AC High voltage Fusor Power Supply?

[Tom Dressel]

I just wanted to throw this out for what it is worth.

I am thinking about building a variable frequency AC high voltage power supply for the fusor. With the chamber grounded, and the inner grid at high voltage AC, the idea would be to tune to the "resonant" frequency of the chamber. The "resonant" frequency presumably determined by the chamber pressure and the applied voltage. This would alternately drive a large percentage of the positive deuterium ions out to the periphery of the chamber during the positive part of the cycle, and then draw them back toward the inner grid during the negative part of the cycle. This at least to my thinking could produce higher ( although intermittent ) deuteron density in the reaction zone

Does anyone have any ideas about why this would or would not work?

Tom Dressel

[Frank Sanns]

Hi Tom.

Check out the Images Section for the amplifier that I use for supplying RF to the fusor. Many people talk about microwaves for ionization but for resonance of the deuterons in the fusor, the frequency is between 10 MHz and 14 Mhz.

As soon as I get some good 1 inch and 3 inch inner grids welded up I will be resuming this work with good neutron production numbers. Good luck if you get to it before me as I think it is an interesting area.

Frank

[Tom Dressel]

Frank:

How did you come up with the resonant frequency of 10 Mhz to 12 Mhz?

Tom Dressel

[Frank Sanns]

The resonant frequency is the frequency that the ions (ie deuterons) are oscillating in step, in and out of the inner grid. For the sake of simplicity for this particular post I will not go too much into the details but rather just show the calculation for one condition as several fusor parameters have to be considered to find the resonant frequency.

The idea is to calculate the transit time of a deuteron for half of the diameter of a fusor so the deuterons can cycle around the inner grid and just before they crash into the fusor walls. This changes for every voltage. We will assume a voltage on the lower end of fusion voltages of 25 KV.

1ev=1.6E-19 Joules
kinetic energy (Joules)=1/2(mv^2)

Using the above formulas, at 25KV and substituting in the mass of a deuteron (3.3E-27 kg) that gives a deuteron velocity of 1.6E6 m/sec.

Dist/rate=time and for my fusor the diameter is 10 inches. Thus the radius is 0.13 meters. This gives a transit time of 8.4E-8 seconds. The reciprocal of the period is the frequency which is 11.9 Mhz for these conditions.

Yes there are assumptions as the voltage is sinusoidal and not constant but I hope this shows how one can calculate a resonant frequency for a particular set of conditons.

Frank

[DaveC]

Tom -

There could be some operational efficiencies that the DC system doesn't have. With a steady DC potential,, the ions will gradually converge to the central regions. The AC will gather them back and re-sling them...

Would be really interesting if you could get an entire spherical shell of ions moving in and out synchronously, .would probably eliminate the need to be concerned with recirculation.. since it would be happening.

I need to think about this a bit more.. interesting idea.

Dave Cooper

[DaveC]

Frank -

I think the freq is 1.19 MHz.
(8.4E-7 = 0.84E-6 and 1/0.84E-6 is 1.19 E6 Hz.)

That makes the drive electronics a bit easier.

Dave Cooper
Dave Cooper

[Richard Hull]

There are two ideas regarding resonance in a fusor.

1. Ion resonance which has been discussed here is a form of particle shuttle resonance.

The problem I see is that a perfect sinusoid would still waste energy. What one oscillation would give in central grid collisions, the other would take away as the ions slammed into the outer shell. Perhaps a complex sinusoid of lesser reverse period would do the trick. ....still not good for the return to normal period on the next half cycle would need to be foreshortened. You are still turning ions around twice each cycle and that is lossey.

2. Cavity resonance puts an intense energy node focused within the spherical chamber or cavity. The latter is definitely in the gigahertz range.

The multicusp effort struggles at cavity resonance.

ion acoustical resonance (waves of timed or bunched ions) sounds interesting, but would demand tremendous oscillatory voltages similar to the current DC applied potentials. This might be a job for a custom driven RF transformer.

Richard Hull

[Frank Sanns]

Thanks Dave. I rechecked my math and found that I just wrote done the wrong exponent on the transit time. The final frequency is indeed correct at 11.9 MHz. It is fixed in the post.

Also check out the post Re: The "runaway reaction" and my thoughts on the subject. from July 2003. More information on this and how this kind of resonance can happen with RF input and MIGHT even happen very very rarely by symmetry in a DC fusor also.

Frank

[Tom Dressel]

Attached is a picture of my function generator. It's max frequency is 2 Mhz. It also has some ability to vary the waveform symmetry as suggested by Richard Hull, and a sweep function.

I was hoping that the power supply might be as simple as a few ignition coils in parallel, driven through some power transistors at a few KHz, but a 30 KV -11 MHz power supply is out of my league.

Tom Dressel

[guest]

You can easily create high frequency oscillators using TTL logic chips. Use a 7414 (or 74LS14, 74HCT14, etc) "Hex NOT logic" chip. Connect pin 7 to Gnd, Pin 14 to +5 Volts. Connect 1K variable resistor to pins 1 and 2. Connect a 5 to 10 pf capacitor from pin 2 to GND. Output is on pin 2.

This will create a simple hi frequency square-wave oscillator. You can adjust the frequency by adjusting the resistance across pins 1 and 2. If you want to lock the freq you can attach a freq crystal across pins 1 and 2 and use a static 1k ohm resistor.

To increase the power output use a hi freq op amp, or a mosfet. Now all is left is the really hard stuff! The problem is going to be building a switching circuit that is capable of driving 500+ Watts at 40 to 50 KV at 10+ Mhz. Good Luck!

[grrr6]

You could also try adding the AC to an existing DC input. That might have some neat effects.