Author Topic: High frequency transformer gigahertz - tesla coil  (Read 4382 times)

0 Members and 1 Guest are viewing this topic.

Offline adrasTopic starter

  • Regular Contributor
  • *
  • Posts: 55
  • Country: de
High frequency transformer gigahertz - tesla coil
« on: September 22, 2023, 02:01:09 am »
Hey guys,
crazy I got here, but I think my conversation with chatGPT is at a point where the real experts come into play. That's you guys, I love you (Sorry I'm a silent reader here)

Now I'm theorizing with tesla coils lately, and without starting some new crazy theories let's just tackle one of the issues I'm currently facing. I want frequency, I recently saw some circuits to power tesla coils which operate in the low megahertz range. That's already pretty cool, but ideally I want even more. Now for a hobbyist with a 100$ budget switching currents at frequencies of 1ghz seems to be challenging. So I got creative.

What I would like to have is a "defined spark gap". Similar to the way you create a neon tube, you take a vacuum tube but fill it with air instead. Air with a certain pressure, humidity etc. Now when you apply a voltage you should get a spark consistently at the same voltage. Unless the outside temperature changes too much.

Now when I look at a low current high voltage generator which is attached to a spark gap, I see that I get a spark, let's say every second. If make the gap shorter, due to the same capacitance on the input side it will charge quicker, and I will see, let's say 2 sparks per second. So given the distance and the capacitance and the amount of power I generate I can adjust the frequency.
And since I hopefully created a "defined spark gap" this should be able be adjusted for the gigaherz range. (At least that's how I understand the crucial functionality of the spark gap in the simple nitrogen laser). As it is a spark gap it can also handle some current, at least as long as the electrodes can handle the heat

What do you guys think?

Online T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21961
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: High frequency transformer gigahertz - tesla coil
« Reply #1 on: September 22, 2023, 04:42:24 am »
You're not going to find much of any experience in GPT here, and I suspect, relatively unbelievable statements as a result.  That is, this will be an area of language, a combination of topics, that it will, as usual, do its best to sound plausible, but make basic inconsistencies along the way.  Again, it doesn't know what words actually are, just that they're more likely to be strung along one way or another; there is no underlying meaning of words, so they don't in general get used consistently, it's mere probability that some words get used consistently in a given (well trained) context.

You might take a different angle, and ask about basic sources relating to such a topic.  There's a chance it can tell you about J.C. Bose,
which would be an excellent jumping-off point (and by "off", I mean off GPT and into some good old fashioned reference-searching and reading; mind that, while you might ask GPT for references, they need not exist; you can, however, do the inverse operation effectively AFAIK, i.e., provide the citation information and ask it to format it in whatever way you require e.g. Chicago, MLA, APA, BibTeX, etc.).

As for the basic idea, you will find:
- Spark gaps have a limiting operating time.  They are not instantaneous (though they can be impressively fast!).
- The de/ionization time depends on gas composition and pressure.
- Mind that composition of a mixture may change due to ionization and heating, even in a sealed enclosure.  Consider the Moore lamp:

For example, air being a mix of N2/O2, accumulates O3 and NOx with discharge.  CO2 breaks down to CO, C and O2 (and I suppose some O3 transiently, but that will combine with CO on its own I think), hence the issues above.  So, characteristics can change over time.  Mostly you need pure gasses or unreactive mixtures, which greatly limits options (especially spectrally, if you were making glow lamps), and preferably inert gasses, limiting options further.  (Typical glow discharge switching tubes use xenon or mercury vapor.)

As it happens, ionization time is related to ionic mass, greatly hindering the use of Xe, Hg, etc. for high performance applications -- typical tubes take microseconds to ionize (even at high pressure -- this sets the performance limit of Xe flash lamps for example), and milliseconds to deionize.  When performance is required, a hydrogen or deuterium thyratron is the preferred choice -- these are usable for radar modulator, transient generator and scientific applications, where high peak power and short risetime are required.

A typical in-air switch can close in a fraction of a nanosecond, and take some ns to deionize. So that's not bad, and in fact is what got Bose a lot of research value.

- When period shortens comparable to deionization time, the spark gap no longer looks like a spark gap, but the on/off ratio becomes much smaller, and it has more of a negative-resistance characteristic instead.  Anyway, at such frequencies, you're working with RF structures, indeed transmission line structures,

- A Tesla coil at very high frequencies, isn't very interesting anymore.  It's useful or distinctive at low frequencies, because the impedance ratio can be quite large: a ratio of hundreds or thousands, on top of the transformer ratio, makes quite high voltages possible.  As frequency goes up, the possible ratio decreases -- or the same geometry is used but at vastly reduced scale, suffering greatly increased losses and electrical breakdown (you want how much voltage in a 1cm space?).  Its concept basically reduces to a standard combination of resonators and coupling networks, optimizing to achieve breakdown field strength at some point in the system -- which might in this case be a pinched waveguide, coupling ring, or something like that.

You certainly can't get any fraction of the visual appeal of a Tesla coil, because the streamers would be multiple wavelengths long.  Ionization only occurs where voltage is highest, and plasma is heated where current is highest; both occur in rapid succession so the generated plasma is very smooth.  This is easy to demonstrate, actually, with home equipment: stand up a lit match vertically in the microwave, putting a heat-resistant glass bowl above it; the flame provides some initial ionization, and a plasma ball quickly ignites.  It doesn't have any interesting streamers or anything because it simply absorbs power from the field, and glows wherever and however it does; it's very hot so it does rise convectively (without a cover, you get more of a Jacob's Ladder effect where the flame/splint ignites a new discharge when the previous one fizzles out, preferably when a 1/4 wave antenna is used as the ignition source).

Related: the general methods themselves aren't really new (ICP and waveguide plasma generators have been around for some decades), but I know one of the guys doing this,
they have a pretty neat setup.  An ignition coil provides initial ionization -- they don't have nearly enough electric field to cause breakdown in air; which also allows the waveguide to be tuned to the plasma impedance more easily.

Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
The following users thanked this post: daqq

Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo