[en] Full text: The cesium-sputter negative-ion source is a fundamental component of accelerator mass spectrometry studies. Improvements in ion current and output emittance allow for higher precision measurements and potential for new applications. The High Voltage Engineering Europa SO-110C ion source has been modeled using Integrated Engineering Software’s Lorentz 2E ion trajectory simulation software. This software includes the mutual space charge interactions between the incoming cesium ions from their production in the ionizer and the outgoing negative ion beam from the target (cathode). Simulations examined the effects of changing the geometry of various source components, electrode potentials, ion currents, position of the target and negative ion mass. The simulations demonstrated that, as the cesium current is increased, the cesium beam broadens by its own space charge repulsion. This affects its focusing and distribution on the sample material. It is important that the cesium cover the outer proportions of the sample well for best usage of the material and stability of the outgoing negative ion beam. Changes in geometry and electrode potentials can mitigate this effect. Informed by these simulations, experiments at the A.E. Lalonde Accelerator Mass Spectrometry Laboratory (Lalonde) recessed the target in incremental 1 mm steps (targets were translated away from the ionizer along the axis of symmetry). Custom target bases were machined to facilitate these recesses without the need to modify the target pressing procedure. The abundant isotopes were measured using the post-accelerator offset Faraday Cups and rare isotopes at the gas ionization detector to compare outputs at various settings. These tests were first run at routine ¹⁴C measurement settings (6 kV target-ionizer potential difference, 115°C cesium oven temperature) on graphite blanks. At these settings, a target recess of 1 mm gave the most stable output with the highest usage of sample material. A second experiment, using targets made from a ¹⁰Be standard, expanded this study by increasing the cesium oven temperature (and hence cesium current) incrementally from 130-140°C, while also varying the target-ionizer potential difference (4-11 kV). This multi-dimensional study gave several promising settings, resulting in the most precise measurement of ¹⁰Be performed to date at Lalonde. A demonstration of key simulations and a comparison with experimental results will be presented.