Re: 3/4 wavelength secondaries
Tesla List wrote:
> >From rwstephens-at-ptbo.igs-dot-netWed Jul 24 21:55:16 1996
> Date: Wed, 24 Jul 1996 18:35:50 -0500
> From: "Robert W. Stephens" <rwstephens-at-ptbo.igs-dot-net>
> To: tesla-at-pupman-dot-com
> Subject: re: 3/4 wavelength secondaries
> >>From sgreiner-at-wwnet-dot-comTue Jul 23 21:59:58 1996
> >Date: Tue, 23 Jul 1996 21:32:37 -0700
> >From: Skip Greiner <sgreiner-at-wwnet-dot-com>
> >To: tesla list <tesla-at-pupman-dot-com>
> >Subject: re: 3/4 wavelength secondaries
> Skip Greiner wrote,
> >Thanks to Richard Hull, Robert W. Stevens, and John H. Couture for their
> >answers to my query re the subject.
> You're welcome. RWS.
Robert, thank you again for your very in depyh reply. However I would like to pose a few
remarks/questions regarding your answer.
> >Robert, I have a coil in which the wire length is "fairly" close to the
> >1/4 wave resonant frequency. This coil will put out discharges exceeding
> >36" from a 4.5" insulator on top with only a 1/4-20 screw for the top
> The relationship between streamer length and terminal voltage is so
> non-linear that the strike distance measured to a grounded target
> cannot be used to accurately guesstimate voltage. As I pointed out
> earlier, once corona is formed in the air, the voltage becomes
> clamped by the ionized atmosphere and will not rise appreciably with further
> increases in the application of power. The length of the streamer will however
> grow from this point on, in a direct relationship with increasing current. Your
> 1/4-20 terminal is probably not yielding more that 60 or 70 kilovolts
> (if that high) at the corona clamping level, but your system can push enough
> current behind the leading edge of the output pulse to push the streamer to
> 36 inches.
> The closer I have been able to get to the 1/4 wavelength match between wirelength and
resonant frequency the longer the discharges become. (this assumes the same input power)
By the way I have wound about 40 different secondaries and each has outperformed the
> > there is also a great amount of corona from the top several
> >turns of the secondary.
> Without any electrostatic field shaping/control at the top end of
> your secondary (which is provided as a side benefit of a toroidal
> topload), corona from the top winding is a common, coil damaging problem.
I fully expected this. mentioned only to indicate this side effect.
> >I have tried to design and wind both 1/4 and 3/4 wave coils to be run
> >with none or very little terminal capacitance with the idea that given
> >the same energy delivered to the terminal capacitance, and using the
> >energy equal to 1/2 cv^2...then decreasing the terminal cap should lead
> >to increasing the voltage.
> If it were possible to put a large radius, smooth top terminal on top of a
> secondary that by some magic not allowed by the laws of physics, did not also
> contribute a substantial amount of isotropic derived capacitance then maximum
> voltage could be achieved. The metal shape would provide the corona holdoff
> we desire (avoiding early corona voltage clamping effect) whil'st not adding the
> capacitive component which causes a so equipped resonator to max out
> considerably short of 90 electrical degrees.
> If you look at any secondary with a toploading capacitive
> terminal from an electrical perspective, you will see that the
> resonator is less than a full 90 electrical degrees long, and therefore cannot
> achieve maximum voltage swing. In the real hardware world though, such
> toploaded, foreshortened coils gain so much benefit from the corona holdoff
> feature provided by a large toroid, that they still provide better
> performance than those full length resonators without toploads.
>I fully agree. I wish to strive for 90 degrees to see what happens. Has anyone shown
that the benefits of a topload outweigh not getting to 90 degrees? I have been unable to
find anything in the literature.
> I believe that the capacitance represented in the toroid also helps
> because it becomes a local storage facility to provide energy behind
> the discharge streamer where and when we need it most, at the moment of
> breakout. For example, consider a 100 pF toroid charged to 3
> megavolts, (this is entirely within the range attainable with a large
> homebuilt coil). 1/2CV squared yields 450 Joules! If discharged in the first
> 180 degrees of a 100 KHz damped sinewave, this stored energy from
> just the toroid represents a pulse power of 90 megawatts! Only with the storage and
> holdoff attributes of a large topload could we reap the benefits of resonant rise
> to build our power to such high levels. Resonance makes a coil
> function, resonant rise makes it function better! Techniques to hold
> off the formation of energy draining corona are the route to maximizing resonant
I do not understand the physics of the coil. If we have AC in the secondary, how can we
store energy in a toroid when the input to the toroid is reversing 200,000 times per
second. In actual practice it appears that regardless of what we put on top of the coil
corona is formed at some point. After the corona forms additional energy in the coil
will flow to that breakdown point and enhance the discharge. Maximizing the energy in
the coil and being able to deliver it to the top terminal at maximum pressure will then
maximize the discharge.
> Richard Hull advises to employ a topload so large that you are
> just able to break away from it in the upper reaches of your system's
> power capability. He's right. This is another fine design balancing
> act however in and of itself! A system that takes 8 KW to form corona is going
> to be an outstanding performer when running at its optimum design power
> level of 10 to 12 kilowatts, but before producing corona it will be beating itself to
> an early death when there are just 7 kilowatts being put into it, no
> corona yet, and nowhere for this power to go. An analogy is a
> powerful radio transmitter operated with the antenna disconnected.
> It doesn't take long to let the smoke out of a 10 KW broadcast
> transmitter under such conditions if there are no fast acting safety shutdown
> circuits in place. Ditto for Tesla coils. There is much more smoke
> stored in the really big systems, and this smoke is expensive!
> My intention at this time is to try to determine the maximum discharge that can be
obtained from a specific power input. (No, I am not suggesting any kind of a contest
here). If there is some relationship between power input and discharge length, the
principle, if it exists, might apply to any size coil. I like to get all of the smoke
into the discharge.
> > Although its been a couple of years since I
> >read Tesla's CSN, I believe he thought he was winding 1/4 wave coils and,
> >if memory serves, he used very small terminal caps. I understand about
> >corona and the fact that a toroid should build up the energy before
> >breakout, but is that really how a TC works? I should not be able to get
> >36" discharges from a 1/4-20 bolt if that assumption is true.
> I see nothing unexpected from the 36 inch discharge size you show in this
> example from merely a 1/4-20 bolt as a top terminal. A 120 MA neon
> transformer represents a lot of power! I'll bet you real money that if you add a
> smooth radius topload to your system, and add primary turns to bring
> it back into tune, that you will add at least 25% to your streamer
> length with no other changes. That is a conservative estimate.
> I have a 15KV-at-60MA powered coil that consistently exceeds 40 inches
> with a 3 inch by 24 inch toroid perched on a 4.25 inch dia. coil
> wound 24 inches long with #26 wire. My system RF capacitor is a little on the
> small size for this size neon at just 0.008 mfd. With a considerably more massive
> topload I have measured 52 inchers to a ground target. I'm sure this
> would increase further with a bit of a capacitor size increase, as
> the cap size (and the voltage it is charged to) sets the system energy.
> I haven't tried to resonate this system without a topload. Maybe
This sounds pretty good but.... do you know if longer discharges are possible with a
slightly different winding? I have a 12"x36" coil that put out discharges of 48" with
720 va input and I believe that 5' to 6' discharges should be possible.
> >In any event I do not believe a 1/4 or 3/4 wave coil can be wound in any
> >"reasonable" configuration. In fact the resonant frequency of 1/4 wave
> >coils is always about 40% higher than the 1/4 wave wire length and a
> >toroid must be used to bring in the frequency even taking into account
> >the increase in cself due to ionization.
> That statement is counter to my experience. I have found that when a
> given freespace 1/4 wave wirelength is made into a coil, the resulting coil
> operates at a longer wavelength (lower frequency) than its equivalent stretched
> out wavelength. For example, I have wound 753 feet of #22 enamelled wire
> onto a 4.1875 inch diameter form to a length of 22.675 inches. 753
> feet of wire represents a freespace 1/4 wavelength at 327.6 kHz. The
> actual measured Fres of this coil with no topload is 317.6 kHz. By
> forming this length of wire into a coil the frequency dropped 3.1%.
> Another way of looking at this is that by winding the coil into a
> spiral, the velocity of propagation down the wire has slowed by 3.1%.
I plan to duplicate your coil and I appreciate your input. This is by far closer than
anything I have been able to do.
> > The resonant frequency of a 3/4
> >wave coil is always lower than the 3/4 wave wire length so the 3/4 wave
> >coil cannot be tuned at all. I have tried configurations from 5" to 48"
> >in diameter and heights up to 190"(absurd, but the program allows it).
> >I have actually built coils up to 48"x48" just to prove the program and
> >the designs usually come in close to 10%.
> Oh, I see now what you meant all along. You're stipulating that you start out
> with 1/4 wavelength of wire measured off to the freespace wavelength
> at some frequency. Then, no matter what shape you wind it into, the
> resulting resonator no longer represents a 1/4 wavelength resonator
> at that given frequency. You didn't come out and say this in your
> original post. Yes, you're absolutely right, and from what I hear it seems that
> this particular anomaly is difficult to predict in mathematical models.
> This is why, even after the computer CAD program tells you final
> parameters to ten decimal places, you still have to build a Tesla
> coil in hardware, and play around with it a bit to make it work. I
> don't even bother with computer models. Several elementary
> electrical formulas on the old scientific calculator and away I go to
> the shop. For me, a lathe, or a drillpress are more essential in Tesla coil design
> than a computer! This is not to say that computerized TC design programs
> won't help save time by calculating tables quickly. Ultimately, it
> is the computer in your head, and your level of understanding which
> will determine your level of achievement with a coil project.
I fully agree but my last coil which is 14"x 30" came within a couple % of the predicted
values for free resonance. The ionization capacitance is obviously much more difficult
> >I would certainly appreciate any comments to this and I would especially
> >like to hear from anyone who has built a 1/4 or 3/4 wave coil that comes
> >in close to the correct frequency.
> I cited an example above in 1/4 wavelength resonance that was just
> 3.1% off. Now that I think I understand your definition of 'correct
> frequency', I suspect that this may not actually be possible. As
> soon as you wind a straight wire into a coil you increase its total
> inductance, and that combined with self capacitance will lower the
> self resonant frequency of that piece of wire. I've learned to live
> with this minor guesswork factor in determining final secondary size, and my
> coils always work just fine, even if they don't work at the 'correct' frequency.
> This seems to be a bit of an art, the feel for which one only gets
> after winding many, many coils. I've read Tesla could merely look at a coil
> and tell you right on what frequency it would resonate at. Tesla had
> his own CAD program (Croatian Aided Design), quite advanced for the
> 'computers' available in his day.
> >Thanks for listening
> Happy Coiling!, regardless of freespace wirelength, rwstephens