Great video! And thanks for keeping the comments section open! In the video you say that the inability to get a perfect linear polarization maintained is in the fiber (namely, its imperfections). In my setups I usually blame waveplates :) Therefore, for this setup I would swap a half-wave plate an a polarizer, since I trust more to the polarizer that nothing except linear polarization survives there. I wonder, have you tried the same experiment with polarizer being set after the waveplate?
@ВладимирЗенин-ч4к It is great that you are considering the components’ limitations and thinking of ways to improve the setup! We agree that the assembly you describe (half-wave plate followed by a linear polarizer (HWP-LP)) likely provides a higher extinction ratio (ER) than the assembly used in the demonstration (linear polarizer followed by a half-wave plate (LP-HWP)). We did not perform a direct comparison of the HWP-LP versus LP-HWP options on the PM fiber’s output polarization state. Both options provided linearly polarized light with an ER sufficient for our application, but for this demonstration we preferred the LP-HWP option, since its alignment was more straight-forward. However, our experience and published research agree that, even when the input light is perfectly linearly polarized and aligned to the PM fiber’s axis, it is expected the light output by the fiber will be slightly elliptically polarized. Stress points are common places for cross-talk to occur between a PM fiber’s orthogonal axes.
@victoriazhang9274 Thanks for your suggestion! Would you help us better understand what you have in mind, since the word “lensed” means different things to different people? For example, some people use the term to describe fiber with a shaped end face (often rounded), which creates a lensing effect for light exiting the fiber. Other people use the term to refer to fibers that are terminated with a GRIN lens or that are connected to a fiber collimator. What does “lensed” mean to you? It would also be great to hear how close you’re placing the fibers and if you’re putting any optical components in between.
Great demonstration. Could you explain why the single mode fiber from the pigtail laser does not cause polarization fluctuations? If I only need a stable polarization, not necessarily linear, then can I just use a single mode fiber instead of a PM fiber?
@user-eg4uc2iy4f Single mode fiber can be a good option for preserving a random polarization state (impossible for a PM fiber), assuming a few conditions are met. The fiber must be protected from mechanical and temperature perturbations, which can cause unexpected changes in the polarization state. Results are also better for shorter lengths of fiber and over shorter time periods. In our demonstration, we secured and stabilized our short single mode fiber and only showed the fiber’s output over a short time period. We added a linear polarizer to the FiberBench to ensure a stable linear polarization state was consistently input to the PM fiber, which is much better than single mode fiber for providing a stable linearly polarized output.
@sadiqnawazkhan333 Yes, shaking the PM fiber measured by the polarization analyzer/polarimeter is an equally valid approach to perturbing the fiber. When using that approach, we find the output polarization state jumps back and forth, and it was faster for us to create full circles on the Poincaré sphere by heating and cooling the fiber. If someone is looking for a method to heat and cool without having to get extra parts, the fiber coil can held flat between your hands/fingers to heat it and then placed on the optical table to cool. We liked using the temperature controlled breadboard, since someone just needs to turn the dial or apply a computer program to start changing the temperature, which leaves both hands free for alignment. Do you see consistent results when using the quarter-wave plate to counteract the excess birefringence? We have seen mixed results when using that approach with PM patch cables, which we thought was likely due to the stress-induced birefringence of the fiber connectors. More specifically, the quarter-wave plate compensated for the birefringence introduced by the first connector to create linearly polarized light within the PM fiber, but the birefringence caused by the second connector still affected the output polarization state.
@@thorlabs I remember the use of a quarter wave plate could not make the polarization circle a single point but only minimized it so that it was enough for our experiment. Also, the time scales at which the light was pulsed were very short and thus any residual impurities in the polarization mode could not contribute to the light intensity fluctuations after the fiber
@@thorlabs Woohoo! Can't wait! Great videos! 👍 Also, a beam circularization video would be great: free space elliptical beam to circular gaussian beam.
Since the PM fiber has a fast & slow axis, does it mean that it could only preserve light that is linearly polarized and aligned to either fast or slow axis? circularly or elliptically polarized light can't be preserved by the PM fiber? Since they do no oscillate along one axis.
@dewenzhang2115 You are correct! PM fiber can only preserve linearly polarization states that are coupled into (aligned with) either the fast or the slow axis. You are also correct in stating that PM fiber cannot preserve a specific circular or elliptical polarization state. As you say, the PM fiber will not preserve the polarization state of any input light that has polarization components that oscillate parallel to both the fast and slow fiber axes.
Question 1: when heating the fiber, what is the temperature ramp range? If we want to do long term measurement, do we need ramp up and down the temperature periodically? If the temperature is modulated, will we see the points on the circle moving in opposite direction? Question 2: I was doing similar measurement, when we stress the DUT, we see a the SOP point moves on the Poincare sphere. It forms a circle but the center of circle also moves. How do we explain the circles moving on the sphere?
@USAsuperstore Response 1: The temperature-controlled breadboard we used has a temperature range between 15 °C and 45 °C. We used the range from about 25 °C to 45 °C, which suited this demonstration well, but other applications and fibers may be better served by a different temperature range. The temperature needs to change in order to make the measurement. Temperature cycles can occur periodically to confirm alignment at discrete times or the temperature could be cycled continuous to visualize dynamics. But it is important to remember that temperature change will induce continuous polarization change at the fiber output during the assessment. Yes! The points on the Poincaré sphere will plot in opposite directions, depending on whether the temperature is increasing or decreasing. Response 2: This analysis technique assumes that the fiber’s birefringence changes, but that light does not couple between the fiber’s fast and slow axes. As you’ve noted, physically straining the fiber also changes the fiber’s birefringence. However, physically straining the fiber can also cause microbends and other effects that enable light to more easily couple between the slow and fast axes. When this occurs, even when the input light is perfectly aligned to one fiber axis, some light can couple into the orthogonal axis due to the stress. Since there will then be light in both axis, and the amount depends on the stress applied to the fiber, the center polarization state of the output light can drift.
Excellent video!! Question: If using the LP + power meter approach rather than polarimeter, why do we align the LP such that we are equally collecting from both PM axes? Can we simply align the LP to either slow or fast axis, and then tune the HWP to maximize power such that the input SOP is aligned with either slow or fast axis?
@szu-yulee390 In general when optimizing alignment, it is often necessary to look for small changes in output power when making adjustments. These small changes can be very difficult to detect when they are being added to, or subtracted from, a high-power baseline. Because of this, it can be very difficult to know when the power is truly maximized. Instead, it is often helpful to look for the minimum power, since small changes compared with a low-power baseline can be a lot easier to see. However, both of these approaches assume that the system can provide a constant maximum or minimum value. The PM fiber and linear polarizer system will typically not provide a stable maximum or minimum power output, since some light will always be coupled into the fiber’s orthogonal axis. Rather than try to take this effect into account, we find it easier to optimize alignment by setting the linear polarizer’s axis between the fiber’s axes and minimizing the amplitude of the oscillations.
@@thorlabsthank you very much for the details! I am still one step behind the logic. So the linear polarizer right before the power meter is set to between PM fiber's axes. Then, why minimizing the amplitude of the power oscillations by rotating the HWP means the linearly polarized input is aligned to one of the PM axis? Is it because otherwise the PM fiber output is in general elliptically polarized and we will see power oscillation?
@szu-yulee390 That is correct! Light output by PM fibers almost always includes a small fraction of elliptically polarized light. This is caused by a variety of reasons including: imperfect linear polarizers transmitting a small amount of light polarized in the orthogonal direction, obtaining perfect alignment is difficult, and light aligned to one fiber axis scattering into being aligned to the other fiber axis at perturbations like bends and irregularities in the glass material. This is why, in the demonstration, we could reduce - but not eliminate - the power oscillations.
Why do we need a half-wave plate? I think a single linear polarizer is enough (just rotate the linear polarization axis). Could you give me more information?
@user-gk2or5jb7o Technically, we did not have to use the half-wave plate (HWP), but we chose to use it because we wanted to couple the maximum amount of light into the PM fiber. The elliptically polarized light output by the input fiber had a significant linearly polarized component. We aligned the polarizer parallel to this component and then used the HWP to rotate the orientation of the transmitted linearly polarized light to align with one of the fiber’s axes. If we had used only the linear polarizer, and the orientation of the polarizer’s axis that aligned with the fiber’s axis was not parallel to the orientation of the linearly polarized component in the input light, the polarizer would have blocked light that the polarizer-HWP pair could have coupled into the fiber.
@DiabloMablo There are different options for attaching lenses to fiber end faces. In addition, the end of the fiber can be processed to include a lens directly (e.g. a ball or GRIN lens). It is typical to use lenses to collimate or focus the light from the fiber. One of the quickest options for collimating the light is to attach a fiber collimator (e.g. www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=4353&YVI=28).
Thanks, you really save someone's project with this video :)))))))))
Great video! And thanks for keeping the comments section open! In the video you say that the inability to get a perfect linear polarization maintained is in the fiber (namely, its imperfections). In my setups I usually blame waveplates :) Therefore, for this setup I would swap a half-wave plate an a polarizer, since I trust more to the polarizer that nothing except linear polarization survives there. I wonder, have you tried the same experiment with polarizer being set after the waveplate?
@ВладимирЗенин-ч4к It is great that you are considering the components’ limitations and thinking of ways to improve the setup! We agree that the assembly you describe (half-wave plate followed by a linear polarizer (HWP-LP)) likely provides a higher extinction ratio (ER) than the assembly used in the demonstration (linear polarizer followed by a half-wave plate (LP-HWP)).
We did not perform a direct comparison of the HWP-LP versus LP-HWP options on the PM fiber’s output polarization state. Both options provided linearly polarized light with an ER sufficient for our application, but for this demonstration we preferred the LP-HWP option, since its alignment was more straight-forward.
However, our experience and published research agree that, even when the input light is perfectly linearly polarized and aligned to the PM fiber’s axis, it is expected the light output by the fiber will be slightly elliptically polarized. Stress points are common places for cross-talk to occur between a PM fiber’s orthogonal axes.
@@thorlabs Thanks for detailed reply! Never thought that the cross talk in normally resting pm fiber can be notable
Could you give a tutorial for aligning the axis of two lensed PM fibers? This is commonly used in fiber-based experiments.
@victoriazhang9274 Thanks for your suggestion! Would you help us better understand what you have in mind, since the word “lensed” means different things to different people? For example, some people use the term to describe fiber with a shaped end face (often rounded), which creates a lensing effect for light exiting the fiber. Other people use the term to refer to fibers that are terminated with a GRIN lens or that are connected to a fiber collimator. What does “lensed” mean to you? It would also be great to hear how close you’re placing the fibers and if you’re putting any optical components in between.
Great demonstration. Could you explain why the single mode fiber from the pigtail laser does not cause polarization fluctuations? If I only need a stable polarization, not necessarily linear, then can I just use a single mode fiber instead of a PM fiber?
@user-eg4uc2iy4f Single mode fiber can be a good option for preserving a random polarization state (impossible for a PM fiber), assuming a few conditions are met. The fiber must be protected from mechanical and temperature perturbations, which can cause unexpected changes in the polarization state. Results are also better for shorter lengths of fiber and over shorter time periods. In our demonstration, we secured and stabilized our short single mode fiber and only showed the fiber’s output over a short time period. We added a linear polarizer to the FiberBench to ensure a stable linear polarization state was consistently input to the PM fiber, which is much better than single mode fiber for providing a stable linearly polarized output.
have used this method with SKPolarization analyzer with the help of a 1/2 and a 1/4 waveplate and did gentle shaking of the fiber by hands
@sadiqnawazkhan333 Yes, shaking the PM fiber measured by the polarization analyzer/polarimeter is an equally valid approach to perturbing the fiber. When using that approach, we find the output polarization state jumps back and forth, and it was faster for us to create full circles on the Poincaré sphere by heating and cooling the fiber. If someone is looking for a method to heat and cool without having to get extra parts, the fiber coil can held flat between your hands/fingers to heat it and then placed on the optical table to cool. We liked using the temperature controlled breadboard, since someone just needs to turn the dial or apply a computer program to start changing the temperature, which leaves both hands free for alignment.
Do you see consistent results when using the quarter-wave plate to counteract the excess birefringence? We have seen mixed results when using that approach with PM patch cables, which we thought was likely due to the stress-induced birefringence of the fiber connectors. More specifically, the quarter-wave plate compensated for the birefringence introduced by the first connector to create linearly polarized light within the PM fiber, but the birefringence caused by the second connector still affected the output polarization state.
@@thorlabs I remember the use of a quarter wave plate could not make the polarization circle a single point but only minimized it so that it was enough for our experiment. Also, the time scales at which the light was pulsed were very short and thus any residual impurities in the polarization mode could not contribute to the light intensity fluctuations after the fiber
Thanks Eric.
Next coupling SPDC photons 🤩
@girlonfire8177 Yes! Fiber-coupling single photons can get tricky. We've added that to our list. Thanks very much for the suggestion!
@@thorlabs Woohoo! Can't wait! Great videos! 👍 Also, a beam circularization video would be great: free space elliptical beam to circular gaussian beam.
@girlonfire8117 Another great suggestion, thank you!
Since the PM fiber has a fast & slow axis, does it mean that it could only preserve light that is linearly polarized and aligned to either fast or slow axis? circularly or elliptically polarized light can't be preserved by the PM fiber? Since they do no oscillate along one axis.
@dewenzhang2115 You are correct! PM fiber can only preserve linearly polarization states that are coupled into (aligned with) either the fast or the slow axis. You are also correct in stating that PM fiber cannot preserve a specific circular or elliptical polarization state. As you say, the PM fiber will not preserve the polarization state of any input light that has polarization components that oscillate parallel to both the fast and slow fiber axes.
Question 1: when heating the fiber, what is the temperature ramp range? If we want to do long term measurement, do we need ramp up and down the temperature periodically? If the temperature is modulated, will we see the points on the circle moving in opposite direction?
Question 2: I was doing similar measurement, when we stress the DUT, we see a the SOP point moves on the Poincare sphere. It forms a circle but the center of circle also moves. How do we explain the circles moving on the sphere?
@USAsuperstore Response 1: The temperature-controlled breadboard we used has a temperature range between 15 °C and 45 °C. We used the range from about 25 °C to 45 °C, which suited this demonstration well, but other applications and fibers may be better served by a different temperature range.
The temperature needs to change in order to make the measurement. Temperature cycles can occur periodically to confirm alignment at discrete times or the temperature could be cycled continuous to visualize dynamics. But it is important to remember that temperature change will induce continuous polarization change at the fiber output during the assessment.
Yes! The points on the Poincaré sphere will plot in opposite directions, depending on whether the temperature is increasing or decreasing.
Response 2: This analysis technique assumes that the fiber’s birefringence changes, but that light does not couple between the fiber’s fast and slow axes. As you’ve noted, physically straining the fiber also changes the fiber’s birefringence. However, physically straining the fiber can also cause microbends and other effects that enable light to more easily couple between the slow and fast axes. When this occurs, even when the input light is perfectly aligned to one fiber axis, some light can couple into the orthogonal axis due to the stress. Since there will then be light in both axis, and the amount depends on the stress applied to the fiber, the center polarization state of the output light can drift.
@@thorlabs Excellent and super clear answers. Much appreciated.
Excellent video!!
Question: If using the LP + power meter approach rather than polarimeter, why do we align the LP such that we are equally collecting from both PM axes? Can we simply align the LP to either slow or fast axis, and then tune the HWP to maximize power such that the input SOP is aligned with either slow or fast axis?
@szu-yulee390 In general when optimizing alignment, it is often necessary to look for small changes in output power when making adjustments. These small changes can be very difficult to detect when they are being added to, or subtracted from, a high-power baseline. Because of this, it can be very difficult to know when the power is truly maximized. Instead, it is often helpful to look for the minimum power, since small changes compared with a low-power baseline can be a lot easier to see.
However, both of these approaches assume that the system can provide a constant maximum or minimum value. The PM fiber and linear polarizer system will typically not provide a stable maximum or minimum power output, since some light will always be coupled into the fiber’s orthogonal axis. Rather than try to take this effect into account, we find it easier to optimize alignment by setting the linear polarizer’s axis between the fiber’s axes and minimizing the amplitude of the oscillations.
@@thorlabsthank you very much for the details! I am still one step behind the logic. So the linear polarizer right before the power meter is set to between PM fiber's axes. Then, why minimizing the amplitude of the power oscillations by rotating the HWP means the linearly polarized input is aligned to one of the PM axis?
Is it because otherwise the PM fiber output is in general elliptically polarized and we will see power oscillation?
@szu-yulee390 That is correct! Light output by PM fibers almost always includes a small fraction of elliptically polarized light. This is caused by a variety of reasons including: imperfect linear polarizers transmitting a small amount of light polarized in the orthogonal direction, obtaining perfect alignment is difficult, and light aligned to one fiber axis scattering into being aligned to the other fiber axis at perturbations like bends and irregularities in the glass material. This is why, in the demonstration, we could reduce - but not eliminate - the power oscillations.
@@thorlabsthis is again very clear and insightful explanations to me. Thank you very much. Keep the high quality videos coming!
@szu-yulee390 You are welcome, thanks for asking :) And, if you have a topic suggestion for us, please let us know!
Why do we need a half-wave plate? I think a single linear polarizer is enough (just rotate the linear polarization axis). Could you give me more information?
@user-gk2or5jb7o Technically, we did not have to use the half-wave plate (HWP), but we chose to use it because we wanted to couple the maximum amount of light into the PM fiber. The elliptically polarized light output by the input fiber had a significant linearly polarized component. We aligned the polarizer parallel to this component and then used the HWP to rotate the orientation of the transmitted linearly polarized light to align with one of the fiber’s axes.
If we had used only the linear polarizer, and the orientation of the polarizer’s axis that aligned with the fiber’s axis was not parallel to the orientation of the linearly polarized component in the input light, the polarizer would have blocked light that the polarizer-HWP pair could have coupled into the fiber.
Is it possible to put a lens on the end of a fiber optic cable to project an image out?
@DiabloMablo There are different options for attaching lenses to fiber end faces. In addition, the end of the fiber can be processed to include a lens directly (e.g. a ball or GRIN lens). It is typical to use lenses to collimate or focus the light from the fiber. One of the quickest options for collimating the light is to attach a fiber collimator (e.g. www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=4353&YVI=28).