Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

Australian Journal of Physics

Experimental Study of Noise Temperature and Conversion Loss of SIS Junction Mixers

J.C. Macfarlane, Lewis B. Whitbourn, R. A. BATCHELOR

Published Jun 1, 1986• 2 citations
Publisher

Abstract

Experiments on the performance of superconductor-insulator-superconductor (SIS) tunnel junctions as heterodyne mixers at 40 GHz are reported. The techniques developed for the fabrication of multi-junction SIS series arrays differ from, and are considerably simpler than, other methods described in the literature. Results are reported on mixer noise temperature, conversion loss, saturation, local oscillator power levels and mixer dynamic range. Factors affecting the mixer's conversion performance are identified and comparisons are made with other reported results. It is concluded that mixer performance in the present experiments is limited by imperfect impedance matching at the signal frequency, due to relatively large values of parasitic capacitance in the junctions. Future work is aimed at both reducing the junction capacitance and redesigning the mixer mount to give improved r.f. matching.

DOI
10.1071/ph860379

Related papers

Performance of Arrays of SIS Junctions in Heterodyne Mixers

Denis Crété, William R. McGrath, P. L. Richards

We have made a systematic experimental study of the performance of millimeter-wave quasiparticle heterodyne mixers which use arrays of SIS tunnel junctions. Sets of arrays with N =1, 5, 10, 25, and 50 junctions in series were fabricated by photolithography. All of the arrays in a given set were made on a single silicon wafer so that their response time parameter omega/sub S/R/sub N/C would be the same. Junction areas were scaled so that the total impedance was the same for each array in a set. Sets of arrays from four wafers with values of omega/sub S/R/sub N/C ranging from 2.6 to 13 were evaluated in mixers at 33 and 36 GHz. These measurements showed that the signal power required to saturate the mixers varies as N/sup 2/ and that the conversion efficiency is nearly independent of N for all values of omega/sub S/R/sub N/C. The mixer noise temperature is independent of N for large values of omega/sub S/R/sub N/C. Therefore, the dynamic range of an SIS quasiparticle mixer can increase in proportion to N/sup 2/. For small values of as omega/sub S/R/sub N/C, however, the mixer noise increases systematically with N. This correlation suggests that the junction capacitance affects the coupling between junctions that can contribute to the noise.

IEEE Transactions on Microwave Theory and Techniques•Apr 1, 1987•24 citations

Low noise SIS receiver performance using Pb alloy tunnel junctions

Steven R. Davies, Leslie T. Little

A receiver noise temperature of 61 K (DSB) has been measured for a superconductor–insulator–superconductor (SIS) heterodyne receiver operating at 232 GHz, the lowest value reported to date for such a receiver employing a lead alloy junction. Analysis of the losses of the receiver components yields mixer noise temperature and conversion loss, Tm = 24 K and Lm = 2.1 dB (DSB). These are compared with predictions based on 3-port quantum mixer theory, made using the measured characteristics of the junction and an equivalent circuit for the waveguide mixer mount deduced from scaled modelling experiments.

IEE Proceedings - Microwaves Antennas and Propagation•Jan 1, 1998•0 citations

0.5-THz waveguide SIS mixers with NbN/AlN/NbN technology

Jing Li, Masanori Takeda, Zhen Wang +1 more

Heterodyne mixers based on superconducting SIS (superconductor-insulator-superconductor) tunnel junctions have been demonstrated to be the most sensitive coherent detectors at millimeter and submillimeter wavelengths. In fact, conventional superconducting SIS mixers with Nb/AlOx/Nb junction and Nb/SiO2/Nb tuning circuit have shown good performances with the noise temperature reaching as low as three times the quantum limit below 0.7THz, which is the gap frequency of Nb-based SIS junctions. However, due to the large loss in Nb thin-film superconducting microstrip lines, the noise performance of Nb SIS mixers deteriorates significantly above 0.7THz. With a gap frequency double that of Nb-based SIS junctions, NbN-based SIS junctions are of particular interest for the development of heterodyne mixers in the terahertz region. Considering the bandwidth and output power of local-oscillator (LO) signal sources are quite limited around 1THz, we firstly develop a waveguide NbN-based SIS mixer at 0.5THz. Three types of SIS junctions, i.e., long junction, parallel-connected tunnel junction (PCTJ) and distributed junction array (DJ) are investigated. They are all comprised of NbN-AlN-NbN tri-layer fabricated on an MgO substrate and have the same current density (Jc) of 10kA/cm2. In this paper, we describe their design, fabrication and preliminary experimental results.

Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE•Nov 29, 2007•3 citations

Performance of a two-junction array SIS-mixer operating around 345 GHz

Cornelia E. Honingh, Gert J. de Lange, M. M. T. M. Dierichs +3 more

The authors have made a detailed study of the gain and noise of a SIS (superconductor-insulator-superconductor) heterodyne receiver at 345 GHz. An array of two Nb-Al/sub 2/O/sub 3/-Nb SIS junctions in series are used as the mixing element. The array is operated in a waveguide mount with a backshort and an E-plane tuner. The best receiver noise temperature achieved is 140 K DSB (double sideband). The embedding impedances were determined by fitting theory to the measured pumped curves. High-quality fits were obtained, providing the first detailed test of the Tucker-theory at frequencies above 300 GHz. The impedances found by this method are in very good agreement with impedances measured in a scale model at 3.3 GHz. From these embedding impedances, the gain and noise of the mixer were calculated over a full bias range using the Tucker theory in the three-port low-IF approximation. The measured dependence of mixer gain and noise on bias voltage, pump power and embedding impedance is in good agreement with theory. However the absolute values show discrepancies that appear to be independent of the bias parameters of the mixer.>

IEEE Transactions on Microwave Theory and Techniques•Apr 1, 1993•14 citations

Performance of heterodyne mixers using arrays of SIS junctions

Denis Crété, William R. McGrath, P. L. Richards +1 more

The SIS quasiparticle tunnel junction has been demonstrated to be the most sensitive mixing element for millimeter wave receivers [1,2]. Conversion gain greater than 1 and noise temperature approaching the quantum limit [3,4] has been observed. The required local oscillator power, which can be as small as 10−8W, leads to a low saturation threshold for these mixers [5,6]. The practical benefits, including higher saturation power, of using series arrays are sufficiently important that a substantial number of experiments using arrays have been reported [2,6,7,8,9]. It has not been clear, however, whether there has been any sacrifice in mixer conversion efficiency (gain) or noise. We present a systematic experimental investigation of mixing performance in series arrays of N junctions, where N = 1, 5, 10, 25 and 50.

International Conference on Infrared, Millimeter, and Terahertz Waves (IRMMW-THz)•Dec 1, 1985•0 citations

SIS-Mixer Receiver with Single and Array Tunnel Junctions for the 140 GHZ to 170 GHZ Range

R. Blundell, J. Ibruegger, K. H. Gundlach +1 more

Mixing experiments were made with superconductor-insulator-superconductor (SIS) single tunnel junctions and arrays of junctions. The best receiver noise temperature is 96 K at 150 GHz for a single junction and 67 K at 139.5 GHz for a series array of two junctions. The IF embedding impedance seen by the junction has been measured and its influence on the noise temperature variation over the IF band is shown.

European Microwave Conference (EuMC)•Oct 1, 1984•3 citations
Experiments on the performance of superconductor-insulator-superconductor (SIS) tunnel junctions as heterodyne mixers at 40 GHz are reported. The techniques developed for the fabrication of multi-junction SIS series arrays differ from, and are considerably simpler than, other methods described in the literature. Results are reported on mixer noise temperature, conversion loss, saturation, local oscillator power levels and mixer dynamic range. Factors affecting the mixer's conversion performance are identified and comparisons are made with other reported results. It is concluded that mixer performance in the present experiments is limited by imperfect impedance matching at the signal frequency, due to relatively large values of parasitic capacitance in the junctions. Future work is aimed at both reducing the junction capacitance and redesigning the mixer mount to give improved r.f. matching.
We have made a systematic experimental study of the performance of millimeter-wave quasiparticle heterodyne mixers which use arrays of SIS tunnel junctions. Sets of arrays with N =1, 5, 10, 25, and 50 junctions in series were fabricated by photolithography. All of the arrays in a given set were made on a single silicon wafer so that their response time parameter omega/sub S/R/sub N/C would be the same. Junction areas were scaled so that the total impedance was the same for each array in a set. Sets of arrays from four wafers with values of omega/sub S/R/sub N/C ranging from 2.6 to 13 were evaluated in mixers at 33 and 36 GHz. These measurements showed that the signal power required to saturate the mixers varies as N/sup 2/ and that the conversion efficiency is nearly independent of N for all values of omega/sub S/R/sub N/C. The mixer noise temperature is independent of N for large values of omega/sub S/R/sub N/C. Therefore, the dynamic range of an SIS quasiparticle mixer can increase in proportion to N/sup 2/. For small values of as omega/sub S/R/sub N/C, however, the mixer noise increases systematically with N. This correlation suggests that the junction capacitance affects the coupling between junctions that can contribute to the noise.
A receiver noise temperature of 61 K (DSB) has been measured for a superconductor–insulator–superconductor (SIS) heterodyne receiver operating at 232 GHz, the lowest value reported to date for such a receiver employing a lead alloy junction. Analysis of the losses of the receiver components yields mixer noise temperature and conversion loss, Tm = 24 K and Lm = 2.1 dB (DSB). These are compared with predictions based on 3-port quantum mixer theory, made using the measured characteristics of the junction and an equivalent circuit for the waveguide mixer mount deduced from scaled modelling experiments.
Heterodyne mixers based on superconducting SIS (superconductor-insulator-superconductor) tunnel junctions have been demonstrated to be the most sensitive coherent detectors at millimeter and submillimeter wavelengths. In fact, conventional superconducting SIS mixers with Nb/AlOx/Nb junction and Nb/SiO2/Nb tuning circuit have shown good performances with the noise temperature reaching as low as three times the quantum limit below 0.7THz, which is the gap frequency of Nb-based SIS junctions. However, due to the large loss in Nb thin-film superconducting microstrip lines, the noise performance of Nb SIS mixers deteriorates significantly above 0.7THz. With a gap frequency double that of Nb-based SIS junctions, NbN-based SIS junctions are of particular interest for the development of heterodyne mixers in the terahertz region. Considering the bandwidth and output power of local-oscillator (LO) signal sources are quite limited around 1THz, we firstly develop a waveguide NbN-based SIS mixer at 0.5THz. Three types of SIS junctions, i.e., long junction, parallel-connected tunnel junction (PCTJ) and distributed junction array (DJ) are investigated. They are all comprised of NbN-AlN-NbN tri-layer fabricated on an MgO substrate and have the same current density (Jc) of 10kA/cm2. In this paper, we describe their design, fabrication and preliminary experimental results.
The authors have made a detailed study of the gain and noise of a SIS (superconductor-insulator-superconductor) heterodyne receiver at 345 GHz. An array of two Nb-Al/sub 2/O/sub 3/-Nb SIS junctions in series are used as the mixing element. The array is operated in a waveguide mount with a backshort and an E-plane tuner. The best receiver noise temperature achieved is 140 K DSB (double sideband). The embedding impedances were determined by fitting theory to the measured pumped curves. High-quality fits were obtained, providing the first detailed test of the Tucker-theory at frequencies above 300 GHz. The impedances found by this method are in very good agreement with impedances measured in a scale model at 3.3 GHz. From these embedding impedances, the gain and noise of the mixer were calculated over a full bias range using the Tucker theory in the three-port low-IF approximation. The measured dependence of mixer gain and noise on bias voltage, pump power and embedding impedance is in good agreement with theory. However the absolute values show discrepancies that appear to be independent of the bias parameters of the mixer.>
The SIS quasiparticle tunnel junction has been demonstrated to be the most sensitive mixing element for millimeter wave receivers [1,2]. Conversion gain greater than 1 and noise temperature approaching the quantum limit [3,4] has been observed. The required local oscillator power, which can be as small as 10−8W, leads to a low saturation threshold for these mixers [5,6]. The practical benefits, including higher saturation power, of using series arrays are sufficiently important that a substantial number of experiments using arrays have been reported [2,6,7,8,9]. It has not been clear, however, whether there has been any sacrifice in mixer conversion efficiency (gain) or noise. We present a systematic experimental investigation of mixing performance in series arrays of N junctions, where N = 1, 5, 10, 25 and 50.