Temperature compensation of crystal oscillators using an Artificial Neural Network

Temperature Compensated Crystal Oscillators (TCXOs) are widely used and well known frequency control products. Their performance has improved over the decades with the advent of newer and improved technologies. Evolution of the TCXO from resistor thermistor networks to modern polynomial generators has pushed TCXO temperature stabilities to nearly +/-100ppb deviation over the industrial temperature range of -40 to +85°C. Even with these great advances, users always need tighter stabilities. This paper focuses on a new temperature compensation technique for crystal oscillators. Through the use of an Artificial Neural Network (ANN), temperature compensation of AT cut crystal oscillators can be achieved with better than +/-10ppb stability over the industrial temperature range (-40 to +85 °C). This is more than a 10 fold improvement over state of the art polynomial function generator compensation. The theory of this compensation method will be discussed, and data showing the results of temperature compensation on actual oscillators will be presented.