Automation of Mode Locking in a Nonlinear Polarization Rotation Fiber Laser through Output Polarization Measurements

The overall goal of this procedure is to detect an automate mode locking in a pre-adjusted non-linear polarization rotation fiber laser. This procedure provides an alternate way to existing automating procedure based on an RF spectrum analyzer, an optical spectrum analyzer, a non-linear detecting scheme or a pause counting device. Its main advantages are that it is relatively inexpensive, it is easy to implement and it requires only a small fraction of the laser output to be tapped for monitoring.

This animation shows the concept of mode locking based on non-linear polarization rotation. The signal is polarized by the polarizer and then transformed into an elliptic polarization state by the polarization controller. The current non-linearity in the fibers of the cavity forces a rotation of this ellipse, which is proportional to the instantaneous power of the signal.

Transmission at the final polarizer will thus favor the transmission of high power portions of the signal, leading to the formation of a pulse. By positioning a polarization analyzer just before the final polarizer, detection of a pulse in the cavity is possible by discriminating between polarization states. This is achievable because a pulse will undergo a larger non-linear polarization rotation than a continuous wave signal during a round trip in the cavity.

The first step is to set up a fiber laser system on a stable platform. To begin, set up the laser ring cavity at the core of this experiment on an optical bench. This schematic provides an overview for orientation.

Light propagates clockwise. To understand the setup, start with a polarizer. A fiber pigtailed polarizer is used.

It is fusion spliced with a section of polyimide coated fiber that is inserted in an all-fiber motorized yow type polarization state controller. The presence of this component is essential for the automation procedure. After the polarization controller, place a 980/1550 nanometer wavelength division multiplexer that is pumped by a 976 nanometer laser diode.

Then pass the light through a single-mode erbium-doped optical fiber that is the gain medium. The erbium-doped fiber glows green when it is pumped as shown here. The erbium-doped fiber is fusion spliced with a hybrid 980/1550 nanometer wavelength division multiplexer and a 1550 nanometer isolator.

It is pumped by a second laser diode. The final element in the cavity loop is a 50-50 output coupler that should be put just before the polarizer. The position of the cavity output coupler is really important since the polarization analyzer will receive its input from this coupler.

It must be located just before the polarizer in order to maximize the effect of non-linear polarization rotation. The output from a cavity goes to a 99:1 splitter with 99%going to usable output. The 1%goes through a manual polarization controller and provides feedback for automated mode locking.

The next step is to connect the motorized polarization controller to a computer controller. First connect the intercavity fiber squeezer polarization controller to its driving module. Then connect the driving module to the USB port of the computer.

Return to the laser output fibers on the bench and select the 99%output. Connect the 99%output to an optical spectrum analyzer using a barefit adapter. Turn on the pump lasers and move to the computer to start the instrument interface.

Use the interface to command the polarization controller to rotate 3000 steps clockwise. While rotating, the controller will reach a mechanical stop. Next, command the polarization controller to rotate one approximately degree counter-clockwise.

Observe the spectrum on the optical spectrum analyzer to determine if mode-lock has been reached. The spectra are representative of mode locked quasi-continuous and Q-switched regimes. Compared with the quasi-continuous and Q-switched regimes, the mode locked spectrum is very broad.

Rotate the polarization controller counter-clockwise in steps of one degree. When mode-lock is reached, fix the angle before moving on. At this point, prepare to search for the pump power thresholds.

Reduce the pump power until mode locking is lost. Then increase the pump power to just above the mode-locking threshold. Observe the mode-locking spectrum then turn the laser power off and back on to ensure the laser mode locks by itself.

To confirm the mode-locking state, disconnect the fiber from the spectrum analyzer. Connect the fiber to a fast photo diode setup to display a signal on an oscilloscope. The mode-locking state produces a pulse strain with a repetition rate of about 82 megahertz.

The next step is to prepare to analyze the polarization of the laser output. To do this, use a commercial polarization analyzer also a called a polarimeter and input the fiber from the 1%tap. The final arrangement of equipment for the experiment is depicted in this schematic.

The computer can control the polarization analyzer. Use the computer to put the polarization controller at it is mechanical stop and to stop the polarimeter software. At this point, the optical spectrum analyzer displays a non-mode locked spectrum.

In the polarimeter software, begin polarization measurements by clicking the start button. Step the polarization controller counterclockwise in one degree increments. As the range of angles allowed by the inter-cavity polarization controller is explored, observe the polarization state using the polarimeter software.

Continue exploring the angles until there is evidence of mode-locking. Note that the polarization stays very smoothly with the angle except where mode-locking is reached. Observe the spectrum on the spectrum analyzer to verify the signature of mode-locking is evident.

Return to the computer to continue. Once again rotate the polarization controller clockwise until it can no longer rotate. In the polarimeter software, prepare to monitor the Stokes parameters as a function of angle.

Here they are ordered S zero to S three from top to bottom. Step through the angles of the polarization controller in increments of one degree while recording and viewing the values of S one, S two, and S three. It can be seen that S one undergoes an abrupt variation when the transition to mode-locking occurs.

In a graphical programming language, write a script that will find the mode-locking condition automatically. The logic of this script is given in this flowchart. Basically, the script will start with the polarization state controller at the mechanical stop and will increase its angle by steps of one degree while reading the value of S one from the polarimeter at each step.

As soon as the value of S one increases by more than a pre-determined thresholds value of zero point three in a single step, the script will stop since mode-locking should be achieved. Next, run the script. This version of the script displays the evolution of the Stokes parameter S one as the angle is incremented.

As the angular sweep begins, the laser is not mode-locked. As the polarization controller rotates, the value of S one and the optical spectrum of the laser are evolving. When the script is over, mode-locking is achieved as seen by the broad optical spectrum.

The next step is to make a replacement for the commercial polarimeter. Use an oscilloscope connected to the computer with a GPIB interface. Next, turn attention to the optical components of the replacement polarimeter.

There is a polarizing beam splitter at the center aligned with three FC-APC fiberoptic port collimators. Light is input into the beam splitter from the port on top. Vertically polarized light exits the system from the right port.

Horizontally polarized light is output from the bottom port. The polarization analyzer also requires fabricating two identical electronic circuits. Here, the two transimpedance amplifier circuits are on the same bright board.

The layout of an individual circuit is in this diagram. First, an indium gallium arsenide photodiode detects a 1550 nanometer signal. The photodiode is connected to an operational amplifier, a resistor, and a capacitor.

The output of the circuit is connected to an oscilloscope and provides a measure of the average optical power. At the bench, begin connecting the beam splitter outputs to the circuits. First, connect one output to an indium gallium arsenide photodiode.

Next, connect the output of the transimpedance amplifier circuit to channel one of the oscilloscope. Turn on the oscilloscope and the transimpedance amplifier circuit. Disconnect the 1%output of the laser from the commercial polarimiter.

Connect the output to the input port of the polarizing beam splitter. Turn on the laser at an arbitrary pump power to send a 1550 nanometer optical signal. On the computer run a script to read the average voltage on channel one of the oscilloscope.

Return to the optical elements of the polarization analyzer. Disconnect the output of the polarizing beam splitter from the photodiode. Instead connect the beam splitter output to a commercial power meter.

Read and record the optical power for this pump power. Continue by varying the power of the input optical signal then measuring the average voltage and optical power. After several measurements, a plot of the voltage versus power should be linear.

Determine the coefficients of this linear relation. Follow the same procedure with the second output of the beam splitter to arrive at the final polarization analyzer setup. Both the vertical and horizontal polarization outputs are connected to amplifier circuits.

Each circuit has a dedicated oscilloscope channel, which is read by the computer. Now incorporate the new polarization analyzer into the automated mode-locking process. Open the scripts written to search for discontinuity in the Stokes phenomenon S one as a function of the polarization controller angle.

To modify it only requires a change in determining S one from the apparatus. For this analyzer, compute S one using this formula. The power values PX and PY are found from the measured linear relation between voltage and power for each polarization.

When done, start the script. This version of the script displays the evolution of the Stokes parameter S one as the angle is incremented. As the angular sweep begins, the laser is not mode-locked.

As the polarization controller rotates, the value of S one and the optical spectrum of the laser are evolving. When the script is over, mode-locking is achieved as seen by the broad optical spectrum. This is a typical plot of the Stokes parameter S one versus the angle of the motorized polarization state controller.

S one is calcualted using power values measured from a non-commercial polarization analyzer. An abrupt change occurs when the laser reaches the mode-locking state. An automated script written to stop varying the polarization controller angle at a discontinuity can find mode-lock within a few minutes.

The procedure can find mode-locking within a few minutes. Its implementation does not affect the laser cavity design and requires only 1%of the output signal to be monitored leaving 99%for intended applications. For the procedure to perform the laser parameters, such as its bump power and the birefringence of the polarization controller must be pre-adjusted properly in order to avoid undesired regimes of operation such as multiple pulsing or noise-like pulses.

Additional work will be required to study the applicability of this procedure to different laser designs and operating wavelengths. We believe this procedure could be used in commercial fiber laser systems where one expects modelocking to occur automatically at start-up.