Bullet Proof Crockcroft Walton multiplier- all tubes

[3l]

under construction:
Read this or Die!

The procedures described here are lethal voltages.
A double whammy in that it is high voltage plus high amperage..
raises the danger geometrically.
Use proper high voltage procedures.
Diodes will emit xrays as a matter of course.
Never run equipment with out safety interlocks or disharge
shunts for safety after a run.
Alway allow 3 times the discharge time for added safety.

My design parameters are as follows
470 pf charging caps for easy discharge after shutdown.
2x2 tube diodes for fast response and invunerability to
overload or accidental sparking.
High voltage isolation on each filament on each diode.
A relatively cheap high voltage driver to run the cascade.
USE COMMON PARTS!
Seems like a tall order until you realise that tubes simplify the dickens out of the cascade multiplier.
Any high power triodes will work.
A Russian substitute for the 811A is ok.
I picked the 811A in order to match the transformer with the fewest tubes.
Heres the specs for the 811A triode
Plate 1500 volts ......close to 2000 volts
will handle 65 W dissipation ...2 tubes 130w..close to the mot wattage
175 ma 2 tubes make 250 ma... Four would be better but a variac to control kick back ...reduces the amps to a resonable amperage. (voltage too!)
grid 50 ma
filament 6.3 volts @ 4 amps
I use a rewound mot tranformer (Mot2 on the Diagram)
Remove the secondary and wind on 2 1/2 turns #10 Rhomex
wire...repeat tie the two closest ends together to make the center tap.
To get a usable fusor output voltage of say 40 kv @10 ma
would take a 20 stage multiplier.
400 w is required at this level.
The impedence (ohms) V=IR ( 4.00 x 10 ^4 volts ) =(1.00 x 10 -2 amps) x ( ? ohms)
Ohms = 4 x 10 ^6 ohms or 4 megaohms
470 pf /20 = 23.5 pf
The rc constant is (2.35 x 10^ -11)(4 x 10^6)=9.4 x 10^-5 sec
at this rc rate 150 khz is in order.
To power the diode filaments, an old trick brought up to date is used.
In the old days they would have a motor generator set to run the diode filaments.
The motor and the generator would be attached to each other with an insulative drive line.
In the 1920's that meant a huge setup.
Today with high strength magnets , two racing RC motors tied through a plastic rod will do.
Diodes are not as power hungry as the triodes.
Any 12 volt trani the would have the required isolation.
to run the motors.
An rc network tied to the triode grids makes the frequency that the triodes will conduct at.
So 150 kz is 1.5 x 10^ 6 hertz or a time period of 1/freq roughly 9.4 x 10^-6 sec. That is the rc time constant to make.

if you use a .0005 mfd mica cap.
thats 5x 10-10 farad x (x)=9.40 x 10^-6 sec
x= roughly 5.3 x 10 ^5 ohms or a pot of 50 kohms
So I would chose a 100 kohm pot so that 50 kohms would be half way.
These calcs are just estimates ... you must make the parts you want to order fit within the operational voltages you desire.

Follow these steps

0... find the wattage for the desired voltage and amperage
1.... get the rating of the transformer
2 .......get the rating of the diode tubes
3......... Calculate the required Tc for the size caps you have.
4 ...........size the triodes to the load required
5 .............design the rc network to make the triodes work at the required frequency.
6 ..................check the triodes cutoff or working frequency


I will post diagrams below.

Happy fusoring!
Larry Leins
Fusor Tech

[3l]

A side to side comparison of a tube multiplier stage and a solid state stage

[stinger]

Very interesting idea,appeals to me because I have about a hundred 2X2's and seven brand new National 811A's just "waiting" for such an application. I do Have one question however, I'm not really clear on how RC motors and and any 12 volt transformer can power a cascade arrangement of vacuum diode filaments.....perhaps I'm dense,could you go into this particular aspect in a little more detail for me? thanks.....Mark

[3l]

Hi Mark:

I learned about this stuff when Crockcroft's junior assistant
lectured at Trinity University in San Antonio Tx. in 1977.
He brought a gob of pics of the old Rutherford Labs.
I asked him about the equipment and such.
The 2 motor trick is very very old. You need two permanent magnet motors. If you drive a pm motor it will make juice.
The motor hooked to the filament will be at that tubes working voltage so you connect it by a long plastic rod to the motor driving it. The driving motor is at ground potential so it can be wired to a normal 12 transformer hooked to the wall. This system has been used as high as 10 million electron volts. My old physics buddy assured me that this method is only limited by the modulus of deflection of the material making up the drive train. They used a dry ash rod. To size the motor simply take the filament volts times amps the tube draws divided by 747 gives the horse motor you will need. Formally the setup I've described is called the RF modulated cascade multiplier. The motors of the old days weighted in at 200-300 lbs, they used anico magnets. With modern
magnets such as samarium cobalt or neodynium iron the motor is very miniscule. Large scale RC racing cars use fairly large motors to achieve scale speed of 100 -120 mph. They would be ideal as they are small enough to fit below the 2x2 tube.

Happy Fusoring!
Larry Leins
Fusor Tech

[stinger]

Thanks Larry,Its really amazing how basic yet effective that particular scheme is-and useable up tp 10 Mev. It speaks volumes on the ingenuity that existed at that time. There's another way to implement powering the 2X2 filaments and get the proper isolation; One could place a series of small(20W) ferrite E-cores(one for each filament) place three sets of windings on each core-maybe 2-3 turns each. One winding goes to a "power oscillator" that runs at @20 khz and will provide 2.5 volts at about 30-40 amps, the second winding goes to your tube filament-the third(tertiary) winding distributes the 2.5v at 20Khz to the next ferrite core-and up the stack we go.
Okay...I said my piece....now I'm ready for anyone to "punch" some holes in the idea! Mark

[3l]

The 2x2 tube's filament is 2.5 volts at 1.75 amps.
Roughly 5 watts / tube
So 5/747 =roughly 7 x 10-3 hp or 7 millihorse.
Rc motors territory!
I plan on overdriving on their rated amperage of of just 7.5 ma watts to twice that to 15 ma.
However if you want long life then put two 2x2s together to make a 5 v device and use a 6 volt Rc Motor to drive it.
Their voltage range is 4500 Volts so you could put two mots in series and still use these tubes.
In order to make say 40kv would take only ten stages.

Happy Fusoring!
Larry Leins
Fusor Tech

[Mike Veldman]

I gave your psudo-dynamotor filiment supply a try last evening. Not that I didn't have anything else to do but, I was in one of my storage areas looking for something and came across a box of PM motors. So naturally an hour or so was lost on an interesting detour. I've included some pictures. First I did the two identical motors, DC to DC so to speak. I was able to get adequate voltage and current, but the drive motor was pretty warm. So I moved to plan B, I used my "Okie Dremel Tool" (sewing machine motor with a chuck) to run the PM generator. I didn't get around to putting a tach on the rig but the RPM to get the amp and a half for the 2X2 being used as a load seemed critical. Are you sure you can do this consistantly with slot car motors? Didn't try putting high voltage on the rectifier yet, I'll save that for the next time I need an interesting detour. Sorry the pictures don't show it running, I ran out of hands while taking the pics.
mike

By the way, I have about 50 of these PM motors, if anybody'd like some to play with, I have no clue as to the motor specs.

[3l]

The drive motor always gets hot!
you can safely blow air on it.
I use an old hair dryer with the heater removed.
or liquid cool it.
The ones I had in mind are not slot car motors but RC racing motor that run 1/4 scale and up. They have real ball bearings and are not very cheap. A better alternative would be two dc drill motors chucked together. The motors without bearings get pretty warm due to friction.
I just went shopping at the college pawn shop...picked up two dc drills for 5 bucks each. A critical rpm item is how big the wire is in the motor....you need not the largest or the smallest but medium size pm motors. The motors you are using seem to be the right ones. It is the high starting inertia that is a real head ache but the two simmilar motors prevent over speed. Just wait till you put on a long stick! When you sock on the juice let the filament warm up for two minutes or so to prevent it from exploding.

Keep on Fusing!
Larry Leins
Fusor Tech