Chip-Scale Precision Timing Unit for PicoSatellites

This project intends to develop a chip-scale timing unit that offers an order of magnitude higher performance compared to existing solutions. Current quartz-based clocks are not ideally suited to the high temperatures and extreme acceleration typical of space applications. The chip-scale precision clocks developed under this research offer reduced thermal sensitivity, and susceptibility to shock and acceleration using an array of micro-mechanical resonators with different temperature coefficient of frequencies. The resonators are passively compensated in a broad temperature range and offer a high quality factor and a small motional impedance, all characteristics required for achieving low-phase noise and low-power clocks. The resonators are placed in an array, and the frequency is estimated based on the weighted average output of the resonators with an accuracy that is at least an order of magnitude better than that of a single resonator clock. This research results in more precise mechanical clocks operating in harsh environment. The main objective of Year 1 was to demonstrate stand-alone high-Q micro-mechanical resonators with different TCF values, which are the building blocks of the proposed multi-resonator clocks. In Year 2, a prototype multi-resonator clock was demonstrated using PCB-based control circuit and off-chip components that showed a marked frequency stability with temperature (20 × better than the individual temperature compensated oscillators). In Year 3, we focused our effort on implementing the readout circuit to realize a complete chip-scale clock. In the extension period, we aimed at improving the temperature stability by exciting both modes of resonance in a single resonating body. 

Significant technical accomplishments of this research in each year include: 

Year 1 

• Demonstration of electrostatically actuated Lamé-mode resonators with Qs more than 500,000 and f×Q products better than 1.23 × 1013. The f×Q of the resonators demonstrated in Year 1 is one of the highest ever reported for silicon resonators. The f×Q product is the most important metric for micromechanical resonators. 
• Development of a passive temperature compensation technique using oxide-refilled trenches and demonstration of high-Q resonators with different turnover temperatures as required for the final multi-resonator clock. 
• Optimization of the anchor geometry using finite element model (FEM) to achieve the ultimate limit of Q for Lamé-mode resonators. The design of anchor geometry should be individually optimized for each resonator in the array as they exhibit different modulus when temperature compensated. The FEM model was used to design high-Q temperature-compensated Lamé mode resonators for timing applications. 
• Demonstration of compensated resonators having an f×Q of 0.36 x 1013, which is also one of the highest reported to date for passively temperature-compensated silicon resonators. 

Year 2 

• First time demonstration of a multi-resonator clock consisting of three ring resonators with different frequency-temperature characteristics.
• Demonstration of a novel passive compensation technique applied to ring resonators that provides minimum Q degradation and careful control over turnover temperature. We filed for a patent on this temperature compensation technique. 
• Development of an algorithm to enable temperature stable multi-resonator clock utilizing multiple temperature compensated resonators with different turnover temperatures. 
• Demonstration of a temperature compensated clock with total frequency shift of ±4 ppm measured across -20 °C to +60 °C in a proof of concept demonstration of the system. Compared with the standalone oscillators, the clock system shows a 20× reduction in its temperature sensitivity (this already meets the 16× reduction noted in the original proposal). 

Year 3

• Demonstration of a novel PLL-based oven-control to improve the frequency stability of MEMS oscillators under external temperature variations by several orders of magnitude.
• Demonstration of prototypes based on this novel solution showing an overall frequency drift within +/-5 ppm over the entire temperature range of -40 °C to 70 °C. It can be found that, with the proposed active compensation system, the effective TCF of a MEMS oscillator has been reduced to 36 ppb/K, an almost three orders of magnitude improvement compared to an uncompensated silicon MEMS oscillator (TCF of ~-30 ppm/K). The residual frequency drift seen is due to the non-ideal characteristics of the current MEMS platform design. 
• Demonstration of the effect of charge redistribution loss, the dominant loss mechanisms in piezo driven resonators. Extension Period 
• Demonstration of high quality acoustically engineered phonon-trap devices with dual-mode operation and a TCF differential of +17 ppm/K at its ideal operating temperature. 
• Demonstration of an oscillator using one of the two modes of the dual-mode resonator.