pupman-dot-com
> Subject: primary coil
> Date: Wednesday, February 05, 1997 12:25 AM
> 
> Subscriber: laakkone-at-icenet.fi Tue Feb  4 22:09:04 1997
> Date: Tue, 04 Feb 97 15:47:37 PST
> From: Calle Laakkonen <laakkone-at-icenet.fi>
> To: tesla-at-pupman-dot-com
> Subject: primary coil
> 
> I need to how to build the primary coil.
> Some TC instructions doesn't tell much.
> Does somebody know what kind is good primary ?
> 
> 
> 
>      - Calle Laakkonen

Calle,
in case you missed the original post, here is a copy of a short article I
wrote on primary coils. It may help you understand what the primary is
doing and make decisions as to what type of primary is appropriate for your
particular coil.

It is not a construction guide, but there is enough information in here to
give you a good idea.

Fr. Tom McGahee

Subscriber: tom_mcgahee-at-sigmais-dot-com Tue Jan 28 23:15:20 1997
Date: Tue, 28 Jan 1997 23:11:20 -0500
From: Thomas McGahee <tom_mcgahee-at-sigmais-dot-com>
To: tesla-at-pupman-dot-com
Subject: Guide to Primaries rev 1.01

    [The following text is in the "ISO-8859-1" character set]
    [Your display is set for the "US-ASCII" character set]
    [Some characters may be displayed incorrectly]

THE GUIDE: TESLA COIL PRIMARIES
Rev. 1.01 January 28, 1996


At its simplest level the primary of a Tesla coil is just a coil
of wire that creates a magnetic field when you pump current through it.

One of our basic goals as Coilers is to find ways to create repeating
pulses of very high current in the Tesla primary. This will produce intense
magnetic fields. If we can get these magnetic fields effectively coupled to
the secondary and operating at the self-resonant frequency of the
secondary, then we can throw some pretty big sparks. So the design of the
primary coil is important.

So here are the challenges before us:
1) Make the primary capable of handling repetitive high currents.
2) Keep the the time it takes for a pulse of high current to rise and fall
as short as possible.
3) Make the primary a tuned resonant circuit.
4) Use the high level primary currents to create intense magnetic fields.
5) Shape the resulting magnetic field so that as much of the energy as
possible is transferred to the secondary.
6) Minimize all sources of power loss.
7) Try to do all this and not implode your wallet.

Why do we want repetitive currents that rise and fall rapidly?
One of the factors affecting the strength of a magnetic
field around a wire carrying current is the current itself. Current is
defined in terms of the number of electrons flowing through a conductor per
second. If we store the electrons up (in a capacitor) over a fair period of
time and then dump them all at once, as fast as we possibly can, then even
though our average current flow may be small, we can still make the peak
value very high! At the instant we dump all those electrons into the
primary we can get really obscene peak values if we can keep the resistance
of the primary really really low. 

OK, so how do we go about making the primary resistance low? When dealing
with high frequency circuits we want to maximize the surface area.
With DC circuits we would want to increase the cross sectional area, but
high frequencies flow mostly over the surface of conductors, instead of
through them as DC would. This is great, because if we use a thin ribbon of
copper that is fairly wide, it can readily carry lots of high frequency
currents with only a fraction of the total material that a solid conductor
would need. 

Flat ribbons have these nasty sharp edges. We don't like sharp endges on
RF conductors, because it promotes energy loss through corona discharge. So
we come up with an almost perfect answer: a conductor with lots of surface
area, little or no wasted cross sectional area, and as non pointy as you
can get (and still be a long conductor). Behold hollow copper tubing! Meets
all the criteria and is readily available to boot. Why copper tubing? Why
not make it out of lead pipe or iron pipe? Because we also need the lowest
possible DC resistance, and the metal should be non-ferromagnetic. If it
was ferromagnetic, then when it produced a magnetic field, the magnetic
field would want to stay close to the iron, because iron conducts
magnetic flux quite well. But we want to get the magnetic field to couple
with the secondary, not hoard it. If you used a ferromagnetic material, the
energy loss due to eddy currents would be Horrendous. Much of the energy
would be expended in heating the pipe. This would be bad news, so we don't
do that. We use copper. Copper is king when it comes to Tesla coils. Some
people plate their copper primary with a coating of silver. This gives
about a 4% increase in conductivity.

For smaller coils the 1/4 inch copper refrigerator tubing is often used.
Many medium to large size Tesla coils employ 3/8 inch copper tubing. Really
big coils use 1/2 inch and larger copper tubing. It is best if you can get
a single piece to form the primary, but if not, then pieces can be soldered
or brazed together. If you put a sleeve on the outside to make a splice,
then you might have a corona problem later. An alternative is to stick a
short section of thinner copper tubing partially inside both halves and
then solder. Use fine sand paper to remove any spots that are bumpy and
might cause corona problems. When bending copper tubing, you want to avoid
having the inside of the bend crimp or buckle. One method is to make sure
the inside bend is firmly pressing up against something and to only bend it
a small amount at a time.

We'll cover the topic of the shape of the primary later in this article,
but we need to mention inductive reactance now. Whenever you have a coil,
it builds up a magnetic field when you pass current through it. Because of
the adjacent turns in the coil, self-induction exists in the coil. That is
the tendency of the coil to oppose changes in current flow. A pulse would
normally cause a coil to have an increasing magnetic field. This magnetic
field will produce a current in the adjacent windings that will tend to
oppose increases in current. Then, when the magnetic field collapses, it
will produce a current that will add to the decaying current such that it
tends to prolong the current flow insteads of letting it decay.
This tendency to oppose changes in current (it's called inductive
reactance) is a function of the inductance of the coil and the Effective
Frequency of the applied Pulse. I say the effective frequency, because a
pulse will always appear to a coil as if it had a Frequency higher than its
repetition rate. A pure sine wave will have no pulses, and its frequency
and repetition rate will be the same.

Oh Great! So we are supposed to keep the resistance of the primary coil as
low as possible, and now we find that it's got this stupid inductive
reactance thing that makes it act like a dumb resistor. Sheesh! Now what?
How do you  cancel out the effects of inductance? It's really fairly
simple. You put the coil in series with a capacitor that has capacitive
reactance that is equal to the coil's inductive reactance at the effective
frequency of the pulse. This works because inductive reactance and
capacitive reactance are 180 out of phase with one another.

For any given inductor/capacitor combination, there is ONE frequency where
The two reactances are EQUAL in Reactance value and opposite in phase. This
frequency is the resonant rrequency. (It's the same frequency that we
ultimately want to also equal the self-resonant frequency of the secondary
coil.

Hey, what if we use the same exact capacitor that we used for energy
storage to act as this capacitive reactance thingie?? It happens to be
exactly the idea that Tesla had. (You are getting SO smart!)

By the way, do you see how all the parts constantly interact? The
Transformer and Capacitor are ALSO supposed to be matched for Maximum
Effeciency, but at the frequency of the AC line, which is 50 or 60Hz in
most cases.

Most of the really spectacular things that a Tesla coil can do occur only
when the tuning of the
primary circuit matches the tuning of the natural self-resonant frequency
of
the entire secondary circuit (That includes the toroid, if any is used). 

Consider coupling to be the way the primary's magnetic field interacts with

the secondary. If coupling is too loose, the coil is inefficient and wimpy.
If coupling is too tight, then you may over-stress the primary and have
voltage breakdown occuring along the secondary or between the primary and
the secondary coil. In olden days there was much emphasis on tight
coupling, because many coil builders erroneously thought that a Tesla coil
worked mainly by transformer action. Not so. A Tesla coil is a
transformer that is tuned to resonance. The resonance is just as important
as the magnetic effects. Even more so if you are talking about a Tesla
Magnifier!

The shape of the primary affects its inductance and its coupling to the
secondary. The inductance affects the resonant frequency of the primary
circuit, and the coupling affects the overall energy transfer from the
primary circuit to the secondary circuit.

Many of the early Tesla coils used a rising helical primary that was
closely coupled to the secondary. Flashover from secondary to primary was
common with this design, and many a secondary coil was destroyed because of
the sparks destroying the insulation. When coupling is too close, the
windings of the secondary get over-stressed and the secondary circuit can
experience breakdown between windings, or sparking between the primary and
the secondary. Some people try to compensate for this by cramming as much
solid or liquid insulation as they can between the primary and the
secondary. That works somewhat, but it has its limits and its limitations.
Eventually almost any insulation can be broken down, even oil insulation. 


The flat pancake coil and the saucer shaped primary are the most popular
primaries among serious coilers, but there are specialty coils that still
use the rising helical primary.

Spiral coils have less of an over-coupling problem. The usual arrangement
is what is sometimes called an
Archimedes or Archimedian Spiral (because the Greek mathemetician
Archimedes was the one who formulated its characteristics). In this kind of
spiral the distance between adjacent turns is kept constant. If the spiral
is kept flat, it is often referred to as a "pancake" coil. If the coil is
not kept flat, but instead each turn of the spiral also includes a rise,
then you have an inverted cone or saucer shaped primary.

The question usually comes up as to how far apart from the outside of the
secondary should the primary begin. Most coilers have the first primary
turn beginning at a distance of one to two inches from the outside of the
secondary coil. This may be adjusted to a larger starting diameter,
especially if the coil requires looser coupling. Coupling can be adjusted
somewhat if the primary and secondary are made moveable with respect to one
another. Some coilers arrange things so that the secondary may be raised
and lowered. Others prefer to move the primary and leave the secondary
fixed. Which method is used is more a matter of convenience and personal
preference than anything else.

Keep in mind, however, that just because most people do something a
particular way does not mean that that is the only way it can be done. Nor
does it necessarily imply that it is always the best way that it can be
done. For example, Malcolm Watts reports that he has made primary coils
whose inside diameters have been less than that of his secondary coils, and
he reports good results from this arrangement. One of his coils had as its
primary a flat pancake spiral made of 3/8" copper tubing with the first
turn having an inner diameter of about 5", and an outer diameter of about
20". The secondary diameter was about 17". He states: "I have noticed no
substantial difference between performance of a helical primary and spiral
primary of the same inductance, using same cap and transformer, and coupled
to the secondary with the same k. I submit that the gap losses (reduced
with higher values of k and/or higher primary surge impedance) determine
system performance if all other components are of good quality and
unchanged in value. I think Bert's SPICE simulations bear this out."

Why are flat spirals and saucer shaped primaries so popular? Because they
work so well. Why do they work so well? Because each creates a magnetic
field that is large and encompasses (ideally) the entire secondary. You can
actually SEE the beautiful shape of this field if you operate a powerful
Tesla coil in the dark. The corona discharge from the primary will engulf
the entire secondary in a kind of inverted parabolic curve when the
coupling and geometry are just right. 

*** Note: there will probably be separate chapters dealing with the nitty
gritty construction details of the primary. This chapter is priamrily
designed to give the reader a good idea as to how the primary works, and
what it does, and how to maximize that in general terms. Specifics of
construction are probably best placed in a separate article dealing with
construction issues.
*** end note

A Tesla coil is a 1/4 wave resonant device. When it is operating properly
the base of the secondary has a low voltage and a high current, while the
top of the secondary has a high voltage and a relatively low current. It
may be useful if you think of it in terms of a standing wave: Imagine one
cycle of a sine wave. It reaches its Peak value at 90 degrees. That is a
quarter of a full wave. If you get a Tesla coil to operate at resonance,
you have a standing wave in which the top of the secondary is operating at
this 90 degree point. It will therefore have maximum voltage at that point.

There is a simple experiment you can try with a child's jumprope that will
illustrate standing waves. Have someone hold one end of the rope tightly in
both hands, with their hands held tight against their stomach so that the
rope at that end will be anchored fairly securely. You grab the other end
of the rope tightly in one hand, and begin moving the rope up and down
about one foot. Start off slowly, and then gradually increase the speed at
which you are moving the rope. Keep the up and down distance you move as
constant as you can. Every time your hand moves the rope up and down it
will send a traveling wave down the rope. But you will reach one particular
rate at which you will no longer see a wave traveling down the rope.
Instead you will see that there is a point right at the center of the rope
that appears to be standing relatively still. At a distance halfway from
the center of the rope to where you are you will see that the rope is
whipping up and down pretty good. That is the 1/4 wave point. That is where
the greatest activity is. When a Tesla coil is operating at resonance, it
is operating at its 1/4 wavelength. Notice that you can get the rope to
have standing waves at higher frequencies. Struggle really hard with the
rope and you might be able to get a standing wave with two nodes standing
still, each about one third of the way down the rope. Notice that it
requires a lot more work on your part to maintain a standing wave with two
nodes, and that the amplitude of the wave is not as big as when you only
had one node. In the same way, you can force a Tesla coil to operate at the
wrong frequency, but it is not achieving full resonance. If a Tesla coil is
improperly tuned the applied energy is wasted because returning waves
cancel the original waves.

The key, then is to achieve this 1/4 wave resonant point. And you want to
do it with efficiency. That's where all the little nitty gritty details
come into play. The frequency we tune the primary circuit to should ideally
match the natural self-resonant frequency of the secondary. Of course, in
real life nothing is quite as simple as that. The secondary's natural
self-resonant frequency varies (in real time!)
***** end of main article on primaries. It is still very incomplete.



*** The following discussion of Q will probably eventually be moved to a
dictionary of terms. I pulled it out of the main text so it would not be a
hindrance to the flow of the article on primaries ***

DEFINITION OF Q

The accepted engineering definition of Q is that it is the 
inverse of the dissipation factor of the circuit - in other words, it 
relates to circuit losses in reactive circuits. 

But Q is not used just in that one restrictive sense. Q is often a general
term that is applied to the relative effectiveness of a circuit or circuit
element. Generally the Higher the Q factor, the better or more efficient a
circuit or circuit element is at what it is supposed to be doing. For
example, if a primary is said to be a high Q primary, you can mentally
decode that to say "This is a high Quality primary. It can do what it is
supposed to do very well." If some other coil design was twice as good at
doing the same exact thing, you could say that it had a Q that was twice as
large as the other coil's Q. The same term "Q" always relates to relative
quality, but the exact thing that Q measures can be quite different. For
example, the Q of the spark gap relates to its ability to go from a
non-conducting to a conducting state in a short period of time, conduct
large quantities of current, and then turn off rapidly. Making any one of
these factors better would increase the Q of the spark gap. One might also
refer to the Q factor of just the quenching. 

For a particular Q, say the Q of the secondary, there may be very specific
formulae available that will allow us to quantify the actual value for Q.
Thus you may hear someone say that the Q of a particular secondary coil is
300.

Q, in short, is a statement of relative performance.  It need not be
fettered or limited to reactive components or circuitry except within
engineering (where it has a nice equation and rigid definition). A
thoughtful person can expand the wonderful concept of Q as a time saving
verbal entity to all processes in coiling where one process or component is
of greater merit or efficacy than another without precise formula or rigid
definition.

*** end of Q discussion

*** The following fragment probably belongs in the discussion of the Tesla
secondary, so for now I am removing it from the Guide for the primary. I
could also include it in both sections if the list finds that desireable. I
am trying to write the Guide such that each Module sort of stands on its
own, though they are obviously all linked in one way or another. 

*** Fragment(s) to be moved to Guide on secondary:

(1)
Some people wind the secondary wires further apart. Normally the secondary
is closewound, but there are times when space winding is used to reduce
winding losses (and thus increase Q), by reducing the proximity effects in
the winding. Proximity effects are an extension of the the skin effect. By
immersing adjacent wires in each other's magnetic fields, you get an uneven
distribution of current around the periphery of the wire. In other words,
it results in current bunching. This raises the effective resistance of the
wire. This is most pronounced in single layer windings. In multilayer
windings, only the outside turns suffer this badly. 

Which method of winding you use depends upon what the prevalent factors are
in the circuit that you are building. It is often wrong to simply say "this
is always the best method to use."  All the factors work in concert. The
idea is to increase Q wherever possible, by whatever means available.

(2)
It seems that it would be ideal
to have a wire graded in thickness from bottom to top to match the 
current distribution in the wire. Problem is that most of the 
effective inductance appears below the top of the resonator, thus 
robbing the coil of inductance at the bottom where it is most 
effective (subjected to the greatest loading). There may be a drawback
to a closewound design if this grading of the secondary is done.
It has been noted by Mark Barton with his conical coils that the symmetry
of the winding is effectively disrupted when the conical secondary is
closewound with wire graded in thickness to match the current distribution.
He found that Q was greatly lowered, as evidenced by poorer 
selectivity. He noted that he could tap the primary virtually 
anywhere and still get a significant output.

(3)
Ever notice that sometimes going near an operating Tesla coil can cause it
to change the size and even the nature of the sparks? That is because there
are interesting electrostatic effects also at work. The secondary coil has
a capacitance that must be taken into account. The problem is the
capacitance of the secondary is affected by many physical parameters such
as thickness of insulation, kind of insulation, number of turns, spacing of
turns, exposed surface area on the outside of the coil, height of the coil,
width of the coil, closeness to other objects, and of course, anything you
put on the top of the secondary, such as a toroid or sphere.

There are well-tested formulae available that can help us get a very
accurate value for the self-capacitance of a coil. Some of these formulae
have even been integrated into very nice computer programs, and are
extremely
useful for anyone designing a Tesla coil. But the person experimenting with
Tesla coils must always be aware that we just can't know all the little
effects which are at work. Some of the least suspected influences, (when
acting in concert), can pull a secondary off frequency. Sometimes enough
out of tune that even if totally isolated it will require another half to
full turn of tapping on the primary over or under the supposed tune point
recommended by all our wonderful equations! This is one of the things that
makes coiling so interesting. There are still mysteries out there to be
probed and understood.


*** end of secondary fragment

This is an on-going work in progress.
See any errors? Did I leave something out?
Anything that is now confusing you even more than before?

Fr. Thomas McGahee