I recommend performing this analysis as I have one of these quarter wave plates that is incorrectly labeled. I performed this experiment on two different QWPs and found one of them was incorrect.
We’re so happy you found the video useful! We also recommend confirming the optic orientation, since it can be different than expected for various reasons. One is that the engravings on waveplates can be hard to interpret. In fact, we know that the engravings on ours manufactured prior to October 2018 have led to a fair amount of confusion. Another reason is that some people, in my lab at least :) , like to take the optic out of the mount, use the optic in something else, and then orient the optic incorrectly when they sneak it back into the mount. Please let us know if there are any other procedures you would like to see demonstrated or if you have any useful tips you would like to share. -Bill Warger, Video Insights Director
@Kyle Throssell Thank you! While we are not a member of the International DOI Foundation, we appreciate the enthusiasm and interest in our video! For those who would like to cite a video, we suggest following the formatting guidance specific to the citation style being used.
You note that "Note that this technique cannot be used to distinguish the fast and slow axes of a half-wave plate." Does ThorLabs have a video for that particular experiment? Thanks, this was a great video.
@user-ch5eq2mu3f Thorlabs doesn’t currently have a Video Insight demonstrating a method to distinguish a HWP’s slow and fast axis, but thanks for asking - it’s now on our list! It is typical to use measurements of optical path length to determine which HWP axes are fast and slow. The optical path length traveled by linearly polarized light aligned with the HWP’s slow axis is longer than the optical path length when the linearly polarized light is aligned with the HWP’s fast axis.
@davidjoy2128 The power sensor we used in this demonstration is a silicon-based photodiode ( www.thorlabs.com/thorproduct.cfm?partnumber=S130C&YVI=14 ), which is sensitive to visible wavelengths and has a semiconductor PN junction at its core. This article ( www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=14218&YVI=14 ) gives a quick overview of photodiodes, as well as providing some information about their wavelength-dependent performance. We also mention at 03:24 that we attached a couple of components to the power sensor that minimize the amount of stray light reaching the sensor. One of these components is the long black tube (lens tube), and the other is the iris that is attached to the input of the lens tube. The combination of the lens tube and the iris allows only the light travelling along the laser beam’s path to reach the power sensor, which reduces the amount of non-signal light incident on the sensor. We have provided part numbers for all setup components (including the lens tube and iris) in the description, if you would like additional information about them.
Hi. I am using an achromatic quarter wave plate which is a composite material. Can you make a video or point me to a resource to minimize the error due to tilt of the plate?
@mukulsharma3447 Thanks for the suggestion! There are different approaches to minimizing the error. One approach is to adjust the wave plate’s alignment, until the wave plate provides as close to optimal retardance as possible. For example, input linearly polarized light to the wave plate, while simultaneously adjusting the wave plate’s alignment and using a polarimeter to analyze the output light. Another approach, which can be used alone or in addition to the previous approach, is to mathematically correct the results provided by the system after making a measurement. This requires measuring the wave plate’s actual retardance. When this is known, the actual retardance can be used when analyzing the results provided by the optical setup. The approach is described in Section 3.A.3 of Paul A. William’s paper (Applied Optics, Vol. 38, No. 31 (1999) 6508).
While we chose to cross the transmission axes of the two polarizers, the technique will also work if their transmission axes are oriented parallel to one another. However, if the two polarizers are oriented parallel to one another, the equation for R needs to be modified: in the numerator of the equation for R, ±Delta should become ±(-Delta). If the equation is not modified, but the two polarizers are oriented parallel to one another, the "fast axis horizontal" curve changes places with the "fast axis vertical" curve. A good reason for orienting their transmission axes either parallel or perpendicular to one another is that either of these two cases simplifies the equation for R. For those wondering about our mirror, we used an unprotected gold mirror, instead of a cold mirror. Cold mirrors (www.thorlabs.com/navigation.cfm?guide_id=2234 ) are designed to transmit IR wavelengths while reflecting visible wavelengths. We did not use a cold mirror because the complex refractive index of gold is more readily available.
Thank you for great explanations. You mentioned thath using unprotected mirrors makes life easier. If I will uae protected gold mirror, how much it will influence the result? Is it still posiible use this method?
@Namig Alasgarzade An unprotected gold mirror (item # PF10-03-M03) was recommended since the equations used to calculate the reflection coefficients are simpler when the reflector consists of a bare metal surface. A protected gold mirror or other coated metal surface can be used with this method, but the equations used to calculate the reflection coefficient depend on the number of coating layers, their thicknesses, and their refractive indices. Not only are the equations more complex than the ones shown in this video, but it can also be difficult to determine the information needed to perform the calculations. That said, we tested a protected gold mirror (item # PF10-03-M01) in our setup, and we measured power values very similar to those we measured in the video. So, under certain circumstances, the protective coating layer may not have a significant impact on the measurement, but in general that can't be assumed.
@@thorlabs I used the protected gold mirror (#PF20-03-M01) and found the results are opposite to yours. When engraving is horizontal, it has smaller reflected power than when it's vertical. I don't know why? I use the 800 nm laser. Moreover, when I tested with a HWP, the power when engraving is horizontal is larger than when it's vertical. Is it right for the HWP?
@@xinhejiang4027 It is hard to know the reason your results differ from ours without having a better understanding of your setup and the measurements that you are making. Having said that, there are two things that typically result in measurements that are opposite to the ones we made during our demonstration. One is the orientation of the analyzing polarizer's transmission axis. In our setup, the transmission axes of the generating and analyzing polarizers are crossed. When the two polarizers' axes are parallel, the power measured at the detector when the quarter-wave plate's (QWP's) axis is horizontal would follow the lower-power, red curve in the plot shown in the video (6:47). The power measured at the detector when the QWP's axis is vertical would follow the higher-power, blue curve in that plot. Something else that would result in your measurements being the opposite of ours is if the engraving on your QWP indicates the slow axis, rather than the fast axis of the QWP. This is definitely a possibility. However, it is confusing that, when a half-wave plate (HWP) is inserted in your setup, significantly different powers are measured when the fast axis of the wave plate is vertical and horizontal. It is expected that both power measurements would be the same. This can be seen by considering the polarization orientation of the light output by the HWP. The light output by the HWP should be the same when HWP's fast axis is oriented both vertically and horizontally. (Due to this, the method described in this video is not appropriate for distinguishing the fast and slow axes of a HWP.) The measurements you obtained with the HWP in the setup indicates that your setup may differ from the one we demonstrated in unknown ways. We advise comparing your setup with ours and contacting Tech Support (techsupport@thorlabs.com , or www.thorlabs.com/locations.cfm) if you'd like to discuss it further.
@@thorlabs Hello! I am sure the transmission axes of the generating and analyzing polarizers are crossed. I used the QWP (WPQ05M-808) to test. But I think this QWP is an old component. The engraving on the QWP is only "FAST/". Not "FAST/Axis". I see that the powers are almost the same (like 8.10 mW and 7.82 mW) for the HWP. We used a 810 laser and the set-up is almost like yours. I think I need to send an email to your Tech Support.
@fethallahkarim-pj7so We expect that the polarizing beamsplitter (PBS) would transmit some horizontally polarized light. This is because linearly polarized laser sources do not provide perfectly linearly polarized light, and because the PBS does not perfectly separate the two orthogonal orientations of the polarized light. Most light output from a linearly polarized light source is polarized in the desired direction, which is vertical in your case. However, it is typical that a small fraction of the output light is polarized in the orthogonal direction, which is horizontal in your case. The amount of laser light polarized along the desired direction is commonly specified using the polarization extinction ratio (PER). This ratio is calculated by dividing the intensity of the light polarized along the desired direction by the intensity of the light polarized along the orthogonal direction. Ideally, the PER would be infinite. Similarly, the performance of a PBS is described by its extinction ratio (ER), which is typically measured and specified for only the transmitted direction. One approach to measure the ER for the transmitted direction is to place the PBS in a highly polarized beam of light, measure the power in the transmitted beam as the PBS rotates 360° around the optical axis, and divide the average of the two highest powers by the average of the two lowest powers. It is possible to estimate the amount of horizontally polarized light that passes through the PBS in the case you brought up. For example, consider vertically polarized input laser light with a PER of 15 dB. Assume that vertical polarization is aligned to the P-polarized direction of the PBS, the ER of the PBS’ transmitted path is > 1000:1, and the reflection efficiency of S-polarized light in the PBS is >95%. Since the laser’s PER is 15 dB, approximately 3.2% of the input light is expected to be in the orthogonal (horizontal, unwanted) polarization direction. With an ER >1000:1, we would expect no more than 0.0032% of the incident power to be horizontally polarized after traveling straight through the cube. With a reflection efficiency >95%, we would expect about 3% of the incident power to be reflected from the cube as horizontally polarized light.
I recommend performing this analysis as I have one of these quarter wave plates that is incorrectly labeled. I performed this experiment on two different QWPs and found one of them was incorrect.
We’re so happy you found the video useful! We also recommend confirming the optic orientation, since it can be different than expected for various reasons. One is that the engravings on waveplates can be hard to interpret. In fact, we know that the engravings on ours manufactured prior to October 2018 have led to a fair amount of confusion. Another reason is that some people, in my lab at least :) , like to take the optic out of the mount, use the optic in something else, and then orient the optic incorrectly when they sneak it back into the mount.
Please let us know if there are any other procedures you would like to see demonstrated or if you have any useful tips you would like to share.
-Bill Warger, Video Insights Director
This video needs a doi for when it inevitably becomes the key information for a grad student to understand their work and publish!
@Kyle Throssell Thank you! While we are not a member of the International DOI Foundation, we appreciate the enthusiasm and interest in our video! For those who would like to cite a video, we suggest following the formatting guidance specific to the citation style being used.
You note that "Note that this technique cannot be used to distinguish the fast and slow axes of a half-wave plate." Does ThorLabs have a video for that particular experiment? Thanks, this was a great video.
@user-ch5eq2mu3f Thorlabs doesn’t currently have a Video Insight demonstrating a method to distinguish a HWP’s slow and fast axis, but thanks for asking - it’s now on our list!
It is typical to use measurements of optical path length to determine which HWP axes are fast and slow. The optical path length traveled by linearly polarized light aligned with the HWP’s slow axis is longer than the optical path length when the linearly polarized light is aligned with the HWP’s fast axis.
Nice video. I have a question, you mentioned photodiode at 03:24, how does it work?
@davidjoy2128 The power sensor we used in this demonstration is a silicon-based photodiode ( www.thorlabs.com/thorproduct.cfm?partnumber=S130C&YVI=14 ), which is sensitive to visible wavelengths and has a semiconductor PN junction at its core. This article ( www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=14218&YVI=14 ) gives a quick overview of photodiodes, as well as providing some information about their wavelength-dependent performance. We also mention at 03:24 that we attached a couple of components to the power sensor that minimize the amount of stray light reaching the sensor. One of these components is the long black tube (lens tube), and the other is the iris that is attached to the input of the lens tube. The combination of the lens tube and the iris allows only the light travelling along the laser beam’s path to reach the power sensor, which reduces the amount of non-signal light incident on the sensor. We have provided part numbers for all setup components (including the lens tube and iris) in the description, if you would like additional information about them.
Thanks!
Hi. I am using an achromatic quarter wave plate which is a composite material. Can you make a video or point me to a resource to minimize the error due to tilt of the plate?
@mukulsharma3447 Thanks for the suggestion! There are different approaches to minimizing the error. One approach is to adjust the wave plate’s alignment, until the wave plate provides as close to optimal retardance as possible. For example, input linearly polarized light to the wave plate, while simultaneously adjusting the wave plate’s alignment and using a polarimeter to analyze the output light.
Another approach, which can be used alone or in addition to the previous approach, is to mathematically correct the results provided by the system after making a measurement. This requires measuring the wave plate’s actual retardance. When this is known, the actual retardance can be used when analyzing the results provided by the optical setup. The approach is described in Section 3.A.3 of Paul A. William’s paper (Applied Optics, Vol. 38, No. 31 (1999) 6508).
Thanks for the video. I have a question, why the polarizer after cold mirror have to be cross polarized to the first polarizer?
While we chose to cross the transmission axes of the two polarizers, the technique will also work if their transmission axes are oriented parallel to one another. However, if the two polarizers are oriented parallel to one another, the equation for R needs to be modified: in the numerator of the equation for R, ±Delta should become ±(-Delta). If the equation is not modified, but the two polarizers are oriented parallel to one another, the "fast axis horizontal" curve changes places with the "fast axis vertical" curve. A good reason for orienting their transmission axes either parallel or perpendicular to one another is that either of these two cases simplifies the equation for R.
For those wondering about our mirror, we used an unprotected gold mirror, instead of a cold mirror. Cold mirrors (www.thorlabs.com/navigation.cfm?guide_id=2234 ) are designed to transmit IR wavelengths while reflecting visible wavelengths. We did not use a cold mirror because the complex refractive index of gold is more readily available.
Thank you for great explanations. You mentioned thath using unprotected mirrors makes life easier. If I will uae protected gold mirror, how much it will influence the result? Is it still posiible use this method?
@Namig Alasgarzade An unprotected gold mirror (item # PF10-03-M03) was recommended since the equations used to calculate the reflection coefficients are simpler when the reflector consists of a bare metal surface. A protected gold mirror or other coated metal surface can be used with this method, but the equations used to calculate the reflection coefficient depend on the number of coating layers, their thicknesses, and their refractive indices. Not only are the equations more complex than the ones shown in this video, but it can also be difficult to determine the information needed to perform the calculations. That said, we tested a protected gold mirror (item # PF10-03-M01) in our setup, and we measured power values very similar to those we measured in the video. So, under certain circumstances, the protective coating layer may not have a significant impact on the measurement, but in general that can't be assumed.
@@thorlabs Thanks a lot :). I
@@thorlabs I used the protected gold mirror (#PF20-03-M01) and found the results are opposite to yours. When engraving is horizontal, it has smaller reflected power than when it's vertical. I don't know why? I use the 800 nm laser. Moreover, when I tested with a HWP, the power when engraving is horizontal is larger than when it's vertical. Is it right for the HWP?
@@xinhejiang4027 It is hard to know the reason your results differ from ours without having a better understanding of your setup and the measurements that you are making. Having said that, there are two things that typically result in measurements that are opposite to the ones we made during our demonstration. One is the orientation of the analyzing polarizer's transmission axis. In our setup, the transmission axes of the generating and analyzing polarizers are crossed. When the two polarizers' axes are parallel, the power measured at the detector when the quarter-wave plate's (QWP's) axis is horizontal would follow the lower-power, red curve in the plot shown in the video (6:47). The power measured at the detector when the QWP's axis is vertical would follow the higher-power, blue curve in that plot. Something else that would result in your measurements being the opposite of ours is if the engraving on your QWP indicates the slow axis, rather than the fast axis of the QWP. This is definitely a possibility.
However, it is confusing that, when a half-wave plate (HWP) is inserted in your setup, significantly different powers are measured when the fast axis of the wave plate is vertical and horizontal. It is expected that both power measurements would be the same. This can be seen by considering the polarization orientation of the light output by the HWP. The light output by the HWP should be the same when HWP's fast axis is oriented both vertically and horizontally. (Due to this, the method described in this video is not appropriate for distinguishing the fast and slow axes of a HWP.) The measurements you obtained with the HWP in the setup indicates that your setup may differ from the one we demonstrated in unknown ways. We advise comparing your setup with ours and contacting Tech Support (techsupport@thorlabs.com , or www.thorlabs.com/locations.cfm) if you'd like to discuss it further.
@@thorlabs Hello! I am sure the transmission axes of the generating and analyzing polarizers are crossed. I used the QWP (WPQ05M-808) to test. But I think this QWP is an old component. The engraving on the QWP is only "FAST/". Not "FAST/Axis". I see that the powers are almost the same (like 8.10 mW and 7.82 mW) for the HWP. We used a 810 laser and the set-up is almost like yours. I think I need to send an email to your Tech Support.
Hi if I use a laser initially vertical polarized and i place a PBS IS there any horizontal light After this PBS?
@fethallahkarim-pj7so We expect that the polarizing beamsplitter (PBS) would transmit some horizontally polarized light. This is because linearly polarized laser sources do not provide perfectly linearly polarized light, and because the PBS does not perfectly separate the two orthogonal orientations of the polarized light.
Most light output from a linearly polarized light source is polarized in the desired direction, which is vertical in your case. However, it is typical that a small fraction of the output light is polarized in the orthogonal direction, which is horizontal in your case. The amount of laser light polarized along the desired direction is commonly specified using the polarization extinction ratio (PER). This ratio is calculated by dividing the intensity of the light polarized along the desired direction by the intensity of the light polarized along the orthogonal direction. Ideally, the PER would be infinite.
Similarly, the performance of a PBS is described by its extinction ratio (ER), which is typically measured and specified for only the transmitted direction. One approach to measure the ER for the transmitted direction is to place the PBS in a highly polarized beam of light, measure the power in the transmitted beam as the PBS rotates 360° around the optical axis, and divide the average of the two highest powers by the average of the two lowest powers.
It is possible to estimate the amount of horizontally polarized light that passes through the PBS in the case you brought up. For example, consider vertically polarized input laser light with a PER of 15 dB. Assume that vertical polarization is aligned to the P-polarized direction of the PBS, the ER of the PBS’ transmitted path is > 1000:1, and the reflection efficiency of S-polarized light in the PBS is >95%. Since the laser’s PER is 15 dB, approximately 3.2% of the input light is expected to be in the orthogonal (horizontal, unwanted) polarization direction. With an ER >1000:1, we would expect no more than 0.0032% of the incident power to be horizontally polarized after traveling straight through the cube. With a reflection efficiency >95%, we would expect about 3% of the incident power to be reflected from the cube as horizontally polarized light.