- #1
Choisai
- 26
- 1
Hi there!
I'm interested in Spatial Light Modulators for my project and I was hoping you (whoever you are) could help me. I try to explain everything as compact as possible but it's still quite a story. Sorry for that!
For spatial modulation of light, a 'spatial light modulator' is needed. These are abbreviated as 'SLM' and come in many forms, such as: liquid crystal SLMs (working basically as LCD screens), Micro Optical Electric Mechanical Systems (MOEMS or MEMS) and Deformable Mirrors (DM's). Also available are DOE's (Diffractive Optical Element). These however are not actually SLMs because SLMs are usually dynamic, where as the DOE's are static.
My goal is to use a 2 mW 589 nm laser to image a Bose Einstein condensate. The current width is around 1 cm. We use several imaging techniques, but most used of all is the absorption technique. We shine the laser on the condensate which casts a shadow on the camera that is behind it. The intensity profile showes us the BEC and tells us much about its properties. Imaging takes place by taking one picture without the condensate and one (or several) with the condensate. Dividing the condensate picture by the picture with no condensate filters out the background.
However, the current beam has to go through many lenses and mirrors, all with small aberrations, small amounts of damage etcetera. This causes fringes to appear in our picture of the intensity profile which need to be filtered out.
The beam is currently also somewhat Gaussian. That gives us an inhomogenous intensity profile which clouds our judgement and calculations. So we need that to be a tophat beam profile.
For this we want to use an SLM. What we expect the SLM to do is the following:
every morning we can run an algorithm to turn our 'dirty Gaussian' beam into a 'clean tophat' beam which we can then use for the rest of the day to image the condensate.
So we are now looking for suitable SLMs and what type is suitable. Deformable Mirrors are out of our budget (the Thorlabs kilo DM costs 100 000 USD) so that's no option.
So LC SLMs and DMD's remain.
The bad thing with LC SLMs is that they have to be refreshed because otherwise the molecules start to relax and their phase and amplitude shifting properties become less accurate. Boulder Nonlinear Systems SLMs repress the droop in phaseshifting caused by this relaxation by using a refreshrate of 6 kHz. They however, have a low resolution (max 512x512). Hamamatsu SLMs have a large resolution but a refreshrate of 480 Hz, which causes an error of 0,011 pi in the phaseshifting.
The other bad thing about LC SLMs is that their refreshing causes 'blinking' of the light. Shining with 'blinking' light instead of a continuous beam on the condensate will cause it to heat up. And heating up the condensate is definitely not something we want!
A solution to our problem might be Digital Micromirror Devices. These are of the category 'MOEMS' (also called MEMS). These do not have the 'blinking' effect LC-SLMs have.
Proof that these can produce top hats and are more suitable for cold-atom physics is found in a paper of the university of München:
http://greiner.physics.harvard.edu/PDF%20Files/zupancic_thesis.pdf
He uses spatial modulation for not only top-hat generation, but also for fun stuff, like the producing of the Batman symbol (as can be seen in the thesis). We just want a circle shaped beam with a top hat intensity profile and no distortions in the beam profile.
Now my question is: what is more suitable for high precision top hat generation? The LC-SLM or the DMD-SLM? LC-SLMs have the 'blinking' but DMD-SLMs don't have a grayscale (it's tiny mirrors are either on or off). Which one is more suitable if we want to have an as clean as possible top hat beam profile?
I'm interested in Spatial Light Modulators for my project and I was hoping you (whoever you are) could help me. I try to explain everything as compact as possible but it's still quite a story. Sorry for that!
For spatial modulation of light, a 'spatial light modulator' is needed. These are abbreviated as 'SLM' and come in many forms, such as: liquid crystal SLMs (working basically as LCD screens), Micro Optical Electric Mechanical Systems (MOEMS or MEMS) and Deformable Mirrors (DM's). Also available are DOE's (Diffractive Optical Element). These however are not actually SLMs because SLMs are usually dynamic, where as the DOE's are static.
My goal is to use a 2 mW 589 nm laser to image a Bose Einstein condensate. The current width is around 1 cm. We use several imaging techniques, but most used of all is the absorption technique. We shine the laser on the condensate which casts a shadow on the camera that is behind it. The intensity profile showes us the BEC and tells us much about its properties. Imaging takes place by taking one picture without the condensate and one (or several) with the condensate. Dividing the condensate picture by the picture with no condensate filters out the background.
However, the current beam has to go through many lenses and mirrors, all with small aberrations, small amounts of damage etcetera. This causes fringes to appear in our picture of the intensity profile which need to be filtered out.
The beam is currently also somewhat Gaussian. That gives us an inhomogenous intensity profile which clouds our judgement and calculations. So we need that to be a tophat beam profile.
For this we want to use an SLM. What we expect the SLM to do is the following:
every morning we can run an algorithm to turn our 'dirty Gaussian' beam into a 'clean tophat' beam which we can then use for the rest of the day to image the condensate.
So we are now looking for suitable SLMs and what type is suitable. Deformable Mirrors are out of our budget (the Thorlabs kilo DM costs 100 000 USD) so that's no option.
So LC SLMs and DMD's remain.
The bad thing with LC SLMs is that they have to be refreshed because otherwise the molecules start to relax and their phase and amplitude shifting properties become less accurate. Boulder Nonlinear Systems SLMs repress the droop in phaseshifting caused by this relaxation by using a refreshrate of 6 kHz. They however, have a low resolution (max 512x512). Hamamatsu SLMs have a large resolution but a refreshrate of 480 Hz, which causes an error of 0,011 pi in the phaseshifting.
The other bad thing about LC SLMs is that their refreshing causes 'blinking' of the light. Shining with 'blinking' light instead of a continuous beam on the condensate will cause it to heat up. And heating up the condensate is definitely not something we want!
A solution to our problem might be Digital Micromirror Devices. These are of the category 'MOEMS' (also called MEMS). These do not have the 'blinking' effect LC-SLMs have.
Proof that these can produce top hats and are more suitable for cold-atom physics is found in a paper of the university of München:
http://greiner.physics.harvard.edu/PDF%20Files/zupancic_thesis.pdf
He uses spatial modulation for not only top-hat generation, but also for fun stuff, like the producing of the Batman symbol (as can be seen in the thesis). We just want a circle shaped beam with a top hat intensity profile and no distortions in the beam profile.
Now my question is: what is more suitable for high precision top hat generation? The LC-SLM or the DMD-SLM? LC-SLMs have the 'blinking' but DMD-SLMs don't have a grayscale (it's tiny mirrors are either on or off). Which one is more suitable if we want to have an as clean as possible top hat beam profile?
Last edited by a moderator: