Short-Term Frequency Stability of Precision Oscillators and Frequency Generators

When the phrase "short-term frequency stability" is mentioned, one can immediately infer that frequency varies with time and that a short time interval is involved. Many questions are, at this point, unanswered: How short is the interval? How is stability defined? What system malfunctions can result from excessive instability? What definitions are appropriate for the specific application and why? Some answers to these questions are developed here; others must be determined by the user in terms of the specific system involved. If a measurement of short-term frequency stability is to be of significant use, each of these questions must be resolved. Although ever-increasing interest in short-term stability has existed for 25 or 30 years, no universally accepted definition of short-term stability exists. Primarily, early research in this area was extremely application-oriented. Little general work was performed with the result that many definitions were developed. Individuals and organizations defined and used stability in terms of their own applications. The resultant confusion showed that a more general definition applicable to the majority of cases was desirable in the topic of short-term frequency stability. Today, many users talk about stable oscillators without understanding why or even if such precision is necessary. Manufacturers add to the present confusion through lack of rigor in their performance specifications. That is, they often merely quote a number without defining what, how, or why they are measuring the stability of their equipment.