Beam shaping using a Digital Multimirror Device

This project was worked on under the supervision of Prof. Lindsay LeBlanc at the University of Alberta. The project involved developing the optical setup for the neutral atom trapping experiment.  I like to divide my work during this time into 3 parts. 

During the first and second months, I was involved in working on the detection of Rubidium-87 in the MOT. The MOT lasers were not perfectly aligned, so the idea was to align them using the feedback of the data collected from the CCD cameras. Essentially, it was a Python GUI that displayed the camera feed in real-time with some unique features. These features included drawing a box on the video feed, summing the pixel values within that box, and plotting the result in a real-time, continuously updating graph at the bottom.  The idea was to make slight adjustments to the laser and observe this plot, and spikes in the pixel sum indicate fluorescence of the atoms, i.e., MOT is being aligned. 

During the second and third months, I worked on the AOD and subsequently the optical setup for the neutral atom traps. One of the key challenges in implementing the AOD was synthesizing the frequency waveform during the experimental cycle for the atom rearrangement¹.  This proved to be a discrete-time signal problem. Generating these waveforms during the experimental cycle was too slow. The solution that could work would be to pre-calculate every possible rearrangement movement and run the desired waveform depending on the random loading pattern. I never got to implementing this, but I did spend countless hours playing with a benchtop AWG at the lab to integrate this setup when we stream a saved waveform on demand, controlled by a Python script.  

During the third and fourth months, I worked on integrating the DMD into the optical setup. I was inspired to dive into this after watching a video on holography by 3Blue1Brown. For some reason, I was convinced I could recreate this in the lab with our 810nm laser. I planned to record the phase map on the CCD. The use of computer simulation to see if I can recreate the image digitally.  The simulation was basically just shining an ideal reference beam on the pattern and propagating the light field. I have made an explaination to these concepts in my "learn from me" section. It did not work; the reference beam I was using was not even close to being ideal, so my ideal beam in the simulation did nothing. Thats my hypothesis of where it went wrong. 

Over the past month, I focused on developing the core idea behind my poster: using holographic techniques to generate custom beam shapes at the image plane of our optical setup. This is achieved by placing a digital micromirror device (DMD) in the Fourier plane and projecting onto it the Fourier transform of the desired field, overlaid with a diffraction grating. At first glance, this approach may seem unnecessarily complex. However, the motivation is that the DMD is an amplitude-only modulation device, and we aim to correct for both phase and amplitude aberrations at the image plane. By operating in the Fourier plane, we gain the flexibility needed to encode phase information indirectly through amplitude modulation. In practice, varying the width of the grating slits or overlaying an aberration-correction map onto the diffraction grating allows us to compensate for both phase and amplitude distortions in the optical system. To enable this correction, we first measure the phase aberrations of the optical setup using a novel interferometric technique. The DMD is divided into multiple patches, one of which is designated as a reference path and remains open at all times. The remaining patches are then sequentially activated. This produces a series of interference patterns, each referenced to the same sample path. By analyzing these patterns—specifically, by extracting the effective slit widths—we reconstruct a phase map of the system. This phase map can then be used to actively correct phase aberrations within the optical setup.

 Finally, during the month of September, I worked on mounting the new NIR DMD and getting it set up. During october I worked on getting it alligned which I partially completed. This was my fall semester in 3rd year of Engineering, so I could not get more done during this time.

Cool pictures from this experience-