Ion Generation Troubleshooting: Overcoming Limitations in Small Ion Accelerators

[Javier Lopez]

I have to increase ion generation in a small ion accelerator but I have troubles to go over 20mA (1.27e17 ions/second). I can modulate up to 40 Mhz at 1kV. What ion generator it is recommended gor large ion flux?, I can work in pulse generation.

 

[Henryk]

it is hard to offer any suggestions not knowing what is your ion generator like. However, from my experience, your ion current might be limited by space charge.

 

[Javier Lopez]

My actual "generator" not works. In the designs no more than 20mA can be done but heating a lot the copper to generate electrons. Pressure is in the 1-5Pa range.

I would like use 2.4-3Ghz microwaves to heat hydrogen to ionic state instead of using hot OBe to emmit electrons and use electrons to ionize hydrogen. I can use solid state amplifiers or magnetron, but before buy and mounting I should like have the efficiency. The question is: what could be the efficiency of the RF?
(I should use solid state amplifier or magnetron)

About the space charge, it would help to introduce RF connected to the grid.

Can be the output charge: Q=CV*f?
Where C is the capacitance of the acceleration plates, f the signal frequency and V the voltage.
I suppose there is a maximum frequency at that the ions can not reach the second plate at the output speed

 

[trurle]

My actual "generator" not works. In the designs no more than 20mA can be done but heating a lot the copper to generate electrons. Pressure is in the 1-5Pa range.

I would like use 2.4-3Ghz microwaves to heat hydrogen to ionic state instead of using hot OBe to emmit electrons and use electrons to ionize hydrogen. I can use solid state amplifiers or magnetron, but before buy and mounting I should like have the efficiency. The question is: what could be the efficiency of the RF?
(I should use solid state amplifier or magnetron)

About the space charge, it would help to introduce RF connected to the grid.

Can be the output charge: Q=CV*f?
Where C is the capacitance of the acceleration plates, f the signal frequency and V the voltage.
I suppose there is a maximum frequency at that the ions can not reach the second plate at the output speed

Are you making a sort of plasma thruster? Seems like Hall thruster upgrade to Vasimr or E-IMPAcT configuration.

Typical ionization efficiency can be about 50% in these setups.

 

[Javier Lopez]

Ok, I come back again.
Thank you for the Vasimr sugestion, I were looking for a system similar to mine. My ionizer is closer to a plasma thruster due I use higher plasma density than a particle accelerator.
Is this the vasimir setup?:

My idea was to heat directly the plasma using a magnetron at 2.4Ghz. I know a japonesse scientist that uses the helicon antenna so I can speak to him to ask about. I like from helicons that I can work at few MHz and high power (I can have 40MHz and 50kW :) )

I do not understand because I think that the efficiency must be lower than 50% because thrust is done by heating the plasma 100eV instead ionizing using 10eV and then accelerate using electrostatics.
I made some tests using Lorentz but I obtained efficiencies under 14% so I am not happy with heating fuel after that

 

[trurle]

Ok, I come back again.
Thank you for the Vasimr sugestion, I were looking for a system similar to mine. My ionizer is closer to a plasma thruster due I use higher plasma density than a particle accelerator.
Is this the vasimir setup?:

Of course. Because you attached wikipedia illustration from VASIMR page.

My idea was to heat directly the plasma using a magnetron at 2.4Ghz. I know a japonesse scientist that uses the helicon antenna so I can speak to him to ask about. I like from helicons that I can work at few MHz and high power (I can have 40MHz and 50kW :) )
I do not understand because I think that the efficiency must be lower than 50% because thrust is done by heating the plasma 100eV instead ionizing using 10eV and then accelerate using electrostatics.
I made some tests using Lorentz but I obtained efficiencies under 14% so I am not happy with heating fuel after that

If you use one-step heater, your efficiency may be up to 65% which is simply nozzle efficiency. VASIMR loses additional 15% on collisions between counter-spiralling electrons and ions, and to generate RF from DC.

Regarding efficiency, may be you have too high frequency for double-layer helicon thruster. I do not know which simulator do you use though. In general, do not forget: simulators indicate the possible solution, not measure the solution.

 

[Javier Lopez]

OK, you convinced me, that approach seems perfect to me, I do not need the microwaves and can set the helicon outside the vacuum, great!, I have to decide now if using pulsed power or just the RF 50kW power system

Note-1: I think an ion thruster have a lot more Newtons using pulsed power than continuous using the same power
Note-2: I suppose the thrust is made due magnets compress the plasma

Do you have the calculus of the helicon antenna dimensions and inductance using frequency and plasma characteristics?

 

[trurle]

OK, you convinced me, that approach seems perfect to me, I do not need the microwaves and can set the helicon outside the vacuum, great!, I have to decide now if using pulsed power or just the RF 50kW power system

Note-1: I think an ion thruster have a lot more Newtons using pulsed power than continuous using the same power
Note-2: I suppose the thrust is made due magnets compress the plasma

Do you have the calculus of the helicon antenna dimensions and inductance using frequency and plasma characteristics?

Both statements are mostly incorrect.
1. Pulse thrusters have larger peak thrust, but its average thrust is actually lower. Prime reason why pulse thrusters are useful is what they can be build lighter and slightly higher power efficiency in the case of the limited solar panel power (i.e. pulse thrusters make sense if they draw peak power from battery).
2. Thrust (at least with helicon design) will be produced even without magnets, but with lower specific impulse (arcjet mode) and damage to the engine itself.

Regarding calculus, plasma thrusters are not standard design yet - you need to decide your own configuration and calculate/measure performance specifically for it

 

[Javier Lopez]

Well, you can extract pulse power from capacitors better than from batteries at full efficiency
(About accelerate ions using only joule heating leads in low thrust due specific energy to heat ions is very high, so is a lot more efficient to accelerate plasma using electrostatic acceleration or magnetic compression)
I calculated with impulse proportional to m*v, so if you send 10 times more mass you obtain 10 times more specific impulse.
So I made following calculations that could be wrong so I would appreciate if you check them
I made an excel table with theoretical data (not related to real rocket) with following results:
A 500kg rocket with 380kg fuel that uses 100 megajoules by using 10 pulses (at 100% efficiency) would be accelerated to 29.6km/s
Using 20 pulses would be 17km/s and 100 pulses would lead in 3.8km/s only:

 

[trurle]

Energy spent on acceleration of propellant is square of impulse carried by propellant. Spreadsheet above is in error.

 

[Javier Lopez]

Then I do not understand because in electrostatic acceleration E=nq*V that is proportional to number of charges, not the squared, where V is the accelerator voltage (of course the ionization energy must be accounted as losses)
And wasted energy from capacitors is V*I*t where I =n*q appears again, then, where the square of nomber of mass (or charges) come from?

It is important as long as I would like measure how many ions are accelerated by measuring the capacitor voltage drop