Octavio Part Three

. donderdag 27 oktober 2011
0 reacties

So, in octavio part two the conclusion was that the pinch coil configuration (current in the outer coils opposite to the current in the core) cannot create a virtual cathode.
So the only configuration left now is the core with focus coils:

The focus coils with the core will create a magnetic field which is zero in the center of the core and with straight field lines between the core and the focus coils.
To understand the idea, take a look at the following 2D analogy (not a cross section, just an analogy, you can click the image to see it enlarged):
Outside the core the electrons will spiral in the magnetic field, near the center they will widen their spiraling path and they will be repelled outwards by the electric field of the virtual cathode.
As explained in octavio part two the electrons will slow down because they loose energy when they are accelerated/de-accelerated by the electric field between the focus coils and the core, and when spiraling in the magnetic field, bremsstrahlung and synchrotron radiation losses.
As explained in part two, to compensate for the losses an AC voltage can be superimposed on the DC potential of the focus coils, this will also introduce extra losses (bunch induction, explained in part two), which can be compensated by the same AC power.
This will create an oscillating virtual cathode, the idea is to create a situation where the DC current on the high voltage positive core is almost zero.
The voltage between the wehnelt cylinders and the focus coils should be negative most of the time, only when the AC power is at it´s positive peak the electric field between the wehnelt cylinders and the focus coils should allow some electrons to be injected into the system, the AC power maximum positive peak is the moment when most of the electrons are near the focus coils, the negative space charge in the center of the focus coils at that moment can prevent too much electron injection (in other words, the input current is self regulating when the focus coils have the right negative DC bias).
Some preliminary calculations show that the resonant frequency can easily get higher than two gigahertz, but it will be lower when the magnetic field is stronger. Feeding AC power to such a construction at such high frequencies is another challenge.
I've been thinking about a self-oscillating octavio, but naturally every oscillation will get it's AC power by slowing down electrons, keeping the electrons at a high speed will always dissipate AC power. Any self oscillation (some kind of buncher/catcher or reflex-klystron-like construction) should be between the the electron guns and the focus coils, not between the focus coils and the core.
Still lots of calculations and simulation to do, and I have to find a way to make it self-oscillating.

Polishing the ground plate

. dinsdag 25 oktober 2011
0 reacties

Did some polishing last weekend, here is picture of the ground plate, nice and shiny:

The ground plate will have two 1-pole, two 8-pole and one 4-pole feedthrough.
Also tried to fix the mass spectrometer, still have the problem that after bake-out all the QMG064 sensors fail.
I tried to clean the grid with fine quartz powder and ultrasound cleaning, no effect at all..... to be continued.

Octavio Part Two

. vrijdag 7 oktober 2011
0 reacties

So, how would it be possible to make a virtual cathode with the spherical octahedron configuration?
What we need is a volume where the magnetic field strength is low or zero, confined within a stronger magnetic field.
Take a look at the following picture:

Still the same core, but now it's surrounded by 8 coils, this picture also shows the 8 electron guns.
The core is at a high positive electric potential (e.g. +50kV), the outer coils (pinch-coils) are at a medium positive potential (e.g. +500...1000V) and the wehnelt cylinders are at a low positive potential (e.g. +50V), the hot cathodes of the electron guns are slightly more positive (e.g. +55V) than the wehnelt cylinders. (to prevent electrons flowing from the electron gun cathodes to the grounded reactor wall the electron gun cathodes are not at ground potential.)
The current through the outer coils creates a magnetic field in the opposite direction of the magnetic field created by the sphere triangle it is facing, cancelling the magnetic field in the center of the sphere triangle.
They will "pinch holes" in the magnetic field of the core, that's why I call them pinch-coils. (originally I thought these coils would have to have their magnetic field in the opposite direction, making the field lines outside the core straight in stead of diverging, creating a strong focus effect for the electron beam, it might still be interesting to do that experiment also, in that case I would call them focus-coils).
The volume where the magnetic field is low or zero takes the shape of a "stretched cube with long points", as if you would take a cube and pull the corners until they are long enough to just touch the centre of the outer coils.
This is exactly the shape needed to confine the electrons in the centre and be able to inject the electron beams.
Just as in the polywell, this could create a cloud of electrons in the centre of the core.
But that still means that electrons loose energy and slow down, the fastest electrons passing through the centre of the core and the slower electrons drifting towards the grid and hitting the grid.
As in the polywell this process will "blur" the virtual cathode.
When the electrons are leaving the core, they slow down, turn around, get accelerated in the opposite direction and enter the core again. In this process they loose energy (brehmsstralung again), so they will slow down and blur the virtual cathode.
This can be prevented by superimposing an AC voltage on the pinch coil potential. This AC voltage should have the same period as the time it takes for an electron to travel from one pinch coil through the core to the pinch coil on the other side of the core.
The idea is to create an oscillating virtual cathode (in resonance), where the electrons don't slow down, in other words, the losses are compensated by the AC power applied to the pinch coil potential. (The load on the AC power supply behaves  as a capacitor with a high dielectric loss, the "dielectric loss" is the power needed to prevent the electrons from slowing down).
Not only the brehmstrahlung losses have to be compensated, when we have an oscillating virtual cathode the electrons form "bunches", just like they do in a klystron, these bunches induce currents in the core, this energy is also lost and can be compensated by the AC power on the pinch coils.
This is completely different compared to the polywell, in the polywell the electrons have a wide energy distribution and the virtual cathode is a statistical effect, a cloud of electrons, in the resonating octavio the electrons have a narrow energy distribution and the virtual cathode is oscillating and sharp.
Next step would be to see if it's possible to apply POPS, maybe by superimposing an extra AC voltage of a lower frequency on the potential of the pinch coils, maybe it's enough if the the octavio resonance frequency is a multiple of the POPS frequency.

This sounds great, if it was true.....

However, if an electron would be kept on it's track by the "magnetic tunnel" created by the octavio core and the pinch coils, then it's path would be bent in the wrong way when the electron travels the other way.
So we have to use focus coils, not pinch coils, where the electron path outside the core will be a spiral.
I will explane more about the focus-coil configuration in octavio "part three", in part one I explaned that the core alone would not work, the core with focus coils might create a virtual cathode, probably a spherical "shell like" virtual cathode.
The experiment will be about the core with focus coils and an AC voltage on the focus coils to compensate bremsstrahlung, synchrotron radiation and bunch-induction losses.

Octavio part one

0 reacties

The goal of "the Octavio" is to create a virtual cathode without consuming a lot of power.
Ideally it would have the same effect as the polywell, but with less electrons hitting the magrid and without needing enormous currents through the coils.
Why do the coils in the polywell need to carry such a high current?
Because in the polywell the magnetic fields of the opposite coils are canceling each other, electrons are not always hitting the grid because of a magnetic field very close to the grid, the rest of the magnetic field is canceled, it works but it is very inefficient. For "the real thing" big and superconducting coils are needed, if that would be enough.
So I started to think about a configuration where there is no direct cancellation of magnetic fields.
When using six coils in a cube like configuration like the polywell, there will always be asymmetry if the magnetic fields are not canceling each other.
After thinking about all kinds of configurations, hallbach spheres, doughnut shaped configurations etc, the conclusion was that the most simple symmetric configuration where there is no direct cancellation of magnetic fields is the octahedron.
You can see a drawing of a spherical octahedron with the direction of the currents on the top-left of this blog.
Just like the polywell it will be at a high positive electric potential and surrounded by electron guns (8 for the octavio), the electrons entering the Octavio core should create a virtual cathode in the center.
The following picture shows the direction of the magnetic fields, in the center of the core the magnetic field is zero:

Actually Bussard mentions an octahedron configuration in "Some Physics Considerations", a paper he wrote in 1991.
So why did he not try it?
I don't know why Bussard didn't try such a configuration, but I think that the "Octavio core" alone would not create a virtual cathode.
When electrons travel from the electron guns through the sphere triangles of the core, the magnetic field is the strongest at the center of the sphere triangles of the core. It could be strong enough to create beam focusing. Strong enough to make the electrons "follow" the field lines (actually they a spiraling around the field lines, but if the magnetic field is strong enough, the spiral radius is small enough to think of it as electrons "following" the field lines).
If you imagine the magnetic field lines, you can see that the beam would diverge near the center and focus again when traveling past  the center. Maybe this could still create a virtual cathode. However, the field lines are diverging in the outward direction outside the core, so electrons would start circling around the wires of the grid when they leave the core. Also, electrons will loose energy when they change direction or when the are spiraling (brehmsstrahlung) and when they slow down their orbital radius around the magrid coil wires will become smaller, until they hit the grid, effectively creating a substantial negative space charge which will cancel the electric field created by the virtual cathode.
When I have build the core, I will measure the effect of magnetic fields on the grid current, but unless I'm completely wrong (which is quite possible) there won't be any useful virtual cathode with only the core.

PS: Because the core is made of 8 sphere triangles it's called the "Octavio", the name was taken from the book "The Light Fantastic" by Terry Pratchett, in which the Octavio is a book with 8 spells.


. dinsdag 4 oktober 2011
0 reacties

The bottom and top plates of the large vacuum chamber are almost finished!
Took a lot of machining, I still need to do some polishing.
Also found 60kV cable from an X-ray machine, a 80kV 0.2A diode and an X-ray transformer.
I cannot find any transformer oil (unless I buy 50 litre), anyone out there who can sell me some good transformer oil (no pcb oil from old transformers)?

And I want 5 of these, but they are a bit expensive.

This is the upper plate with the viewport: