Feasibility study of a frequency division multiplexing read-out system with superconducting tunnel junctions as rectifiers and frequency mixers

Abstract

Nowadays many studies in astrophysics focus on the detection of far-infrared and submillimeter radiation. The most sensitive detectors for these wavelengths are superconducting detectors that have to be cooled to temperatures of order 100 mK. Large scale imaging or spectroscopy require large arrays of detectors (e.g. Transition Edge Sensors) containing hundreds or thousands of elements. In the last decade the size of the array increased to several 100 pixels, however, a jump to an array containing thousands of pixels requires a fundamental redesign of this existing technology. A serious limitation in the size of current cryogenic TES detector arrays is the large wire count from the room temperature electronics to the detectors at cryogenic temperatures. This gives rise to a lot of engineering problems such as electromagnetic interferences and thermal issues, especially for space instrumentation where the cooling power at low temperatures is very limited. In this thesis we study the feasibility of a cryogenic , low power, frequency multiplexed read-out scheme for a large number of Transition Edge Sensor (TES) detectors. This is done with help of Superconducting-Isolator-Superconducting (SIS) devices operating at GHz frequencies and working as frequency up- and down- converters and as RF-to- DC converters together with a Superconducting Quantum Interference Device (SQUID) for simultaneous local read-out. Future implementations of this design would allow the read-out and control of thousands of pixels with one single coaxial cable instead of thousand separate wires. For demonstration of the SIS-based RF-to-DC conversion and down-conversion, three di erent Printed Circuit Boards with commercial discrete components and a superconducting multiplexing frequency division scheme with Niobium microstrip lines were simulated, produced and tested. The results were very satisfying. The SIS devices have shown an excellent performance when acting as RF-to-DC converter at 2.5 GHz and the initial results of our bias system for control of a feedback coil of a SQUID were better than we could have expected, although improvements in the noise measurements could easily be done in future iterations. Furthermore the performance of the SIS as a GHz to MHz mixer was shown and nally a multiplexing system with a superconducting Nb microstrip line lter bank working around 4-6 GHz showed a performance close to simulations and demonstrated a next step in possible miniaturisation of the superconducting multiplexing electronics. These results give a proof of principle and indicate that with some improvements in the following iterations we can get a whole independent working scheme that could possibly be included in future astrophysical projects.