Figure 1. Illustration of the Time Tagger setup for frequency stability analysis. One input serves as the reference signal, where the ReferenceClock is enabled, while the other channels receive oscillator signals under test. The Time Tagger collects and timestamps all signals with picosecond precision, transferring the time tags to the PC, where Swabian Instruments’ software architecture enables parallel timing and frequency analyses.
Periodic electrical signals are the heartbeat of modern technology. Stable frequency sources are critical across a wide range of disciplines, from oscillators that regulate computer processors and industrial machinery to high-precision atomic clocks used in scientific and navigational applications. These oscillators come in various forms, ranging from classical crystal oscillators and MEMS devices to highly accurate chip-scale atomic clocks, each optimized for specific performance and cost constraints [1]. Ensuring synchronization across different oscillators and maintaining a common standard is essential in many fields, including telecommunications, global navigation satellite systems (GNSS), radar systems, and quantum technologies.
A key challenge in working with frequency sources is accurately characterizing and analyzing their stability over different time scales. The stability requirements vary depending on the application: some systems demand minimal phase noise over short time scales (milliseconds to seconds), while others prioritize long-term frequency stability over hours or even days. A simple standard deviation of timing errors is insufficient for capturing these fluctuations, as it fails to reveal how they evolve over time. To address this, specialized metrics for frequency stability analysis have been developed [1].
The Allan Deviation (ADEV) and its extensions are the most widely used, quantifying frequency variations over different averaging times [2]. Unlike standard deviation, which provides a single value, ADEV produces a function that describes oscillator stability across multiple time scales, making it a valuable tool for characterizing jitter, drift, and other noise sources affecting frequency signals [3]. To evaluate an oscillator’s stability, it is typically compared to a higher-quality reference clock. This reference is chosen based on the required precision, ranging from high-quality quartz oscillators to hydrogen masers and atomic clocks. Swabian Instruments’ Time Taggers simplify this process by offering a software-based phase-locked loop (PLL), allowing users to employ an arbitrary reference external clock for precision timing measurements.
The Time Tagger timestamps the rising and falling edges of oscillator signals with picosecond precision, capturing event times without predefined gate intervals. In a typical setup, multiple clock signals, such as devices under test and a stable reference oscillator, are connected to standard inputs of the Time Tagger. The ReferenceClock feature is enabled on the channel of the reference oscillator. This feature locks the time-tag stream in software to the external reference, rescaling all incoming time-tags according to the periodic clock signal to ensure accurate characterization of high-precision oscillators.
Using these high-resolution raw timestamps, the FrequencyStability measurement class improves the timing resolution even further by averaging over groups of timestamps. Based on these averaged timestamps, FrequencyStability computes key stability metrics, including ADEV, modified Allan Deviation (MDEV), Time Deviation (TDEV), and Hadamard Deviation (HDEV), providing insights into oscillator performance across different averaging times. On the other hand, the FrequencyCounter measurement class, emulating an Ω-Type Frequency Counter, reconstructs continuous frequency and phase traces, enabling real-time analysis of fluctuations.
Frequency stability varies over different timescales, from nanoseconds to days. Traditional frequency counters rely on fixed gate times, making it challengingto track fluctuations dynamically. This limitation often results in missing critical short-term jitter or long-term drift, preventing a complete understanding of an oscillator's stability.
Many frequency stability analyzers restrict users to fixed internal reference clocks or require external synchronization hardware. This limitation makes it challenging to compare multiple oscillators simultaneously or adapt to different experimental conditions. Without flexible reference clocking and multi-channel support, synchronization across numerous signals becomes cumbersome.
Many conventional frequency counters and phase noise analyzers rely on rigid, hardware-dependent software, limiting their flexibility in real-time data processing. If the software lacks adaptability, extracting meaningful insights from large datasets, automating stability analysis, and integrating with custom workflows can be challenging.
Most available tools specialize in either frequency counting, phase noise analysis, or time interval measurement, requiring multiple instruments to perform a complete stability analysis. Researchers and engineers need a single, versatile solution that integrates all these capabilities while supporting additional timing applications, such as 1PPS monitoring.
The FrequencyStability measurement class coming with the Time Tagger allows you to measure ADEV, MDEV, TDEV, HDEV, and other key stability metrics - all simultaneously. Unlike traditional frequency counters that require separate measurements for each deviation type, the Time Tagger processes multiple stability indicators in real-time, ensuring a comprehensive analysis of oscillator performance across different timescales. Additionally, all results conform to the IEEE 1139 standard, making the Time Tagger an industry-compliant solution for precision frequency stability analysis.
The Time Tagger removes the constraints of fixed gate times by continuously timestamping frequency signals with picosecond precision. This enables a gap-free, software-driven approach to frequency stability analysis, ensuring that both short-term jitter and long-term drift are captured within the same dataset. Unlike traditional counters, which force users to choose between resolution and duration, the Time Tagger enables real-time ADEV calculations at multiple averaging times from a single data acquisition.
The scalability of the Time Tagger enables highly parallelized characterization of up to 160 oscillators simultaneously, making it an ideal tool for large-scale oscillator production testing and research applications. Each oscillator is analyzed by an independent software instance, allowing for long-term monitoring over weeks on some channels while conducting short-term stability tests on others. This parallelization capability is unmatched by traditional counters, which typically analyze only one or a few oscillators at a time.
The characterization of an oscillator requires comparison with a higher-stability reference clock. The Time Tagger allows you to use any external reference oscillator of your choice, from high-quality quartz oscillators to hydrogen masers and atomic clocks. Its software-defined clocking mechanism accepts arbitrary input frequencies from 1 kHz to 700 MHz, enabling maximum flexibility in stability measurements without requiring specialized hardware modifications.
The Time Tagger provides full software-defined measurement capabilities, enabling real-time frequency stability processing and automation workflows. It is natively compatible with C++, Python, MATLAB, and LabVIEW, allowing researchers to customize data analysis, automate stability testing, and implement advanced statistical processing within their preferred environment. Unlike hardware-restricted frequency analyzers, the Time Tagger offers unmatched flexibility in integrating with modern research and industrial applications.
The Time Tagger consolidates multiple frequency and timing measurement functions into one versatile device. It supports frequency stability analysis, phase noise characterization, 1PPS monitoring and synchronization, time interval measurements, jitter analysis, and frequency counting, eliminating the need for multiple instruments. With its high-performance, all-in-one solution, the Time Tagger simplifies experimental setups and reduces system complexity, making it the ideal tool for precision timing research and oscillator characterization.
The Time Tagger software enables real-time merging of time-tag streams from remote locations, allowing distributed frequency stability analyses. By integrating various synchronization technologies, such as White Rabbit, multiple Time Taggers operate on a unified time base with picosecond-level precision, ensuring accurate cross-node oscillator characterization and phase drift monitoring. With real-time data processing on a central PC, the Time Tagger simplifies remote clock comparisons and distributed frequency metrology, making it a powerful tool for large-scale oscillator characterization and networked frequency analysis.
With the FrequencyCounter measurement class, the Time Tagger functions as an Ω-type frequency counter, capable of efficiently recording phase and frequency time traces. Unlike conventional frequency counters that rely on fixed gate times, the Time Tagger dynamically timestamps each event with high precision. Its sampling resolution is limited only by the clock period of the specific Time Tagger model, which is on the order of nanoseconds.
Figure 2. (a) Modified Allan Deviation (MDEV) of three 10 MHz oscillator signals, evaluated using the FrequencyStability measurement class of the Time Tagger software. The dashed line represents the stability floor of the Time Tagger X (very similar to the one of the Time Tagger Ultra), indicating its noise limit. (b) The phase evolution of each oscillator relative to the reference of the last 100 seconds of the measurement, extracted using FrequencyStabilityData.getTracePhase. c) Frequency measurement results obtained using the FrequencyCounter measurement class for 10 seconds. The Time Tagger continuously timestamps the 10 MHz oscillator signals with a sampling interval of 1 ms and a fitting window of 0.1 ms, allowing tracking of frequency variations over time.
[1]W.J. Riley, A. D. Howe, "Handbook of frequency stability analysis." (2008).
[2] D. W. Allan, "Statistics of atomic frequency standards." Proc. IEEE (1966).
[3] E. Rubiola, "Phase Noise and Frequency Stability in Oscillators," Cambridge Univ. Press, 2008.
Last week, our team met with the Mayor of Stuttgart, Dr. Nopper, and former Minister-President of Baden-Württemberg Mr. Oettinger, at the Center for Applied Quantum Technology (ZAQ) in Stuttgart.
Read more