Abstract: We have measured the noise temperature of a single, sensitive superconducting NbN hot electron bolometer (HEB) mixer in a frequency range from 1.6 to 5.3 THz, using a setup with all the key components in vacuum. By analyzing the measured receiver noise temperature using a quantum noise (QN) model for HEB mixers, we confirm the effect of QN. The QN is found to be responsible for about half of the receiver noise at the highest frequency in our measurements. The beta-factor (the quantum efficiency of the HEB) obtained experimentally agrees reasonably well with the calculated value.

Abstract: The thesis reports the development of Hot Electron Bolometer (HEB) mixers for radio astronomy heterodyne receivers in THz frequency range. Part of this work is the fabrication of HEB devices, which are based on NbN or NbTiN superconducting thin films (â‰<a4>5 nm). They are integrated with wideband spiral or double-slot planar antennas. The mixer chips are incorporated into a quasi-optical receiver. The experimental part of this work focuses on the characterization of the receiver as a whole, and the HEB mixers as a part. Double side band receiver noise temperature and the IF bandwidth are reported for frequencies from 0.7 THz up to 2.6 THz. The spectrum of the direct response of HEB integrated with dierent antennas are measured using Fourier Transform Spectrometer (FTS). The effect of the bolometer size on total receiver performance and the LO power requirements is also discussed. A high-yield and reliable process for fabrication of NbN HEB mixers have been achieved. Over 100 devices with different bolometer geometry, film property and also different antennas have been fabricated and measured. The measured data enables us to discuss the impact of different parameters to the receiver overall performance.
This work has provided NbN HEB mixers to the following receivers:
TREND (Terahertz REceiver with NbN HEB Device) operating at 1.25-1.5 THz, installed in AST/RO Submillimeter Wave Telescope, Amundsen/Scott South Pole Station, in 2002-2003.
Band 6-low (1.410-1.700 THz) and 6-high (1.700-1.920 THz) of the HIFI (Heterodyne Instrument for Far Infra-red) in the Herschel Space Observatory, due to launch in 2007 by ESA (European Space Agency).
Besides, there has been continuous efforts to develop better models to explain the mixer performance more accurately. They are based on two temperature model for electrons and phonons and solving one-dimensional heat balance equations along the bolometer. The principles of these models are illustrated and the calculated results are compared with measured data.

Abstract: Microcalorimeters and bolometers are noise-limited devices, therefore, a proper understanding of all noise sources is essential to predict and interpret their performance. In this paper, I review the fundamental noise processes contributing to Transition Edge Sensor (TES) microcalorimeters and bolometers and their effect on device performance. In particular, I will start with a simple, monolithic device model, moving to a more complex one involving discrete components, to finally move to today's more realistic, comprehensive model. In addition to the basic noise contribution (equilibrium Johnson noise and phonon noise), TES are significantly affected by extra noise, which is commonly referred to as excess noise. Different fundamental processes have been proposed and investigated to explain the origin of this excess noise, in particular near equilibrium non-linear Johnson noise, flux-flow noise, and internal thermal fluctuation noise. Experimental evidence shows that all three processes are real and contribute, at different levels, to the TES noise, although different processes become important at different regimes. It is therefore time to discard the term “excess noise” and consider these terms part of the “fundamental noise processes” instead.