The PEGASES thruster is a new plasma thruster developed at LPP. In a classical electric thruster, thrust is produced by positive ions accelerated from an electropositive plasma. The novelty introduced by PEGASES is that thrust is provided by successive acceleration of positive and negative ions from a continuous dual ion source. The PEGASES thruster can be divided into three stages: (i) the ionization stage consisting in an inductively coupled plasma source to generate an electronegative plasma from halogen-contained gases, (ii) the magnetic electron filtering stage to obtain an ion-ion plasma, and (iii) the extraction and acceleration stage using alternately biased grids to create an accelerating electrostatic field with time-varying direction in order to accelerate both positive and negative ions. During this thesis I worked mainly experimentally on the first and third stages of two PEGASES prototypes. An ion-ion plasma was obtained in the first prototype using SF6 and a strong magnetic filter. But the first prototype had some inherent limitations. We therefore moved to a planar ICP source for the second prototype, using a transformer coupled matching network and a ferrite core to obtain an enhanced efficiency. The quest for power transfer efficiency led to use a low-loss step-down transformer. A reduction in the stray capacitive coupling was obtained by improving the impedance matching network and investigated via plasma potential oscillations measurement. I used Langmuir Probes and a Retarding Field Energy Analyser (RFEA) to investigate the plasma in the two prototypes. I showed that the potential of an ion-ion plasma could be controlled with a biased electrode in contact with the plasma. Then I investigated the ion acceleration from the ion-ion plasma with continuous grid bias and with square voltage waveforms at 1 kHz and acceleration voltages in the 0 - ± 350 V range. I showed that the positive and negative ions are accelerated in the positive and negative bias half-periods, respectively, and that the respective beam energies can be controlled independently. With these results is provided the first proof-of-concept, and PEGASES is moving from concept towards realization.
