Kaizer VTTC I

WARNING!: Working with electricity is dangerous, all information found on my site is for educational purpose and I accept no responsibility for others actions using the information found on this site.

Read this document about safety! http://www.pupman.com/safety.htm

Introduction

A VTTC is a Vacuum Tube Tesla Coil and it uses vacuum tubes / valves for the oscillator that is self biasing from a grid leak circuit.

I chose to build a VTTC because the operation and components were simple and few. It did turn out to be hard to find all the required components at a reasonable price in Denmark.

Considerations

After spending many evenings reading about vacuum tube Tesla coils and comparing tubes, I chose to build a Tesla coil utilizing two 811A transmitter tubes. These tubes have been used by others with great success, they are fairly cheap and have a moderate output power for a tabletop size Tesla coil.

Having studied and read about all the VTTCs I could find on the internet, it became clear to me that there are no pre written text book example on how to design your own VTTC from the bottom, as there do for other types of Tesla coils. Its about finding the old radio amateur books and manuals and combine some of the theory from those about oscillators and the knowledge gained by those who have build VTTCs.

The schematics for this dual 811A VTTC are by most people recognized as Steve Wards, but they do in fact go way back to a article from the magazine “Science Experimenter” fall 1963 issue, which are the drawings most dual 811A VTTCs are based on. I have not been able to find earlier material about this Tesla coil design.

I will build a Tesla coil that resembles the one Steve Ward build, in order to be able to compare the two coils. He succeeded in producing 180mm long sparks, due to it being his first VTTC he wrote there was room for improvement. I have tried to correct some of the issues he described and my goal was to achieve at least 200 mm long sparks. A major limitation to this project is money, so its being designed around components I have already available.

Here are some of the things I want to improve.

The high voltage supply needs to be able to exceed the maximum ratings of the tubes. For very short run times the tubes can be driven at a higher voltage / current than they are rated at, in order to achieve longer sparks.

The secondary coil will be bigger, from the unconfirmed theory saying that larger secondary coils gives longer sparks at the same power input in a VTTC.

Adding a Staccato controller to be able to adjust the break rate of the sparks, this will make me able to apply a higher voltage without excessive heating of the tuber, longer sparks can be achieved this way.

Specifications

Tubes 2x RCA 811A transmitter tubes
High voltage supply 230 / 2300VAC MOT with voltage doubler.
Supplied with 160VAC through a variac for 3200V.
Primary coil 180 mm diameter, 1.78 mm diameter copper wire, 26 windings, tapped at 22 windings.
Capacitor tank 4 nF Fribourg capacitor 20kV pulse rated.
Secondary coil 75 mm diameter, 440 mm long, 1600 windings, 0.25 mm enamelled copper wire.
Topload 45 mm small diameter, 152 mm large diameter, John Freud toroid.
Resonant frequency Around 300kHz.
Grid leak circuit 19 windings feedback coil, 1 nF capacitor and 500R to 2K5 grid leak resistor.
Input power 300W at 150V.
900W at 180V.
50W for the tubes heating.
Spark length up to 330mm long sparks.

Schematic

Here is the schematic with all the component values. Here is a few notes on choosing some of the components.

All ground connections are made at the same ground rail.

R2 is a wire wound resistor with L4 wound around it, one for each tube.

C4 is a decoupling capacitor mounted directly under the socket, one for each tube.

R1 needs a high wattage rating, bigger is better, or supply sufficient cooling.

Calculations

Nearly all the following equations is originated from John Freau and is to be found in Steve Wards VTTC FAQ, there is a link at the end of this article. These calculations are not exact and will only work as a guide to building a system that is easier to tweak into good performance.

My goal with these calculations are to determine a Q value and a value for my tank capacitor.

First I will calculate the load impedance for my two 811A tubes. This equation is for class C amps supplied with AC or pulsed DC. If you want to calculate the tube impedance for a filtered DC supply use R = V / (2 * I) instead.

The maximum ICAS plate current rating for my RCA 811A tubes are 175 mA each.

R = V / (4 * I)

R = 3200 / (4 * 0.350) = 2285 Ohm

Next I want to find the primary reactance, f is my resonant frequency in Hertz and L is the primary inductance in Henry.

X = 2 * π * f * L

X = 2 * π * 300000 * (102 * 10^-6) = 192 Ohm

Now the Q value can be determined

Q = R / X

Q = 2285 / 192 = 11.72

Generally a Q of 10 to 20 is ideal. Lower Q gives less tube efficiency, more difficult coupling, but less tank losses. Higher Q gives higher tube efficiency, easier coupling, but more tank losses.

Now I can calculate the capacitance needed for my tank capacitor to work at the desired Q for the tubes I use. Result will be in nF.

C = (Q / (2 * π * f * R)) * 10^9

C = (11.72 / (2 * π * 300000 * 2285)) * 10^9 = 2.72 nF

The closest value capacitor suited for this high frequency and current was three 10 nF capacitors in series for 3.33 nF. The resulting Q for this tank capacitor value is as follows.

Q = 2 * π * f * R * C

Q = 2 * π * 300000 * 2285 * (3.33 * 10^-9) = 14.34

For my final tank capacitor I ended up using a 4 nF capacitor to only have one capacitor instead of a series connection of other types, this capacitor is intended for use in radio oscillators so it can handle the frequency and current, the losses are despite the higher Q value not enough to heat the tank capacitor. The resulting Q value for the 4 nF tank capacitor is as follows.

Q = 2 * π * 300000 * 2285 * (4 * 10^-9) = 17.23

Construction

14th July 2008 to 6th  August 2008

It takes its time to find all the special and at times rare components and materials at a price where I feel its worth it. It would take 1½ month to have found all the things that I needed.

The two 811A vacuum tubes are imported from USA through Ebay. Sockets and terminals are imported from England through Ebay. Heater transformer is bought from a private person in Denmark. The high voltage supply are from old microwave ovens with their capacitors and rectifiers. The rest of the components and materials are mostly thrown out stuff from various industry and private persons.

9th August 2008

The staccato controller is assembled on a breadboard and tested.

12th August 2008

The  staccato controller is build on a vero board.

Schematics for the staccato controller, there are slight changes from the original schematics. I do not use a centre tapped transformer and have removed the polarity switch.

Here is the layout for a vero board, green traces are the traces of the vero board, remember to cut them where they do not join. Red traces are copper wire jumpers on the component side of the board. Notice that the red jumper next to D2 is connected to 14VAC in N along its way, this is the only jumper that does.

30 – 31st August 2008

The secondary coil is wound and enamelled with 3 layers of varnish.

75 mm drainpipe with 1600 windings of 0.25 mm enamelled copper wire.

5th September 2008

A test setup have been build to experiment with the optimal number of windings on the primary and feedback coils, adjustable from 18 to 26 windings. With these options I can adjust the systems efficiency, frequency and coupling. In the first nights of testing 80 mm long sparks are achieved.

6th September 2008

The Tesla coil is tuned and adjusted

2300V high voltage supply, 1K/1nF grid leak circuit, 24 primary coil windings, capacitor tank 3.33nF.

160mm long sparks are achieved.

This is where I discover that my vacuum tubes are not conducting evenly, which they should as they were bought as a matched pair. I solved this problem by driving the weaker tube with 1 winding less on the grid leak coil, that was enough to adjust the screen voltage to a level where the tubes would conduct very close to matched conditions.

20th September 2008

A voltage doubler is added to the high voltage supply.

3200V high voltage supply, 3K/1 nF grid leak circuit, 24 primary coil windings, capacitor tank 3.33 nF.

260 mm long sparks are achieved.

3rd October 2008

3200V high voltage supply, 0.5K/1 nF grid leak circuit, 22 primary coil windings, capacitor tank 4 nF.

280 mm long sparks are achieved.

15th October 2008

The staccato controller is connected to the Tesla coil and it is now a completely different world, to modulate the sparks. It can vary the number of sparks per second from 1 to 50 outbreaks, the sound and shape of the sparks changes all the way through that scale.

Up to 330 mm long sparks can now be achieved through over driving the tubes further with a lower break rate.

27th October 2008

New clear PVC have been bought along with nylon supports for the primary, screws, shoes and fittings.

1st November 2008

Removing the isolation from 15 meters of 1.78 mm hard copper wire took a real long time as I need it to stay in its circle form. It did cost me a deep and violently bleeding cut to my thumb because I did not pay enough attention when cutting off the isolation. But I got the job done in all my bandages.

A platform was build from thrown out wood and clear PVC, that does not mean its not in good condition.

High voltage, filament and 12VDC transformers are placed on the platform, the capacitor is for the high voltage doubler.

All wiring is done, staccato controller added and the grid leak circuit is in place, everything is in its place.

Sparks

The complete coil in all its beauty, the blue and purple sparks go very well with the warm soft glow of the tubes.

This series of close up pictures really show the thin sword like sparks that are characteristic for a VTTC.

Demonstration

Running in CW mode.

Running in interrupted mode

Conclusion

The most important goal was to achieve at least 200 mm sparks and I went through the roof with sparks leaping out to 330mm.

I had assumed that the grid leak voltage would be so low that the coil could be wound on top of its own layers, but eventually it burned through and I had to find a place for the new grid leak coil wound in one layer. The 3 positions I had made for the grid leak coil in the nylon supports were no longer of any use unless I took the whole primary support apart and remade them, which I decided not to do.

The new grid leak coil is still made from 0.5 mm diameter enamelled copper wire, this is not optimal as its isolation is not thick enough, ideally normal wire with PVC isolation would be to prefer.

A higher Q value seems to be preferred if you have a tank capacitor that is able to withstand the heavier load.

Too much of the construction was not drawn in hand before building, that is a sure way to run into problems, but I built most of it from my head and took the battles along its construction, something that I do not advice others to do and I will try to avoid this in the future. 2D/3D CAD drawings would have been a great help, so for the future this will be standard in bigger projects.

 

Further reading

http://stevehv.4hv.org/VTTCindex.htm Steve Ward have documented 5 of his VTTCs, the staccato controller and written a good FAQ that is a must for anyone wanting to build a VTTC.

http://drspark.org/ Christopher Hooper have documented 3 of his VTTCs.

http://www.mif.pg.gda.pl/homepages/frank/vs.html Vacuum tube database, almost any tube data sheet can be found here.

22 Responses to Kaizer VTTC I

  1. Richard Harris says:

    Very nice results! I just completed the staccato part of my single 811-A tube coil, but for some reason R18 (being the 100 ohm resistor in your schematic) keeps getting extremely hot. The half watt resistors kept burning up, so I put in a 2 watt resistor, and that puppy starts smoking if the coil runs for too long. Everything seems correct on the board, as I have checked it several times. Could I be hooking the SCR up incorrectly? What do you think would cause this problem? Thanks, Richard Harris

  2. Mads Barnkob says:

    Hey Richard Harris

    Sounds like something with the SCR is connected in a way so that it tries to draw current into the cathode or ground from your gate drive to the SCR, that way it pulls way too much current.

    This is what I could just think of, might be something else :)

    Kind regards
    Mads

  3. Richard Harris says:

    Unbelievable! Had the SCR hooked up incorrectly…LOL!

  4. Richard Harris says:

    You were absolutely correct Mads. Thanks so much for the troubleshooting help :)

  5. Mads Barnkob says:

    No problem, I am here to help :)

  6. Rob says:

    Very nice coil. I have built the same dual 811a coil and would like to add a staccato controller but am unsure how to choose the right SCR?

  7. Mads Barnkob says:

    Hey Rob

    I just used one I had lying around, just be sure it got a voltage rating 2-3 times higher than your filament voltage and maybe 1-2 times higher current rating than your filament…

    That should do it :)

  8. Mike says:

    I’m still wondering about the SCR. My filament voltage is 6.3volts, but won’t the SCR see potentially 1000’s of volts when it’s turned off as the filaments will float to high voltage?

  9. Mads Barnkob says:

    The filament will only be capacitively coupled with the plate voltages.

  10. Mike says:

    Hey Mads, I’ve been bouncing back and forth between this writeup and Steve’s VTTC FAQ, and I think you might have your Tube impedance equation wrong. Using two tubes I think it should be R = 3200 / (4 * 0.175), or R = 3200 / (2 * 0.350), since the value of K in a Class C oscillator should be 2.

    Is this a mistake, or did you do this deliberately because you plan to overdrive the tubes to 200% of plate current?

    Either way I can’t argue with the results.. 330mm is fantastic!

  11. Mads Barnkob says:

    Hey Mike

    As written this was an idea by John Freud to double the figures if we used pulsed DC or AC supply, you are right about what you say, for a filtered DC supply. I changed a bit in my description to make it clear.

    Thanks for kind words :)

    Kind regards
    Mads

  12. Mike says:

    Thanks Mads, I missed that. It’s all starting to make sense to me now :)

  13. Jeff says:

    Does anybody know what way the feedback coil is connected ie does the top go to the grid of the tube?

  14. Mads Barnkob says:

    Hello Jeff

    It depends on the polarity of your primary, in my case the top of the feedback coil goes back to grids. If it does not work one way, try to reverse polarity of the feedback.

    Kind regards
    Mads

  15. cyborgcnc says:

    Hello, GREAT BUILD and information….what are the white bars, bolted together with hardware on this picture? http://kaizerpowerelectronics.dk/wp-content/gallery/2008_11_06_-_kaizer_vttc_i/IMG_1023.jpg..are they resistors?

  16. Mads Barnkob says:

    Hello cyborgcnc

    Thanks for the kind words and yes, the white bars are power resistors, set up in this way to make a tap-able grid leak resistor so I could vary the load from 500R to 2500R in steps of 500R.

    Kind regards
    Mads

  17. cyborgcnc says:

    Hello again, and thank you for your reply. I was wondering if you have calculated the resonant frequencies of your Secondary and Primary coils? I am running through some calculations/equations as I have a couple of 811a’s around, and thinking of building one of these things!

  18. Mads Barnkob says:

    Hi cyborgcnc

    It is around 300 kHz, if you want a precise figure you can plot my numbers listed under specifications into the JAVATC calculator. http://www.classictesla.com/java/javatc/javatc.html

    I do also suggest that you use this wonderful tool to design your Tesla coil.

    Kind regards
    Mads

  19. Spiro says:

    One last question if you do not mind: How do you ground your Tesla Coil, in other words, how did you ground your secondary? Did you simply use the mains ground to ground it, or did you use some sort of an external ground?

  20. Mads Barnkob says:

    Hey Spiro

    You can use mains ground if the power levels are modest, something you could have on a table. Just be careful, if you are unlucky you might damage other electronics that is connected to mains.
    You can also use a normal earth connection, a 2 meter spear in the soil.
    A artificial ground plane, make a ground plane of sheet metal, aluminium foil that the coil stands upon and connect that as ground.

    Kind regards
    Mads

  21. vasil says:

    Beautiful coil Mads. Good looking and nice sparks.

  22. Mads Barnkob says:

    Hey Vasil. Thanks :)

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title="" rel=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>