It is now possible to develop a new class of receivers for the millimeter and shorter wavelength region by application of Josephson and other electron tunneling phenomena which occur in superconducting materials. This development represents a major advance in receiver sensitivity and will be applicable over a wide frequency interval. Operating in the conventional heterodyne detection mode this receiver has a theoretical noise temperature as low as 20 Kelvin at 100 GHz. Three basic components required are: a superconductor-semiconductor Schottky diode mixer; a Josephson junction phase-locked multiplier local oscillator; and a Josephson parametric amplifier. The component and system design and limiting factors will be discussed.
It is now possible to develop a new class of receivers for the millimeter and shorter wavelength region by application of Josephson and other electron tunneling phenomena which occur in superconducting materials. This development represents a major advance in receiver sensitivity and will be applicable over a wide frequency interval. Operating in the conventional heterodyne detection mode this receiver has a theoretical noise temperature as low as 20 Kelvin at 100 GHz. Three basic components required are: a superconductor-semiconductor Schottky diode mixer; a Josephson junction phase-locked multiplier local oscillator; and a Josephson parametric amplifier. The component and system design and limiting factors will be discussed.
The progress made in relation to the development of sensitive submillimeter heterodyne receiver systems is reviewed. Sensitive superconductor-insulator-superconductor (SIS) tunnel junction mixers were demonstrated at frequencies of greater than 1 THz, with noise temperatures of below 1000 K. An alternative mixer technology is that of the superconducting hot electron bolometer, for which sensitivities of 1.2 THz and 2.5 THz were demonstrated. Local oscillator technology is under development. Gunn oscillators followed by Schottky varactor frequency multipliers have sufficient output power to drive SIS mixers at 1 THz. It is considered that optical photomixer local oscillators have potential for driving hot electron bolometers above 1 THz.
This chapter discusses the high sensitivity superconducting mixers, i.e. detectors used in heterodyne receivers. Today, there are essentially two very successful concepts of superconducting mixers, or frequency down-converters, the superconductor-insulator-superconductor tunnel junction SIS mixer and the superconducting hot-electron bolometer mixer. Moving up to higher frequencies in the microwave and millimeter wave electromagnetic spectra makes it more and more difficult to retrieve a good signal-to-noise ratio with ordinary solid-state amplifiers. The driving force for using superconducting mixers is their excellent down-conversion efficiency and capability to down-convert RF signals from microwave frequencies up to the terahertz region. The main purpose of a receiver is to pick up and amplify a weak signal, which can then be accurately determined in amplitude and phase. The mixer noise temperature and conversion gain are the most important parameters for the mixing device. The Josephson junction mixer was one of the first suggested devices based on superconducting point contacts and later tunnel junctions.
Josephson junctions are highly nonlinear and respond up to infrared frequencies. A wide variety of high frequency applications have been explored. In addition, several high frequency superconducting devices are being developed which do not use the Josephson effect. These include a superconducting version of the Schottky barrier diode for use as a heterodyne mixer and a video detector, and also a composite superconducting bolometer. In this review, the state of development of superconducting mixers, detectors, and parametric amplifiers at microwave, millimeter, and submillimeter wavelengths is discussed. Emphasis will be placed on developments which have taken place since 1974 when a previous review was circulated.
As radioastronomers continue to push their observations to shorter wavelengths and fainter signals, their progress is impeded by a lack of sufficiently sensitive millimeter-wave receivers capable of operating at frequencies above 100 GHz. But now a spinoff from Josephson-junction fabrication technology appears to be coming to the rescue. New metallurgical and lithographic techniques developed for Josephson-effect computers are permitting the construction of superconducting diode junctions that promise to function as high-frequency photon detectors and heterodyne mixers near the theoretical limits of sensitivity.
It has been shown both theoretically [1] and experimentally [2,3] that the very sharp nonlinearity in the conductance of superconducting tunnel junctions can be used to implement millimeter wavelength mixers having high gain and very low noise, leading to the possibility of receivers that achieve the quantum noise limit. The junctions involved contain a thin insulating barrier between two superconductors, and are thus known as superconductor-insulator-superconductor (SIS) junctions. They are physically identical to Josephson junctions, but in this application the conductance is via the tunneling of quasiparticles (unpaired electrons), and so the Josephson effect is not involved.
Heterodyne receivers based on superconductor-insulator-superconductor (SIS) tunnel diodes are the most sensitive available for near-millimeter wavelengths. Since the late seventies, receivers based on SISs have been used for millimeter band observations at radio observatories around the world. The work described here was carried out with the elusive goal in mind of developing ideal SIS receivers for radioastronomy. This thesis describes one researcher's path towards this goal. In the first chapter, a basic description of SIS diodes, their interaction with radiation, and heterodyne detection are given. The important detailed results of J. R. Tucker's tunnel diode heterodyne theory are described, since subsequent chapters rely on Tucker's work quite heavily. Throughout this introductory chapter, extremely simple (by comparison with the algebraic theoretical results) physical models are used to describe superconductivity, SISs, photon-assisted tunneling, and heterodyne detection. It is the author's experience that these physical models can be used to derive correct theoretical results, long before these results are proved rigorously. The second chapter presents a fully quantum mechanical theory of heterodyne detection with diodes. This theory was developed because Tucker's theory for tunnel diodes predicts a greater mixer sensitivity than is possible considering Heisenberg's uncertainty principle for radiation. Tucker does not quantize the radiation incident on the SIS, although his treatment of the isolated tunnel diode is completely quantum mechanical. In chapter 2, the quantization of radiation is carried out for heterodyne diode detectors. The formalism is shown to obey quantum limits on sensitivity. Finally, an ideal SIS mixer is shown to have noise properties identical to those of optical mixers based on ideal photodiodes. In possession of an apparently complete theory for SIS mixers, the third chapter presents a sampling of numerical results from that theory. Four different non-ideal tunnel diodes are used for these calculations so that a quantitative feel for the importance of diode quality can be achieved. The effects of dc and LO bias, signal and image source admittance, frequency of operation, and junction quality are all explored. This information will be useful for the proper engineering of SIS mixers. Finally, the fully optimized performance of the four tunnel diodes is presented as frequency is varied. It is shown that reasonably good quality lead-alloy SISs should behave like photodiodes up to frequencies as high as 1500 GHz. Finally, chapter 4 presents a prototype open-structure SIS mixer. Measurements in the laboratory show this mixer to be quite sensitive for signal frequencies from 115 to 761 GHz. Unlike all other SIS receivers, in which the diode is mounted across a waveguide, this mixer relies on the bowtie-on-quartz antenna structure, which has been investigated by D. B. Rutledge and his students. This difference is essential to the multi-octave spectral coverage of this mixer. It is probable that waveguide designs will never achieve good results above 500 GHz, and as of now, there are no SIS-waveguide mixers which operate well above 300 GHz. Tests at the Owens Valley Radio Observatory verify the suitablility of this mixer for radioastronomy, but these tests have been limited to frequencies of 260 GHz and lower.