Thyratron Switched Tesla coil (THYTC) / Projects / Forums

Forums

4hv.org :: Forums :: Projects

« Previous topic | Next topic »

Thyratron Switched Tesla coil (THYTC)

1 2 3 next

Move Thread

LAN_403

radhoo

Mon Oct 10 2011, 04:01PM

Registered Member #1938 Joined: Sun Jan 25 2009, 12:44PM
Location: Romania
Posts: 701

Even before I got my first Hydrogen Thyratron, I was already anxious to see the pink/purple glow the ionized hydrogen produces. A few weeks later, I got a nice TGI2 400/16 Russian Thyratron (

) so I started one of my longest projects with High Voltage.

As you know a thyratron tube can work as a switch. It's a gas filled triode, with a filament, an anode plate and a grid. Gases like xenon, hydrogen or even mercury vapors are all used in various applications. The grid works like a turn on switch, and starts the ionization inside the thyratron. The tube will remain on, if the plate voltage is kept over a given threshold, so unlike a transistor, the thyratron will not turn off, even if the grid is voltage is removed. To turn the thyratron off, the anode voltage must drop close to zero. Then the ionized gas inside needs to revert to the neutral state in a process that is called "recovery", and the time needed to complete this process is called "recovery time". As you can see a thyratron works similar to a thyristor, but the similarities end here:

Unlike its silicon relatives, a thyratron like the one I got, the TGI2 400/16, can handle pulses of 16KV 400A, impressive numbers to say the least. Currently, hydrogen thyratrons are hard to get, and are usually expensive. For the purpose of this project, hydrogen or even deuterium Thyratrons work best, as the tubes that use these gases have a very quick recovery, unlike mercury vapor thyratrons. So when trying to build a Thyratron switched tesla coil, it has to be a hydrogen thyratron at least.

Even so, my hydrogen TGI2 400/16, only allows frequencies of up to 450Hz, very little for getting any air transformer to resonate. As my good friend vasil indicated, it would be wise to use the setup to drive a pulse transformer instead (using ferromagnetic core).

But since many of you wondered of how a Thyratron Switched Tesla coil would look like, I decided to run a test using an Air transformer, and see the results. Building a device like this requires achieving a few steps:

1) A filament transformer (both for main filament and for the hydrogen reservoir), for the TGI2 400/16 that means 6.3V at 10Amps. I used a rewound mot transformer, see it here:

2) A high voltage supply, delivering 8KV, that is actually a flyback inverter using a Royer oscillator.

3) A Pulse Forming Network, type E, construction details here:

Here's a video with a PFN test:

4) The TGI2 400/16 :

5) Pulse generator for grid control, this is actually a variable pulse duration, frequency and amplitude, PCB design and schematics here:

A video with a pulse generator test, using the neon tube 313CD from Western Electric:

6) An air core transformer with a tesla coil secondary and a few turns of primary, that I had ready from a different project
This device is in fact a relaxation oscillator. I've prepared a long article on this topic here:

In just a few words, the pulse generator provides bursts of 200V on the thyratron grid, that turn the tube on. The capacitors in the pulse forming network discharge via the thyratron in the load. Because of the inductive properties of the PFN, the positive discharge voltage has a tendency to swing negative, turning the thyratron off and ready for the next discharge. The PFN discharge and the Thyratron deionization times give the total recovery time.

Here is a video showing the working Thyratron tube in one of the first tests:

Since I'm among the first to build a device like this (I only saw another single design, I believe it would be nice to have the privilege to name the abbreviation for the Thyratron Switched Tesla coil, a "THYTC" device.

Simplified circuit diagram:

The pulse transformer is replaced by an air core transformer in the video above.

Looking forward to your comments.

Tetris

Tue Oct 11 2011, 02:13AM

Registered Member #4016 Joined: Thu Jul 21 2011, 01:52AM
Location: Gainesville, FL
Posts: 660

So this has got to be one of the "simpler" coils, as far as circuitry goes. :P not as spectacular as a DRSSTC, SSTC, or SGTC, (some VTTCs are neat too), but its still nice for a desktop thing. :D awesomeness.

Dr. ISOTOP

Tue Oct 11 2011, 04:07AM

Registered Member #2919 Joined: Fri Jun 11 2010, 06:30PM
Location: Cambridge, MA
Posts: 652

Holy shizballs, you actually built one of these!
You should see whether it scales up, using those big ceramic thyratrons.

radhoo

Tue Oct 11 2011, 08:51AM

Registered Member #1938 Joined: Sun Jan 25 2009, 12:44PM
Location: Romania
Posts: 701

HighVoltageChick wrote ...

So this has got to be one of the "simpler" coils, as far as circuitry goes. :P not as spectacular as a DRSSTC, SSTC, or SGTC, (some VTTCs are neat too), but its still nice for a desktop thing. :D awesomeness.

Well, the video shows only a "first light" test. I'll be changing that to something more spectacular.

bwang wrote ...

Holy shizballs, you actually built one of these!
You should see whether it scales up, using those big ceramic thyratrons.

Well I had to, you know , having a device that uses hydrogen to transport electric charge is rather intriguing. And regarding the applications I also need to see how a pulse transformer goes, since Tesla coils are not that well suited for this oscillator.
Nevertheless I will be trying to push it to its limits soon. Currently I'm feeding it with only 100W of power, as provided by the high voltage supply I'm using. That's too little. I need to get closer to those 16KV 400Amps pulses.
If I do that, I can start thinking of ceramic thyratrons as a next step.

All in one this was rather an exotic experiment to replace a spark gap with a thyratron and to manage to control the thyratron on and off at the best frequency possible. The PFN and the pulse generator are doing an excellent job here.

teravolt

Tue Oct 11 2011, 01:43PM

Registered Member #195 Joined: Fri Feb 17 2006, 08:27PM
Location: Berkeley, ca.
Posts: 1111

hi radhoo, I gues you know by now you will need a floating filiment transformer equal to your anode voltage. your pfn's pulse width should be equal or under half of a cycle of your resonant frequency of your secondary. the thyratron is only good for about 1khz and rise times suck so the out put pulse of your pfn will have to be around 1-3us hope you can do it. if you pulsed a resonoant primary with the same frequency as your secondary you may make the first pulse tesla

Proud Mary

Tue Oct 11 2011, 08:23PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

radhoo wrote ...

A filament transformer[/b] (both for main filament and for the hydrogen reservoir), for the TGI2 400/16 that means 6.3V at 10Amps

TGi2 400/16 has an indirectly heated cathode, but no auxiliary reservoir heater.

The English data sheet appears to contain errors. For example,

Steepness of the pulse anode current, A /ms 800

which must surely be A/Î¼s for a hydrogen thyratron.

Good to see some original work!

PS I noticed in a technical bulletin from Lianozovo that TGi2 400/16 is no longer manufactured, and is being replaced by the new metal-ceramic thyratron TTI-3-500/16, in order to extend the working life of ageing P-37 radars built between 1980 - 1992.

radhoo

Tue Oct 11 2011, 10:26PM

Registered Member #1938 Joined: Sun Jan 25 2009, 12:44PM
Location: Romania
Posts: 701

Proud Mary wrote ...

TGi2 400/16 has an indirectly heated cathode, but no auxiliary reservoir heater.

The picture below shows the hydrogen reservoir (actually containing a metallic hydride, that generates hydrogen when heated. The reaction is reversible, so the hydrogen will re-form the hydride when the tube cools down).
The reservoir has a central filament heater, that is connected to the main filament via an inductor placed at the bottom of the tube. The picture shows it glowing, after a few minutes of use. I might take a few more pics tomorrow, to better illustrate the setup.

Proud Mary

Wed Oct 12 2011, 07:53AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

You're right about that - it's the datasheet circuit symbol that is incomplete, and does not show all internal elements.

radhoo

Wed Oct 12 2011, 06:19PM

Registered Member #1938 Joined: Sun Jan 25 2009, 12:44PM
Location: Romania
Posts: 701

I've just built a socket for this tube:

It looks better in vertical position, and improves heat dissipation.

Proud Mary

Wed Oct 12 2011, 10:19PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

A further thought on reservoir heaters.

As you know, larger thyratrons often have a separately powered reservoir heater in order to adjust the gas pressure by controlling the heater current. E2V supply switchable resistance boxes with some of their metal-ceramic thyratrons in order to optimise the gas pressure in individual tubes for minimum jitter and anode delay timing drift.

Now with your thyratron, you cannot heat the gas reservoir separately from the cathode heater, but you might be able to supply extra heat to the reservoir by means of laser pulses from outside the glass envelope - which I have never seen done before.

PS My own recent experience with hydrogen thyratrons is in single shot firing at 4.999 kV, so I haven't had the same heat build-up limitations you will have to manage with your application.

1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob