Application Cases

Experiment Name: Pulse Repetition Rate Synchronization Experiment of High-Repetition-Rate Mode-Locked Fiber Lasers

Research Direction:
High-repetition-rate pulses possess short pulse intervals, making them highly valuable in high-speed processing networks. However, when lasers operate freely, environmental disturbances such as mechanical jitter and temperature fluctuations inevitably affect the noise performance of the output pulses. This is particularly evident in high-repetition-rate passively mode-locked fiber lasers, where the ultrashort linear cavity structure makes the output pulse repetition rate more susceptible to environmental noise. Driven by the demands of precision applications, controlling environmental disturbances to improve the noise performance of output pulse sequences from high-repetition-rate passively mode-locked fiber lasers has become a critical research focus. This chapter employs phase-locked loop technology to achieve phase synchronization of the repetition rate signal from high-repetition-rate passively mode-locked fiber lasers with an all-fiber structure, thereby enhancing their noise performance.

Test Equipment:
Voltage amplifier, Signal generator, Bandpass filter, Low-pass filter, Actuator, Controller, Thermoelectric controller, etc.

Experimental Process:

Figure 1: Schematic diagram of the repetition rate locking experimental setup

In the repetition rate synchronization experiment, the repetition rate of a passively mode-locked fiber laser with a pulse repetition rate of 1.27 GHz was phase-synchronized with a commercial microwave reference source. The locking schematic is shown in Figure 1. As illustrated, the mode-locked optical pulses with a repetition rate of 1.27 GHz first pass through an isolator and are then converted into electrical signals using a photodetector. The isolator is added to prevent reflected light from affecting the stable operation of the resonator. The amplified repetition rate electrical signal is filtered by a bandpass filter to extract the fundamental frequency of the pulses. The fundamental frequency signal and the reference signal undergo a phase discrimination process in a double-balanced mixer. The generated error signal is passed through a low-pass filter and then injected into a loop filter composed of an analog PID controller. This PID controller has a working bandwidth of 100 kHz, sufficient to suppress environmental noise in the 10 Hz–1 kHz frequency range. The control signal generated by the PID controller is injected into a piezoelectric controller VCO, which consists of an optical fiber fixed to the surface of a PZT actuator. During the operation of the mode-locked laser, the heat generated by the saturable absorption process in the saturable absorber causes the cavity temperature to rise. To control the cavity temperature, the entire resonator is placed in a structural component made of brass with good thermal conductivity. The actuator for temperature control is a thermoelectric controller, and the control circuit is a commercial temperature control instrument with a precision of 0.02°C.

Figure 2: Static tuning curve of the voltage-controlled oscillator

Before conducting the laser repetition rate locking experiment, the static and dynamic response characteristics of the voltage-controlled oscillator (VCO), which tunes the repetition rate by stretching the optical fiber via a piezoelectric actuator, were tested. This provided a foundation for determining the specific loop locking structure and PID parameter settings. During the static testing of the VCO, a waveform generator was used to apply square wave signals with varying amplitudes to the piezoelectric actuator, and a signal analyzer recorded the repetition rate offset at different amplitudes. The experimentally measured static response curve of the VCO is shown in Figure 2. Within the 0–5 V range, the repetition rate offset exhibits a linear relationship with the applied voltage, with a fitted coefficient of 1.29 kHz/V. Considering that the maximum repetition rate offset within 60 minutes is only 430 Hz, the static response coefficient of 1.29 kHz/V implies that the signal output by the PID controller in the control loop can directly drive the piezoelectric actuator without the need for a voltage amplifier in the control loop. In structures where the cavity length is adjusted by moving the cavity mirror to lock the repetition rate, the VCO static response coefficient is often too low (e.g., 74.2 Hz/V), necessitating the inclusion of a voltage amplifier in the control loop to drive the VCO.

Figure 3: (a) Schematic for testing the dynamic tuning curve of the VCO, (b) Dynamic tuning curve of the VCO

The principle for testing the dynamic response characteristics of the VCO is shown in Figure 3(a). A sinusoidal modulation signal with an amplitude of 2 V from a signal generator is applied to the piezoelectric actuator in the VCO. The modulated optical pulses are converted into electrical signals by a high-speed photodetector, and the frequency of the electrical signals is reduced to within the bandwidth of a lock-in amplifier using a double-balanced mixer and a low-pass filter. The down-converted electrical signal is input to the signal port of the lock-in amplifier, while the modulation signal from the signal generator is input to the reference port of the lock-in amplifier, completing the test loop. The measured dynamic response curve of the VCO is shown in Figure 3(b). It can be observed that when the applied modulation frequency is below 1000 Hz, the intensity response of the VCO is flat. A resonance peak appears in the response curve around 1050 Hz. When the modulation frequency is below 100 Hz, the phase response function remains at 0, and when the modulation frequency reaches ~1050 Hz, the phase response function exhibits a 90° phase shift.

Experimental Results:

Figure 4: Output pulse sequence after resonator repetition rate stabilization: (a) Phase noise, (b) Comparison of relative intensity noise

In the experiment, the controller was set to PI control mode, and the specific PI parameters were determined using the Ziegler-Nichols tuning method. The typical experimental results are shown in Figure 4. From Figure 4(a), it can be seen that through phase-locked loop technology, the phase noise of the output pulses is significantly suppressed within the frequency offset range of 1–100 Hz. The phase-locked loop technology does not affect noise components beyond 1 kHz, indicating that it effectively suppresses the impact of environmental noise on pulse noise performance but has no effect on noise related to pulse dynamics. The accumulated timing jitter decreased from 200 ps to 642 fs, and the corresponding accumulated phase noise reduced from 1.65 rad to 4.99 mrad. The phase noise curve of the frequency-stabilized pulses begins to deviate from the phase noise of the reference source around 20 Hz. Considering that the dynamic response bandwidth of the VCO is as high as 1 kHz, it can be concluded that this deviation is not due to bandwidth limitations. Around 10 kHz, the phase noise of the output pulses becomes lower than that of the reference source. Due to the influence of the phase-locked loop on the VCO, the relative intensity noise of the frequency-stabilized output pulses is significantly higher than that of the unlocked laser source. The corresponding accumulated relative intensity noise increased from 0.228% to 0.400%, and harmonic components of 50 Hz appeared on the relative intensity noise curve.

Figure 5: Frequency stability of the output pulses after laser resonator repetition rate stabilization

After the pulse repetition rate was locked, it was down-converted to ~3.3 kHz using a double-balanced mixer, and the frequency of the output pulses was recorded using a universal frequency counter with a bandwidth of 350 MHz. The stability of the laser repetition rate is shown in Figure 5. The gate time of the frequency counter was set to 100 ms, and 900 data points were recorded. The results show that during the recording period, the repetition rate of the output pulses exhibited random jitter on the order of mHz around the center frequency, but no significant frequency drift was observed. The calculated standard deviation was 9 mHz, and the relative Allan variance for a 1-second averaging time was on the order of 10⁻¹².

Voltage Amplifier Recommendation: ATA-2042

Figure: ATA-2042 High-Voltage Amplifier Specifications

The experimental materials in this article have been compiled and released by Xi'an Aigtek Electronics. For more experimental solutions, please continue to follow the Aigtek official website. Aigtek is a high-tech enterprise in China specializing in the research, development, production, and sales of measurement instruments. The company has consistently focused on the R&D and manufacturing of test instrument products such as high-voltage amplifiers, voltage amplifiers, power amplifier modules, and high-precision current sources.