Assuming you mean winding them on the same toroid: arc-over and the coupling will be too strong. You want poor coupling, preferably with a way to adjust/tune it.
Also, winding on a tube and then bending it into a toroid will try to stretch the windings on the outside of the toroid and compress them on the inside...this may be impractical.
Tesla coils are very prone to arcover causing breakdown. You therefore don't want them toroid shaped. You could, but a linear coil would be much more reliable. I'm also not seeing any upside to a toroidal form. It's harder to wind, and since you don't want any metal in the core you may as well use paper tube, wood, plastic etc which are all more available in linear form.
Even with. The Triad CST206-3A for example rings terribly. The waveform you get is precisely what's explained here, for the same (mechanical analog) reasons:
formatting link
I think they use a powdered iron core or something, which exacerbates the capacitive loading, ensuring the lowest possible resonant frequency for the amount of wire used.
A typical ferrite core (high mu, so pretty lossy in the MHz) exhibits three or four resonances of notable magnitude, and then is pretty much a dumb resistor above there. The modes can be shunted by winding the primary in a particular fashion -- assuming you have the freedom to do so, of course.
A toroid without a ferrous core is magnetically interesting.
--
John Larkin Highland Technology, Inc
lunatic fringe electronics
If you hold a bat too tight the ball will not go far.
If your coupling factor is too high you induced voltage will not resonate
high. Mine your coupling factor. A toroid could have a coupling factor as
high as 0.7 where
If you thought it's as simple as doing a 1/4 wave antenna, wrapped around, you're missing something!
What's a 1/4 wave solenoid got? Magnetic field at one end (at the grounded base: voltage node, current antinode), electric field at the other (voltage antinode, current node).
Wrap it around, and you've just hooked a node to an antinode and completely canceled out that mode!
So we know the lowest resonant mode cannot be 1/4 wave. Similarly, it can't be 1/2 wave if we short the ends together. (The connection can be left open to allow the 1/2 wave mode, but now you've just bent around a 'two headed' TC (two top loads on one floating secondary) into a horseshoe for no reason, and now it's inevitably arcing to itself...)
The first resonant mode of a helicotoroidal winding is a full wavelength. Because, 'duh', the wave repeats after exactly one full trip around the loop. :)
The fields are in quadrature, so that one corner is a voltage node, then 90 degrees down is a current node, and so on, alternating around the circle.
The next mode (two waves) alternates nodes every 45 degrees, and so on. Higher modes aren't useful to a TC (less distance between peaks!), so you'd want to tune away from those (or block them with clever winding design, or dampen them with snubbers).
So what you'd actually get, is high voltage between the two sides of the torus -- not at the wire ends (which should be shorted together forming a complete loop, to enforce the continuity condition), or in the middle.
It's something like one of those two-headed Tesla coils, with a pair of "top" loads and the primary in the middle -- the secondary is driven as a
1/2 lambda resonator in this condition (current nodes at the ends, voltage node in the middle). Suppose you ran two of these side-by-side, with their primaries wired together in phase, then shorted the top loads together (because there's no voltage between adjacent tops -- it's all in phase, and presumably, tuned the same, and achieving the same voltages..). Bend the coil former into a smoother shape, and you've got the helicotoroidal secondary!
On the upside, you won't need a "top load", because the bulk of the winding itself holds all the voltage -- it's not a sharp cylinder, so isn't especially prone to breakout. You will probably want to tap the winding to add breakout points (and you can add top loads to these if you like), just so the arcs don't burn up the coil.
Slit down the middle (a "C" shaped slitted cylinder, revolved into a torus), presumably?
Then you just get transmission line and transformer effects, and no fun toroidal resonances. Useful? Idunno. Probably not. Trying to get any voltage on it, is obviously a large design challenge...
Why not a *full* toroid? Connect the ends together? (So it acts like a "w hispering gallery" optical nanoresonator! Giant optical nanodevices, sell them for physics lectures.) The closed ring would be equivalent to two TC secondaries in parallel. Drive it at a mode where two locations on the per imeter will become antinodes and exhibit an immense voltage diff. Or, driv e it at a higher mode where you get four HV antinodes rather than two. Or six, etc.
Also, for loose coupling, take advantage of the fact that Tesla Coils can b e driven capacitively, with a driven plate or sphere held near the secondar y. (Or, for an ungrounded half-wave device, run it with a pair of driven p lates placed at two spots near the center.) A primary coil does not give loose coupling, and in that case the Q of the secondary won't be extremely high Q of the secondary considered alone. (Low Q gives low resonant-rise v olts.)
If I understand correctly, the HV taps would be 90 degrees around the circumference from either side of the primary, and there would be a low voltage node (null) at 180 degrees.
Does the chirality of the secondary winding relative to the primary matter in this case? IOW one side of the secondary CW and the other CCW.
Could this configuration be operated as a phase splitting transformer by grounding the null point?
I presently use a "Slayer" type driver for exps since it automatically keeps the TC in resonance.
Not quite, the taps would be 180 degrees apart, and 90 degrees away from those are the voltage nodes (voltage nulls, where the primary is placed).
Yes, because winding it backwards (left vs. right for the two halves) but carrying the same current would have an opposing magnetic field. (You could cross-connect it, so it looks like two 'Z's rotated and superimposed, and that would support the single-wave resonance: but that again has the high voltage insulation problem.)
Good catch, that's something my "put two coils together" explanation missed!
Now, if it is connected that way (a left-handed half and a right-handed half, joined together), you still get a 1-wave resonance, but it's not available anywhere on the winding, it's not rotationally symmetric. The opposing magnetic fields, where the left- and right-handed windings join, enforces zero current at those nodes, thus they must be voltage peaks instead. Placing the primary 90 degrees from those points will operate as normal.
The difference is that, a perfectly symmetrical secondary (an all-right or all-left wound, helical loop, bent into a torus, rotationally symmetric), will not care where the primary is, and the voltage peaks will always be at
90 degrees to the primary winding. That's not very important, but it becomes important if you don't wind it all the same way. :)
Yes. The two HV nodes are opposite (180 degree) phase, electrically as well as physically (around the torus).
Eugh, a circuit that has undefined peak collector voltage...
It's a circuit that works better with high coupling and high ratio, so you're not driving a double tuned resonant transformer, but closer to a single-tuned transformer with a high ratio. High coupling keeps the collector ringdown (peak voltage and resonant frequency) under control, while the high ratio at least allows high output voltage to be produced.
It's also a self-excited oscillator with no quenching or bias protection, so it's quite capable of melting transistors when it stops working.
It also requires a transformer ratio (including Q multiplication) smaller than transistor hFE at the operating frequency (several times smaller if you want efficient operation).
You can do it "balanced" with two primaries (180 deg apart) or just one.
Sure, use two primaries in opposite phase. Or short it out. :P
Perhaps (the field outside an infinite solenoid, and I suppose possibly around a toroidial winding, dunno), but whether it does is immaterial if there's no induced E as well: there can be no propagating vector (for an antenna) or induced power (for a transformer) without it.
There is most certainly no gap in our knowledge of the electromagnetic field. It's the second discovered physical law, and the first discovered with apparently exact precision!
(E&M takes no corrections from relativity, whereas the first-discovered (Newtonian gravity) does. Indeed, the Lorentz transform arises from electromagnetic relativity! At very high energy densities, there is a correction from quantum mechanics -- as the very vacuum of space itself breaks down, producing matter (pair production), or from general relativity if an awful lot of it is crammed into an impossibly small space somehow (kugelblitz). As far as I know, these do not affect the "small signal" properties of E&M, so that Maxwell's laws, expressed in whatever form (the classic 4 vector calculus relations, or just one tensor written into relativity itself), are always the same proportions.)
This is a problem in any coupled impedance system, indeed a required aspect. Systems that are nonreciprocal (signals or power only flow one way) are the exception, usually by taking advantage of wave mechanics, or requiring power consumption (amplifiers).
Isolators are usually based on Faraday rotation through a magnetized substance. Usually Tb glass or garnet for optics, and ferrite or YIG for microwaves. It can't be done in a wire (a transmission line is one-dimensional, there's no degrees of freedom to pull the same trick with), making it challenging at low frequencies (where, fortunately, amplifiers are cheaper, but a passive varactor circuit has been invented).
For a TC, it's not very meaningful, as you're not matching impedance to a source: you're alternately opening and shorting the primary (which happens to have some initial voltage or current when this happens). The impedance of a spark is very low (ohms), and the impedance of an open is very high (gigs). You mostly want to target an impedance somewhere inbetween, so as to keep efficiency high (relatively small voltage drop in the spark) and voltage and current manageable (so the capacitor and inductor aren't ridiculous ratios, and the charging input is reasonable, say ~10kV from an NST).
It's more meaningful for any CW (constant or gated) exciter, whether SS or VT. In that case, you have to match to the voltage and current specs of the amplifier.
Towards a possible practical implementation of that idea.. Toroid form with primary, at reasonable distance from this toroid centroid, wind secondary. Obviously, half of this winding will be near the center, and the other half will be outside the center primary winding.
These two papers discuss Weird Science: using resonant toroid coils as transmitters and receivers. They're looking for non-radio signals (or, something outside of Maxwell's Eq'ns.)
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.