TESLA COIL DESIGN
* Carbons Sent to: Esondrmn-at-aol-dot-com
> From: Esondrmn-at-aol-dot-com
> To: tesla-at-grendel.objinc-dot-com
> Subject: Re: TESLA COIL RFI
Please take my tone with a tablespoon of salt! R.Q.
ES> I plan to build and operate the Tesla Coil in my basement.
ES> I have a 1.5 inch copper pipe that goes through the cement
ES> and runs underground (5 to 6 feet deep) about 80 feet out to
ES> my well. This should provide a good, non radiating ground.
Sounds good! As long as the pipe is electrically isolated from
any other electrical equipment. You don't want anything else
grounded to the dedicated RF ground. The dedicated ground is
strictly for: the base wire from the secondary coil, safety gap
center post, the core of the step up xfrmr, the motor housing
of the rotary gap, any protective strike shields or points
intended to draw spark, and the primary coil too if you intend
to operate circuits with one side of the primary grounded.
RQ> I like a nice tight wound magnet wire secondary
ES> I have read many books on Tesla coils, but not nearly
Oh NO! Not the books again! I am going to have to burn some
coiling books! What books do you have? Who wrote them? Were they
the dreaded "radio" guys again? The radio guys with theory, but
short with the "hands on" arcs and sparks? <grin of one who has
been there!, no insult intended, only humor>
ES> I thought the reccomendation was to space the secondary
ES> windings somewhat apart - like winding a string along with
ES> the wire then remove the string.
Wrong, wrong, wrong, STOP WINDING! <grin, (been there!) grin>
ES> This makes sense as it would reduce the secondary
ES> capacitance and increase the Q. Is this not the case?
Oww! Oww! (enough pain already!)
The answer to these questions is emphatically NO!. You want
maximum inductance in the secondary coil, within limits mind you,
but maximum inductance per unit volume within the breakdown
strength of the physical coil design. Do not space wind! Do not
use insulated wire! I have dozens and dozens of coils under my
belt. I build hot and I burn long, long, long. Throw the books
away... These guys with books out, with few exceptions, are
designing from radio theory or they are throwbacks to the coil
design schools of the 1920s and 30s. Modern, practical, coiling
experience shows (without a doubt) that the radio theory and
classical coil guys are wrong. Follow my advice and you will get
a coil that: runs cooler, runs longer, runs more efficiently,
uses a smaller primary capacitor, uses a smaller step up
transformer, and, exceeds the "radio" and "classic" school of
design theory in bottom line spark length and peak secondary
current by 30% at identical or lower input power levels.
From: Richard Quick
To: David Bearrow
Subj: 10KVA Tesla Coil
DB> How did you go about winding your coil? What are the specs?
DB> And how much did it cost?
The first step in winding a coil is to select a coil form. The
coil form should be a low loss material (we are talking RF
losses) like polyethylene, polystyrene, or polypropylene, Lexan,
or Plexiglas (acrylic): but a common material is PVC, which is
high loss. Thin wall tubing is best regardless of material.
Ratios of coil height to width are important. Small coils work
best with aspect ratios (height to width) around 5:1 - 4:1,
larger coils (over 8" diam.) have aspect ratios around 3:1. Now
we are talking about the actual winding length here, so allow an
extra inch or so of coil form on each end. Determine the length
required and cut the ends square.
The form must be sanded smooth of surface imperfections, dried
thoroughly, and if PVC is used, it must be sealed. A good sealer
is polyurethane, another is two part epoxy paint. By sealing the
surface of the PVC before you wind on wire you can negate the
excessive losses in PVC plastic coil forms. If necessary the coil
form may be sanded after the sealer had dried.
The coil should be wound with good quality magnet wire. I use
double Formvar enamel coated magnet wire. Magnet wire gives you
maximum inductance. A coil should have over 900 turns, but not
too much over 1000 turns. There is a little leeway here. Select a
gauge of wire which will allow the aspect ratio and number of
turns to fall within this range.
I plug the ends of the coil form and run a dowel through a center
hole so that it will spin. I set up the wire spool on one end of
a pair of sawhorses, and the coil form on the other end. I wind
the wire on by hand, making sure the windings are tight, smooth,
and even. I use a dab of hot glue or tape to hold the first turns
in place, and make sure to leave a tail of wire at either end.
Once the coil is wound, it is sealed to prevent corona leakage
and breakdown. I use the same sealers mentioned above. Coats of
sealer are applied until there are no ridges and valleys in the
wire. In other words the coats must build up until the wire is
completely imbedded in sealer.
The wound, sealed, coil is capped at both ends with plexiglas
plates glued down with epoxy. I cut circles out of plexi sheet
that is about the same thickness as the coil form. I rough up the
surface around the edges to give the epoxy a bite. One small hole
may be drilled into the bottom end plate to allow the air
pressure to equalize, but under no circumstances should any other
holes be drilled. The wire is never allowed inside the form.
I have numerous coils, my largest coil has specs as follows:
10.5" O.D. thin wall PVC flume duct, the coil form is 34" high.
The coil is wound with #21 magnet wire, 1024 turns, actual
winding length is 32", aspect ratio 3.05:1. The coil is sealed
with eight coats of polyurethane on top of the wire and five
coats under the wire so that the wire is not in contact with the
PVC but is suspended in sealer.
Coils take time more than anything to construct. I suppose the
material cost for my 10" coil is around $60.00. Smaller coils are
cheaper of course. What can be expensive is putting together the
rest of the oscillator components. Things like HV pulse
capacitors, xfrmrs, and power controllers like variacs. Beginners
can usually start with a few old neon sign xfrmrs, make some caps
(homemade salt water caps are very cheap) or buy some used, and
fire a small coil for under $150.00 from start to finish.
Date: 10-13-93 00:46
From: Richard Quick
To: David Bearrow
Subj: 10KVA Tesla Coil
>> The coil should be wound with good quality magnet wire. I use
>> double Formvar enamel coated magnet wire. Magnet wire gives
>> you maximum inductance. A coil should have over 900 turns,
>> but not too much over 1000 turns. There is a little leeway
>> here. Select a gauge of wire which will allow the aspect
>> ratio and number of turns to fall within this range.
DB> Did you calculate this so that the capacitive reactance
DB> equals the inductive reactance? I understood that's what was
DB> so interesting about Tesla's coils.
No. The balancing act that you are referring to occurs in the
primary tank circuit. High voltage pulse discharging capacitance
(capacitive reactance) is made to balance off the heavy primary
coil (inductive reactance). The primary coil is made to very low
resistance; like HEAVY cable, strap, or soft copper water pipe.
The inductive reactance is "canceled" by the capacitive
reactance, and a very low resistance "tank" is formed where heavy
currents can oscillate with low loss. In order to excite the tank
circuit, high voltage feed lines are brought in to charge the
capacitors. Voltage rise in the capacitors (as they charge)
breaks down the main system spark gap, and bang, the tank fires.
Currents of hundreds of amps, with voltages in the thousands of
volts, ring through the Tesla Tank Circuit.
The energy delivered can easily reach peak powers in the megawatt
range. Since the main system spark gap fires hundreds of times
per second, these high peak powers are, for most practical
purposes, continuous. This primary circuit energy is transmitted
through inductive coupling to the Tesla Secondary.
The secondary coil that I described in the quoted post is NOT a
balanced coil. On the contrary, this coil is pushing the extreme
of several design limits in the quest for more efficient power
processing. The secondary coil, wound as I instructed, results in
a very high inductance coil; but it has a significant internal
capacitance and resistance due to the closeness of the windings,
the length of the wire, and the number of turns.
The high inductance makes the secondary effective. The higher the
inductance, the more energy can be absorbed from the primary tank
Resistance and internal capacitance limit the Tesla secondary for
obvious reasons. Current, especially RF current, reacts poorly to
resistance, which gets worse in small diameter wire. Internal
capacitance in a coil also reduces throughput, as the capacitance
in the turns of wire slow the current peak.
Designing a potent Tesla secondary balances the maximum
inductance (a positive factor) against the resistance and
internal capacitance (negative factors). It takes into account
the need for "critical coupling" between the primary and
secondary (for good energy transfer), resistance of the wire,
internal capacitance between turns, and the breakdown voltages of
the construction. The design given is well researched and proven.
> BUT THE COIL IS UNBALANCED!!! YOU SAID SO ABOVE!!!
Yes, it is VERY unbalanced. Yes, the Tesla circuit depends on
balance for maximum efficiency. The high inductance of the
secondary is balanced by the addition of a large (even huge) top
capacitance. Donut shaped dischargers, called toroids, are used
as a large capacitive air terminal. This air terminal capacitance
"unloads" the secondary, and allows for current flow through the
high inductance coil. The secondary coil, as I instructed, will
not function well without an effective discharger; a capacitive
reactance to balance it. Without it the coil will not survive
much input energy and will self destruct. We have just pushed the
limits with modern plastics in 1/4 wave Tesla oscillators.
Look at the dates I wrote this. I stand by every word today.
ES> Several books reccomend a L/D ratio of 2.5 : 1 or 3.0 : 1.
ES> This means a shorter coil with a large diameter, of say 10.0
ES> to 14.0 inches - like the coils that Tesla built and we see
ES> in all the old pictures. Do you find that this is not
Tesla was winding coils 100 years ago. He did not have Formvar
covered magnet wire. He did not have hermetically sealed plastic
coil forms. He did not have polyethylene dielectrics. He did not
have polyurethane sealers. He did not have hot glue, plexiglas,
polymer adhesives, or any other of the extremely high Q con-
struction materials around my shop. If he had these materials, he
would have used them, and the coils he designed and built with
these materials would have looked more like mine than the coils
in the old photos.
Tesla was building coils with the most primitive, low Q materials
imaginable. Mineral oil was the highest Q material available.
Wire was cotton or rubber covered: rubber wire of that era was
loaded with soot and could not under any circumstances be
considered an RF "insulator", it represented an RF impedance at
best; cotton covered wire was even worse. Coil forms were open
air wood forms, the lowest Q material possible for construction
of this critical resonator. It is a true credit to the genius
of Tesla that he was able to get any spark discharge at all.
He designed his coils, especially his large coils, as you
describe; simply because that is the only way to get them to
function with the materials he had available.
ES> You mentioned 5.0 inch diameter - which must be about 3.0
ES> feet high? Thanks, Ed Sonderman
These numbers are off the top of my head, and it is getting late,
try something like this: 5.0 inch diam. plastic coil form, number
22 double Formvar covered magnet wire, coil form cut to say 26
inches in length, actual winding length of wire 24 inches, number
of turns (37 tpi for #22 Formvar * 24 inches) = 888 turns, aspect
ratio of winding 4.8, this coil should come in around 350 kHz,
discharger ~14 inches in diameter * 3 inches high should bring
the coil down to around the 280 kHz - 300 kHz range; primary 14
turns of 3/8 inch soft copper water pipe wound in an inverted
conical section (Archimedes spiral wound) with a rising slope of
30 degrees from the inside turn to outside turn, primary coil
inside turn 7 inches in diameter, primary outside diameter ~24
inches, this primary coil will tune with the above secondary with
.015 uF capacitance in the tank circuit and tapped around 8-10
This coil won't need a rotary gap. With a 15kV 60mA neon and a
static gap I could get 2 1/2 - 3 feet of spark from this setup.
Richard (check my math, It's late, I'm tired) Quick
... If all else fails... Throw another megavolt across it!