Week of 01/22/2018 - Progress Report 9

Tamara Jovanovic
Jan 30, 2018
8 min read

Tamara Jovanovic

Topics covered: Final Design and 3D Printing of the Cuvette Stand, recapturing the spectrum in free space, gaining power in mW, initial results of spectrum through distilled water, power readings and spectral data of distilled water

Materials used: SolidWorks Software, StrataSys F270 3D Printer, Optical System Setup

After the rotational stage was finally delivered at the beginning of the week, final edits to the SolidWorks Design were made to the soon-to-be printed cuvette stand. The final design is shown in figures below. The idea was to make a cuvette stand that will be easy to use as soon as it is printed, so two holes were designed at each end. After measuring the diameter of the holes on the rotational stage, it was ensured that the holes on the cuvette stand would be bigger so that a screw can go in them for more stability.

A StrataSys F270 3D Printer was used for 3D printing. After the design from SolidWorks was finalized, it was transferred to the 3D printer’s software and printing began. After some time, the final product was obtained. It is shown in the picture below. It fits perfectly on the rotational stage and a cuvette goes inside of it perfectly as well.

StrataSys 3D printer used for printing.

Printing in process.

Cuvette Stand - final product of 3D printing.

Final product of 3D printing with a cuvette fitting inside of it perfectly.

The cuvette stand fitting perfectly on the rotational stage and in the optical system.

After it was confirmed that the 3D stand was a perfect fit for the system and that no alterations needed to be made and redone, I went back to checking the power at the output collimator of the whole optical system in free space. By adjusting the knobs on the collimator mounts, the maximum achieved output power in free space at the second collimator was 1.65 mW. Considering that the input power from the laser is about 4 mW, this is roughly the amount of power expected to be measured at the output collimator.

1.65 mW power output in free space.

I then looked at the spectrometer to see the spectrum obtained by this power at the output. The linear and log scale results are shown in the two pictures below.

Free space log scale spectrum.

Free space lin scale spectrum.

After this was achieved, a cuvette was cleaned and filled with distilled water and put in the 3D printed cuvette stand on top of the rotational stage. As expected, most of the power was gone, and the reading at the output was about 2 nW. This was a huge power attenuation and something needed to be done about that. I started adjusting the knob on the rotational stage to achieve the maximum angle of refraction on the cuvette with distilled water. The power started to go up. Then, I started to move the rotational stage closer to the output collimator, while at the same time still adjusting the knob on the rotational stage, as well as the knobs on the collimator mount, ensuring I can capture the most amount of power. After some time and fine tuning of the system, the power output at the second collimator was 0.835 mW. The reading of this success is shown in the picture below. This was an amazing achievement, given that it is exactly half of the power output in free space. This is what, hopefully, I was expecting to see. I wanted to double check that the collimators weren’t maybe still just facing each other and missing the cuvette and/or giving me readings from an angle or edge of cuvette in free space, so I removed the cuvette from the cuvette stand and the power reading went back to 1.6 mW. When I put the cuvette back in the middle of the 2 collimators, it went down to 0.8 mW. This ensured that the readings I achieved were accurate.

Power reading through distilled water.

Final set up of the system with 0.83 mW of output power.

How power was measured through the output collimator.

After these power readings were achieved, I wanted to check the spectra on the spectrometer by plugging in the output fiber from the second collimator to the spectrometer. The spectra I obtained was exactly one half of what I got in free space - which was an expected and warmly welcomed outcome. The results are shown below.

Log scale spectra through distilled water.

Lin scale spectra through distilled water.

All of these results are good and they are a very valuable milestone for our project. For next week’s progress, more adjustments should be made to the optical system to try to get even larger output power in free space, which would result in larger output power through substances. I will also work on maximizing the dBm loss. Ideally it should be -3 dBm, but in these readings, it was -20 dBm at best.

Ezequiel Partida

Topics covered: Light research on supercontinuum generation. Trying to setup GPIB interface again. Played with optical setup a bit.

Materials used: Online. Lab Computer. Optical Setup

This week, Tamara finished printing the 3d cuvette holder and placed it in our system with the rotational stage. Prior to her starting to adjust the system and trying to get an output, I went into the lab and set up the adjusting to have 1.8 mW of power output. I tried to get less than a 3 dB loss, however, I was unsuccessful.

After Tamara got 1.6 mW and then a 0.8 mW retention with he cuvette holder in place with water, we were certain that the rotational stage was helping a lot. Previously, when we placed a cuvette with water, we would get mostly noise with no defined spectrum, as shown below.

Thankfully, we now get a spectrum with water, as shown below, even if the power is not so great.

I went into the lab and played with the system after she did for a while, however, I was not able to improve the power output. However, it was my first time using the rotational stage, so I was not very well trained on how sensitive the rotations were. For next week, I will focus more on getting better with the system so we can start getting good glucose data.

Additionally, this week I tried to once again set up the new computer to interface with the GPIB driver. I brought my laptop, which is currently the only system that can interface with the GPIB and spectrometer. When connected, on my laptop, I get the following:

My Laptop’s Control Panel

ON the lab computer’s control panel, I get the following:

Lab Computer Control Panel

The reason why my laptop works, is because when the GPIB converter shows up as a GPIB interface, it shows up under Agilent’s devices. Additionally, when opening Agilent’s IO Expert, the device shows up under there as well, where one can successfully send and receive commands from a connected instrument (shown below).

Agilent IO Expert

However, on the lab computer, the driver that is installed is from Keysight technologies. Perhaps, this might work if Keysight’s IO Expert recognizes the device as an IO instrument, however, the Keysight IO Expert does not even load, as shown below. After this, I uninstalled Keysight and reinstalled it, however, the instrument interface still does not load. Since this is not loading, there is no way to configure the GPIB devices in order to remotely control our Spectrometer.

One last thing that I worked on this week was research on non-linear fibers and photonic crystal fibers. These types of fibers allow for supercontinuum generation, which can enlarge the wavelength range for a laser source. This is useful for us because we can either use our 780 nm or 1550 nm femtosecond laser source and expand its range to be able to use the range for better glucose detection and inspection. After reading ThorLabs’ guide on supercontinuum generation I learned a few things about such fibers. First, I learned that in order to properly generate a larger spectrum without much loss of power or other problems, the laser source should be below the lower half on the intended wavelength range. For example, if we were to use a 1550 nm laser source, the intended range should have more than half of the wavelength range above 1550 nm (e.g. 1400-1800 nm). If we were using the 780 nm source, a non-linear fiber capable of 400-1600 nm should be sufficient. Also, laser pulses apparently have to be sufficient, but I am assuming that our high tech laser should have no problem with that, given that it is a supercontinuum laser source. Another thing to note was that nonlinear fibers really do not work well in non-clean room environments. Therefore, the quality of the environment can greatly affect the quality of the laser outputs. One recommendation from ThorLabs was to buy a fiber cleaver, however those are thousands of dollars. Another alternative is to use sealed caps on the ends of fibers and make sure that tips are always clean.

Given this information, I went online and looked for possible fibers to buy for our system. Based on the wavelength requirements, I was able to find 2 options. The options are shown below.

Based on specs, it would be better to buy the first option, since it provides a greater dispersion wavelength and it has better reviews online. However, the price is way too expensive. Another problem I found is that I did not find any cheap options online. I even went on Corning, and I did not find any fibers that were non-linear or PCF. Even on ebay I could not find any vendors. Perhaps these fibers are too complex and that’s why they are expensive, but I could just not find any cheap options.

Plans for next week:

More research on NL/PCF.
Continue trying for remote access of Spectrometer.
Glucose data/ practice with rotational stage/ spectra

Jonathan De Rouen

Topics covered: Rotational Stage, data collection/ improvements, GPIB cable: interfacing with computer

Materials used: Spectrometer, Lab computer, NI GPIB cable, Agilent GPIB cable

This week I worked on interfacing the lab computer with the spectrometer through the GPIB cable again as well as becoming more familiar with our setup and collection of data. I attached the rotational stage to the optical table in order to have a more stable device in order to collect data accurately. However due to the bulkiness and missing the blue screwdriver set I was only able to attach one of the screws into the optical table. However I am using painters tape as a reference on the table for where the rotational stage should be set. This was to prevent a problem where I would cause the stage to slide across the table and to have a more definite and solid reference. From there I adjusted the rotational stage and obtained some sample data using a regular cuvette and a cuvette with water.

Absorption spectra of Cuvette without water Db /nm

Due to the obvious change in the absorption spectra, it confirms that the light of the laser is passing through the cuvette and through the sample that we place i. I also found that the closer to the receiving collimator from the sample, I was able to retain more power retention, passing through the samples which can be used to optimize our setup for better samples. All the data obtained in this part of the report is at low power, for reference. The cuvette stand was in the middle of the two collimators and the power out of the output collimator was roughly 50 nW.

I also briefly worked on interfacing the computer with the GPIB cable again. Unfortunately I came across the same error as before where the computer in the lab recognized the device as a separate device and not in the communication port. I tried using the NI GPIB cable but unfortunately didn’t have time to install the drivers. I am also uncertain if they will work with matlab due to the specifications on the supported languages of the system. However I will attempt to install said drivers before giving up on interfacing through GPIB with the spectrometer on the Lab computer

For Next Week: