Comparative study of TiAu-based TES microcalorimeters with different geometries H.F.C. Hoevers, M.P. Bruijn, B.P.F. Dirks, L. Gottardi, P.A.J. de Korte, J. van der Kuur, A.M. Popescu, M.L. Ridder, Y. Takei, D.H.J. Takken SRON Netherlands Institute for Space Research, Sorbonnelaan 2, 3584 CA Utrecht, the Netherlands Seven microcalorimeters with different geometries have been tested and their performance is compared. The study, for TiAu TESs with a Cu absorber, indicates the presence of so-called constant voltage noise and internal thermal fluctuation noise. We find no indication that the constant voltage noise can be manipulated through a normal metal pattern on the TES, nor by a magnetic field. The energy resolution of the detectors, having different heat capacities, is between 2.5 and 5.0 eV (at 5.9 keV). PACS numbers: 07.20.Fw, 85.25.Am, 85.25.Oj, 95.55.-n. 1. INTRODUCTION Arrays of cryogenic microcalorimeters, based on superconducting Transition Edge Sensors (TESs), are being developed for high resolution X-ray spectroscopy and are foreseen as detectors in future satellites in the space programmers of ESA (Europe), NASA (USA) and JAXA (Japan). The ultimate performance of these detectors in terms of energy resolving power is, however, still limited by unexplained noise (UN). The aim of this study is a critical redesign of TES-based microcalorimeters and a further optimization of their performance. In the redesign the absorber will completely overlap the TES. In this way the stress concentrations, originating from the absorber, are located outside the TES and should not degrade the steepness of the superconducting transition. An insulator separates absorber and TES, apart from a well defined metallic contact area. As the absorber overlaps the TES, only a small overhang of the absorber will be needed to achieve an array with a high filling factor. The study will also address the observations by NIST that the steepness α = (T/R) (dared) of the superconducting transition and UN can be manipulated by normal metal patterns on the TESs, respectively by a magnetic field 1 . Figure 1 shows the devices that are tested; Ref. 2 describes the device production in more detail. Device (a1) has a 100 x 100 μm 2 central