A compilation of time-of-flight spectra for the key ion species at 25 keV/charge (full post-acceleration for low energy ions) is shown in Fig. 4. Because the major species in the magnetosphere, ionosphere, and solar wind are separated by factors of 2 in mass/charge a clear identification of these ions even with poorer resolution, i.e. for reduced post-acceleration voltages is possible. As can also be seen, ions adjacent on the mass scale, such as N and O, cannot be easily distinguished. However, this was never a goal for an instrument designed to determine the velocity distribution function for the major species that contribute significantly to the dynamics of the plasmas in the diverse regions of the magnetosphere and the solar wind. The goal has been more than met with the present design.
The calibration of the sensor efficiency for all species and the complete energy range provided another challenge. While the response of an ESA remains essentially unchanged for different species and over the entire energy range, the detection efficiency of TOF assemblies changes with species and energy due to dependencies of the secondary electron production and the sensitivity of the MCPs. To reduce the anticipated variations the MCPs are operated in saturation of their gain. Yet there are still significant variations of the efficiency with species, energy and position on the MCPs that may be as large a factor 2. As a result each sensor has been calibrated in a matrix of ion species, energies and MCP position. The variation of the efficiency as a function of energy for a typical sensor model is shown in Fig. 5 for H + , He + , and O + .
The first scans of the sensor in the electrostatic analyzer voltage and the azimuthal angle (in the direction of the ESA deflection) had suggested that the TOF section might even introduce a variation of the response in these parameters. Instead of showing the theoretical ESA response, unanticipated reductions of the detection efficiency were observed as a function of ESA voltage or acceptance angle. However, it was found that the focusing of the incoming ion beam by the post-acceleration field at the TOF entrance is so narrow that the support grid structure becomes visible. This effect is not visible for an ion distribution sufficiently wide in angle and energy as can be assumed for realistic conditions. This could be verified by a combined angle and analyzer voltage scan, as shown in Fig. 6. The variation of the response is shown as a function of ESA voltage (in a) and azimuthal angle (in b). In both cases the response has been integrated over azimuthal angle (in a)) and ESA voltage (in b)), respectively. Now the grid structure becomes invisible and the curves closely match the theoretical response of the ESA as has been confirmed in separate calibrations of the ESA assemblies. The remaining effect of the carbon foil support grids is a reduction of the detected flux according to the transmission of the grids that has also been calibrated. The calibration effort for the RPA of this instrument may be found in a paper by McCarthy and McFadden (1996) in this volume.