As far as I know a thruster also includes a magnetic field allowing for a separate deflection of positive and negative ions currents which can thus  be separately collected

You can read the Following Article "Langmuir Probe Measurements in Electronegative RF Plasma" I will send you.

You are right Claude, I collect negative ions but along all the tube so I can not measure at specific place. This is important to me to optimize plasma acceleration.

About the Langmuir probe perhaps I should use a small magnetic field to separate positive and negative ions and collect them in different plates.

There are two main ways to measure plasma current: either the magnetic field created by the current itself, or with Langmuir probes (as suggested by Tayeb). Usually with Langmuir probes in fusion plasmas you distinguish an "ion side" and an "electron side", since ion and electron currents have the same  orientation, and opposite directions. Usually it is more difficult to measure electron currents, since electrons carry more energy and can damage the probe. I am not expert with thrusters, so in that case maybe the energy carried by electrons is not too much.

In plasma thrusters it is tried to reduce the electron (or negative ions) current and increase the positive ones. I do not know how many energy will came with negative ions and electrons (I should use a magnetic field and separate ion collectors). Perhaps in order to know the current it is better using coils, as in the following drawing, or use the Langmuir probe as you said but connected a current to voltage converter (or simply a resistor).

Accordingly my experience the use of a coil does not work well this devices due giant spikes.

Laser Photo-detachment with a Langmuir probe in principle can be used for measurement of positive ion density and negative ion density. For positive ion density use ion-saturation current in standard Langmuir probe configuration and for -ve ion density launch the laser pulse when probe power-supply ramp reaches in the electron saturation. Laser pulse photo-detaches extra electron from the negative ion and a small jump in the electron saturation current will be registered, which is equivalent to negative ion density. For repeated measurement, make the probe circuit ramp frequency and the laser pulse repeatation rate  same.

You can look at the doppler shift of a spectral line of the ions. If you image the spectrometer output you can also get one spatial dimension. I have seen this done in Argon plasmas using a 2 metre spectrometer fitted with an echelle grating. You'll need to do a few calculations to see if the doppler shift is resolvable in your plasmas. There is also the question of whether the plasma is suitably transparent.

The Doppler method is a lot less intrusive and gives the speed of the particles that is right, but perhaps may not give good for accuracy measurement of the ion current because negative ion emissions must be higher. I have made some calculus for the 656nm emission of hydrogen and for speeds of 10km/s and 17º degrees of the observation line I obtain only 20nm shift

I have added the Doppler formula in the next excel table to be uploaded to RG: "Useful formula and excel tables for plasma physics V06.xls", formula number 96:Ion speed measurement by using Doppler shift

I added also formula 97 to calculate the diffraction by using a diffraction grating.

How fully ionised is your plasma? I know the charge exchange cross section between neutral hydrogen and hydrogen +ve ions is very high. You may measure a current this way. I was thinking of a heavier ion where you would measure the velocity more directly. I know little about plasma thrusters, I had thought xenon was the gas of choice.

20nm seems like an easy to measure shift. Of course you can measure spectral lines of both positive and negative ions independently.

I could not find the link to the file/formula you mentioned.

Plasma must be ionized in the 25-75% range,  I avoid negative charge current by using electron traps, now I designed a simulator that said I could reach 85% electric efficiency.

>I could not find the link to the file/formula you mentioned.

I have not uploaded the document yet, as long as there are very few new formulas to generate a new version, but I will place here for you, I would be pleased if you could check it:

I have not yet gone through your calculations but I spotted that you set the diffraction grating line density to 2500/cm whereas one can buy 2400/mm. Spectral resolution to 6pm is achievable. What is more interesting than spectral resolution is how small a shift on a wide line one can resolve.  The technique I have seen used imaged the line onto the edge of a reflecting prism. This generated two signals that would be equal if the line centre is unshifted. The non equality then gives the line shift. Possibly this can be done on a solid state camera nowadays using software to fit line profiles and look for the shift.