Fusor construction-Discussion-Pulse research

[3l]

Hi folks:

I am getting substantial requests for the pulse method now.
There seems to be confusion about pulse research in general.
The goal and object of pulsing is to get high levels of fusion reactions by increasing the ions volumetrically (Large numbers.)
for a short period. The power is always peak in pulsing. The "real" power is always the average power. The fusor part works identically to the continious fusor operation. The advantages of pulsing are high average peak power so relatively high average power,the possibility that your input power can be less than the output yielding usable power even at very low efficiencies (I still working on it) ,and high numbers of neutrons.
Draw backs are the faster you pulse the fusor the less time to charge your storage ,grid degradation,very fast switching is required at very high amperage.

The last proviso is what I spend most of my time on the high speed high amperage switching.

Relavent posts:

reasoning behind pulse work
2004-03-11 17:14 Re: A FEW fiddly little details part 2 (larry leins)

Facts you should know:
2002-26-08 13:20 pulsed fusor power issues (Richard Hull) [Latest: ]


2003-08-09 08:38 Debueing a 100 kv pulse fusor (larry leins) [Latest: ] (10)
2001-12-30 13:46 Pulsed Fusor #3: Case Study of a 1HP Fusor (3L) [Latest: ] (5)

In fact I have almost thirty posts under Pulsed Fusor #.
They cover the history and developement of pulsing.

Schematics are in the Files Forum under:

2002-03-21 00:42 XFMR Coupled Pulsed Fusor 10024 bytes 1
2002-03-09 15:00 Pulsed Fusor Schemes 16506 bytes 1

Present day pulse work
2004-03-11 17:14 Re: A FEW fiddly little details (larry leins)



Switching methods used: Spark gaps,Thyratron,Triodes,and the latest pseudospark switches.

The fastest speed switched - 1 ns
The most amps - 10 MA (delivered thru a vacuum feedthrough to a grid)
The max peak wattage 1 GW @ 50 kv.

At higher voltages I use spark gaps only so far.
The pseudoswitch is still in the testing and developement stage.
To trigger the high voltage trigatrons I use special methods.
I use Marx generators propelled by an automotive coil.
A marx generator makes a low current due to the fact it stacks caps of low voltage into a higher voltage . Doing that lowers the capacitance hence it lowers the amperage.

Example if I stack 10 -470 pf caps at 30 kv to to get a 300 kv spark. But the total energy is 470 pf /10 =47 pf...e=1/2 CV^2
=.5 x (4.7 x 10^-12 F)(3 x10^4 volts)^2 or about 2.3 x 10^-3 Joules
or 2.3 millijoules. Not enough for fusion but enough to trigger a trigatron. The energy used for fusion so far is in the 10-1026 joule range. About a million times what a Marx can do.
The Marx is usually set to deliver a voltage that is 2 and 1/2 times the switching voltage (The higher the better). I use inductor chokes rather than charging resistors to reduce losses. All my trigatrons are home made. The gap faces are lined with tungsten sheet to reduce errosion.

All caps must be specifically designed for pulse work.
HIGH INDUCTANCE CAPS WILL NOT WORK PERIOD!
I have home built some of them.
They are all flat plate capacitors with out exception.
Before you go absolutely crazy on pulse read these:
2003-10-22 14:48 Capacitor Discussion from Images Section (Frank S.) [Latest: ] (3)

Happy Fusoring!
Larry Leins
Fusor Tech
,

[Mark Rowley]

Larry, you wrote "Draw backs are the faster you pulse the fusor the less time to charge your storage ,grid degradation,very fast switching is required at very high amperage."

Too bad you live 3000 miles away or I would let you use my giant Aerovox and Maxwell pulse caps. Talk about INSTANT grid degradation !!!
Mark Rowley

[DaveC]

Larry -

Not to sound like a broken record... (does anybody still remember records??. heheh ) ... It seems to me that incinerating the grids, whether in CW or pulse mode operation, is telling us that the ion aiming process is not doing its job. In fact it tells me that grids are intercepting MOST of the ions, (or electrons).

Simple electrostatic lenses have a tendency to intercept a large percentage of ions, particularly if they're in the form of apertures. To minimize ion capture, certain electrode shapes are needed.

I have yet to model a complete inner grid, and run ion trajectories, its on my to do list, when I can get time..

But I am guessing that a double cage inner grid, (one grid inside the other) with somewhat different voltages, the innermost being the highest, will allow a larger percentage of the ions to pass through going inward, But that same configuration , may not be efficient in focussing outbound ions to miss the grid wires.

Some detailed experiements, not done with particularly high voltages or currents may elucidate some of our specualtions about the ion paths. I'm still thinking about how...

Also, IGBT switching transistors are probably your best choice. they have on transition times, in the low us range (0.5 to 2 us) but the circuitry is extremely sensitive to stray inductance and capacitance. Sloppy construction here will make for ringing and slooooow.. ristetimes. As to caps, Cornell -Dublier sells several kV ufd size caps with reconstituted mica dielectric. Quite low inductance and work well with a tranformer driven pulse generator. Probably a bit more $$ than making your own...

Dave Cooper

[3l]

Hi Dave:

I too have wondered what the ideal grid would be.
Oh ..IGBT ,they are not as good as a backlit pseudoswitch.
The simpler BPS has killer amperage handling.
My first crude BPS exceeded the IGBT amperage handling capacity by a million fold.
(10^10 amps in 1 nsec @ 100 kv)
With refinement, I should be able to reach 10^11 amps...just like commercial units.

Happy Fusoring!
Larry Leins
Fusor Tech

[Frank Sanns]

Hi Larry,

A way to minimize the number of ions that will hit the inner grids is to turn it off before they get there. Pulsing or using RF is the way to do that. If the chamber is larger and your pulse is short enough or the RF frequency is high enough then the grid will go to zero potential before the ions get there. It is when the ions are getting close to the inner grid that they are most steered towards the inner grid wires becasue of 1/r^2 .

Frank S.

[3l]

Hi Frank:

Thank you.
You are the first person to pick up on this fact.
The pulse duration is determined to shut off just when deuterons reach the grid. By making the duration a nanosecond or less by the time the deuteron reachs the proper speed the accelerating pulse is done. Deuterons in close proximity to the inner grid won't be attacted to it by electrostatic forces.

Happy Fusoring!
Larry Leins
Fusor Tech

[DaveC]

Larry -

I think you need to run the numbers on how fast your ions will be travelling. I doubt you can accelerate D+ ions to a fraction of a meter per nanosecond velocities... right?? C being about 0.3m/ns..

Also, regarding the BLPS (back lit pseudo switch.) 10^10 amps for 1 ns converts to 10 coul of ions. Where did you get this many? Are you speaking of switching a metallic circuit? Got a little bit lost in the description, sorry.

Dave Cooper

[3l]

Hi Dave:

It is good to see lots of questions.

Don't think of the acceleration of just one ion...its is a large mass of them at 100 ma glow discharge milling around at random.
It takes more than one nanosecond to get that mob of ions to fusing velocity. You were wondering how did I get 10^10 ions,well the current availible by my 7kv @100 ma transformer
can make a total of 10^18 singly charged deuteron ions. By making the inner grid fairly small compared to the can the space charge problem can be contained for short time.

The pseudospark apparatus uses the entire inside of the electrodes as the active surface hence incredible total amperage. It acts like an increadibly large vacuum spark gap.

For the test pulse:
I used a variation of explosive generation to generate that pulse.
I use a shot gun shell to propell a magnetic slug past a single copper turn. I have a friend who reloads his shot gun shells.
he loads a special shell for me...a shell with a sabot that has a Neodynium Iron magnet in it.The voltage is not very high but the amps with Ne iron is astounding! A tip of the hat to Mike Veldman again!
The reprints of IEEE pulse articles have been very fruitful.

Thanks Mike.

Happy Fusoring!
Larry Leins
Fusor Tech

[DaveC]

My questions Larry, have to do with the internals of the Fusor, not with inducing a huge current in a metallic conductor. At the normal fusor pressures... 10^-2 torr, you have about 10^-5 atm pressure.. give or take a little. If.. your fusor had a volume of 22.4 liters, then this internal pressure would give 10^-5 moles of Deuterium molecules. or about 6 x 10 ^18 molecules.

With exception of the Bell Jar types, most of the fusors are less than one ft in diameter or somewhat less than 1/2 cu ft volume.. about 14 liters - or about 2/3 of a Gram Molecular Volume of 22.4 liters. Thus the 10^ 18 molecules is ball park.

With only this many molecules, total ionization ( a complete impossiblity) would net you a mere 1 coulomb of ions. How can you get more than 100 coulombs of charge from this.?

Is your linear apparatus that much bigger? (100 times the volume) But perhaps I am missing something important, here.

Also, I think you will have a serious impediment in getting any cathode to emit this sort of current density. LaB6 cathodes will electron "bleach" long before this...yes? And these are some of the stoutest of the high current density cathode materials.

I have been working with some industrial applications lately of pulsed transformer drivers, and when you begin to address the problem of converting a stupendous low potential, current into a high voltage, not-so-stupendous current, there are nasty little realities of leakage inductance, parasitic capacitances and etc... that genuinely complicate life, and often spoil the fun.

To support 20 kV or higher accelerating potentials, the gas density in the fusor must be much lower, and thus the amount of ions is proportionately smaller still.

Not trying to be a wet blanket on creative thought... but there seem to some realities we need to address, if we are to have some hope of success.

Dave Cooper

[3l]

Hi Dave:

Not a problem .
The cathode problem is pretty steadfast pain in the tail.
I have gotten no where near 10^10 ....for all the reasons you sited.
More like 10^9.
Perfect ionization just occurs in physics books...never in real life for sure.

I have stuck with a simple scheme to eliminate the "hairy assed problems of stray inductance".
I have quit using step up transformers after the capacitor for now in the pulse work.
Like you say the waveform goes a little nuts and stray capacitance, inductance are real time killers.

I'm just saying the electronics have the ability to deliver 10^18 ions IF I can figure out the rest of it.

For really high amperage stuff 10^-5 torr is an ironclad requirement.

Ive been playing with photo ionization by UV Laser before the big pulse.
Theoretically it should be better than Glow discharge but we shall see.

I discarded B La emitters 2 years ago.
RF is the way to go.
It is the perfect use for the SS Thyratron that won't destroy it.
The SS Thyratron was benched pretty quickly...pseudospark was replacing it anyway....
It would have been a pretty junk thing in my lab's recycle pile had I not bothered to put a hv fuse before it.

A 13.6 Megahertz RF discharge plasma is the standard now for fusion emitters.
It is fairly simple to build for an amateur .
Two feedthroughs on a conflat.
A loop of copper or ss in a vacuum.
It has to be porcelian coated for durabilty.
A fairly high power RF Generator is detailed in links
Under the post:
2004-04-24 09:28 MultiMegaWatt RF driver (Brian Willard) [Latest: ] (5)

Oak Ridge claims 10^19 ions out of it for the Spalation Neutron Source...but they are the government....I don't think a person of limited means could do those numbers.
Mr Megley got me interested in this method three years ago.

Oh Btw the linear and reflex ion gun are differentially pumped just like the Pontiac fusors....You can hardly get good ion numbers at 10^-6 torr.

New****
It looks like Richard Hull's suspicion of the accuracy of silver activation is well founded. It seems to vary +or minus 20 %.
I half expected the count to rise but it fell instead.
Compared with BF3....numbers so far have fallen down to 1x 10^8 neutrons per 10 pulses at 50 kv pulses. I will reshoot the 100kv pulses to see if it rises any.

Happy Fusoring!
Larry Leins
Fusor Tech

[Richard Hull]

Good man Dave!

I have been out of town for five days and away from discussions. I Would have asked the same questions.

Numbers are important and they must be explained in the light of realizable phyusical operations and constraints. Pulse work can claim some big numbers, but still remain rather pathetic for the most part once intergrated over time. Even if still impressive over time, the losses mount with each ever faster operation.

A coulomb is just a raw number of charges, a coulomb-second is an amp and relates directly to the joule or watt second which is a rate of delivery of charges. Large numbers claimed here must be explained. Because joules are real. Unfortunately, it is the nature of nature to be really lossy at high di/dt with a lot of RF and other coupled and radiative losses. Condensed matter is no friend to rapid change. Try and hussle it along and all it will do is produce heat or radiation at some point forgetting all about what you originally WANTED it to do.

I have done a good bit of high energy pulse work and the motto there is "Speed Kills"................Kills the process, kills the chance of success, kills gear, kills the urge to go much farther along any speedier path.

One only needs to beat one's head against this natural wall for a few times before the warm blood in the eyes makes you stop or, at best, gives pause to your next attempt at wall bashing.

Richard Hull

[3l]

Hi Richard:

Good to have you back.
Pulsing is not for the weak that's for sure.
I tried to stay away from spending too much cash in the specialized pulse area.
In case I did hit a brick wall.
The only exception was my PAR high speed boxcar unit...I wanted it anyway.
Most of my stuff is general purpose.
I pulsed while the linear and continious parts got pulled together.
I now have the parts and ps needed for high power work.
Pulse is cheap at low levels but expensive at higher levels.
To be honest , I think pulse is a dead end as far as high power goes but it kept me entertained for 3 years. I got to recreate all the pulse work of Mather,Milley and others, learned Pseudospark Tech,high speed pulse measurement.
Gathered nim and other instruments ,outfitted my lab.
Was it wasted ?
Nope.
It lead to the high amperage linac colliders(so called wall plug linacs).
Nearly ALL linacs are pulsed in this mode.
All the pulse research on the electrical side will be used.
These devices working at 10^-6 torr have better efficiency than any standard fusor on mean free path.
So the ideal of pulse takes shape in a new form.
Pulse was a make do until high power could come on line.
I've come full circle back to Pontiac fusors.
Now I've earned the right to criticise the pulse work of others (scars to prove it)
but don't worry , I have a soft spot for fusor folks!

It's Lanl and Livermore that got me laughing!
I only blew a coupla hundred bucks and a gob of hours.
But it fell into the fun catagory.
But I have a working lab and got financing for the real deal,
by demonstrating pulse stuff .

What have the major labs done?
Next to nothing!
I am I embarrissed ?
NOPE!.... got plenty of company.

BTW the 100 kv stuff did not raise the neutron output at all.
The maximum I could get was 1X10^8 neutrons per 10 pulses

I'm moving on to linacs and reflex guns where the high power pickings are much better.

Happy Fusoring!
Larry Leins
Fusor Tech
The Linac Fusion Guy

[jst]

Now that your moving on, just think; all custom pseudo spark "vacuum tubes" with no solidstate rectifiers...

Guess what I'm thinking???