Bulk acoustic wave resonators and oscillators at liquid helium temperatures

This thesis presents the results of investigation of bulk-acoustic-wave (BAW) resonators working at cryogenic temperatures (over 3K-15K, in a pulse-tube cryocooler) along with the systems built with these devices. The first aspect of the work is dedicated to the operation of different devices, and particularly the BAW resonators. Their behavior is more systematically studied over a frequency range of 1−90 MHz. Quality factors of 417 * 106 and Q* f-products (Q-factor at frequency f) up to 3.07 * 1016 Hz have been measured for quartz resonators, outstanding values that are word records of this class of devices. It is shown that the intrinsic Q-factor does not depend of the frequency f, at these temperatures, in agreement with the Landau-Rumer theory. Problems and advantages of operating at such temperatures are assessed. Related limitations are discussed. Other passive and active electronic devices, such as transistors, are also under scope of this work. The choice of appropriate components is made based on comparison of their behavior at 4K. The results are confirmed with a successful design of liquid helium amplifiers. The second aspect of the thesis is the modeling and simulation of the studied devices and systems. The report presents a rigorous model of the BAW device phase noise taking into account device nonlinearities based on the averaging approach. This MIMO (multiple-input and multiple-output) model is used to derive impacts of different oscillator parameters to its phase noise and to explain experimental results. Accurate models of other electronics devices working at liquid helium temperatures are also derived. They are used for simulation and optimization of cryogenic, oscillators, etc. The third aspect concerns frequency sources based on these cryogenic BAW resonators: a cryogenic passive closed loop reference system and an oscillator. The implemented systems allowed measuring and characterizing the resonator phase noise at the given conditions. Thus, some noise sources have been identified. The tested feedback stabilization systems achieve fractional frequency stability of 4 * 10−13 at 100 seconds and have stability better than 10−12 between 1 and 2000 seconds. The implemented oscillator exhibits frequency stability of 1.5*10−12 at 200 seconds and better than 10−11 for averaging times greater than 80 ms. Limitations of these systems are discussed based on the obtained data.