Gautam Narasimhan and Ashwin Shankar

Background

The goal of this project was to build an inexpensive photo-spectrometer and compare its performance to the expensive PRS650 in the VISTA lab. The parts chosen for the design are simple off-the-shelf components that are available at any office supply store. While expensive lab-grade spectrometers are used for research purposes including astronomy, gas analysis and crystallography, the inexpensive spectrometer will be used by amateur users for purposes like brewing, checking the fat content in milk and examining soil samples. The inexpensive photospectrometer described below costs less than $10. Components were obtained and an apparatus was designed to focus light on the diffraction grating. The picture of the diffracted pattern was taken with the help of a phone camera and the Spectral workbench software on Public Lab was used to analyze the data.

Optics

Light which has wave-like properties is composed of various colors. A spectrometer is a device that splits light into the various colors it is composed of, which we otherwise cannot distinguish with the naked eye. By viewing a substance through a spectrometer, one can distinguish the exact mixture of colors, which correspond to specific wavelengths of light. These can be compared to other spectra to help identify the sample. The types of spectroscopy depend on the spectrums being examined:

Continuous Spectrum: The colors of the spectrum cannot be distinguished individually. Typically used in crystallography.
Absorption Spectrum: Gases absorb light of a certain wavelength and the spectrum passing through the gas has signature wavelengths missing.
Emission Spectrum: Extraterrestrial objects like hydrogen clouds emit photons of a certain wavelength. This signature can be used to detect the presence of elements in these distant objects.

Apparatus Design

The goal of the designs were to use cheap and easily available items to build the spectrometer. The following items were purchased from an office supply store:
Black Cardboard
Black tape
2 razor blades
1 DVD
30-60-90 Set square

For the purposes of comparison, two designs were built that made use of different diffraction mechanisms:

1) Black Box Design
This design made use of transmissive diffraction. A DVD was used for the diffraction grating. The stickers and silver coating on the DVD were removed to provide a transparent surface for the propagation of the light. Black cardboard was used to make the enclosure. The dimensions of the enclosure are 2 in. x 2 in. x 12 in. On one side of the cardboard box, razor blades were used to make the slit. Razor blades provide a sharp and stable edge that is needed to focus the incoming light. On the other side, the DVD (now transparent) is cut so that the diffraction array appears perpendicular to the slit. The camera will be mounted on the side of the diffraction grating to capture the image of the diffraction pattern.

2) Cereal Box Design
This design made use of reflective diffraction pattern. Light is incident on the DVD diffraction pattern and is reflected off the silvered side of the DVD. The DVD is angled at 60 $\deg$ to the plane of the incoming light. The light hits the diffraction pattern and is reflected off the silvered side of the DVD. The reflected image is captured and is uploaded on Spectral Workbench to examine the spectrum.

Spectra from the above spectra were captured using a cellphone camera. The image of the spectrum was uploaded on the Spectral Workbench software.

The software is first calibrated using a light source without a filter to show positions of the red, green and blue bands. The software scans pixels from and plots the relative intensity for the associated wavelength.

Experimental Method

In order to compare the workings of the simple spectrometer and the PR650 available in lab we pursued two methods of data collection. The setup involved obtaining data from two different light sources and also measuring the same with different color filters.

Here are a few pictures of the setup in lab. The first image shows the PR650 measuring light from the monitor. The second picture shows the inexpensive device measuring light from the incandescent light source with a red color filter.

PR650 setup measuring light from monitor

Black Box device measuring light from incandescent source

The color filters used in the experiment were:

Red
Yellow
Green
Blue

The two spectrometer designs were tested with both light sources. From our initial experiments we noticed that the incandescent source was too bright for the PRS650 to measure directly and while the computer screen even with a white background was not bright enough for the inexpensive device to provide a good spectrum (bright enough to be captured by the camera).

Therefore we also obtained results for the PRS650 and the incandescent light source by focusing on reflected light. The incandescent source was pointed towards a white reflective surface and the resulting light was indirectly captured by the PRS650. The picture below shows the setup for this test.

PRS650 measuring reflected light from incandescent source

The reflected light from the incandescent light source also did not pose as an ideal source for the PRS650. The spectrometer detected large amounts of IR which had the effect of washing out other data from the visible spectrum. This was corrected for in post processing where the IR was digitally filtered out using MATLAB. The resulting PRS650 spectrum provided a good comparison for the inexpensive device.

Results

The PRS650 was first used to measure the spectrum emitted by the computer monitor. The monitor consists of red, blue and green colors. A blue colored filter was used to as a single pass filter to filter out all components of the light except the blue color. The spectrum obtained shows the intensity as a function of wavelength. A peak is seen in the visible range of the blue color.

Spectra using PRS650 and monitor with blue color filter

Next, the same experiment is performed with a red colored filter and a peak is observed around the wavelength for red light.

Spectra using PRS650 and monitor with red color filter

The black box design was now used with the blue and red colored filter. Since the monitor source was too dim to be used with the black box, an incandescent light source was used with the apparatus. The pictures below show the plots for blue and red colored filters. The two pictures show peaks at the respective wavelengths.

Spectra using Black Box and incandescent source with blue color filter

Spectra using Black Box and incandescent source with red color filter

The spectrograph plots shown in the above picture have intensities plotted as a function of wavelength. The intensities are relative (as opposed to the absolute measurements made in the the PRS650 plots). The peaks seen for the red and blue colored filters are not sharp. This could be due to the use of an incandescent light source or errors during calibration of the software.

The cereal box design was then tested with the monitor and the reflected diffraction pattern was sent to the Spectralbench software for analysis However, the calibration of the spectralbench software is built only for transmissive diffraction patterns. Thus, the apparatus could not be calibrated properly and the results show peaking at frequencies beyond the visible light range.

Spectra using Cereal Box and incandescent source with blue color filter

Spectra using Cereal Box and incandescent source with red color filter

In order to make a valid comparison between the working design(black box) and the PRS650, light was made to reflect off a white paper. The intensity of light hitting the spectrometer now decreased. Data was acquired and compared for a spectrum with a blue colored filter.

Spectra using PRS650 with reflected light from an incandescent source with blue color filter

As seen in the above plot, although the reflected light allowed the PRS650 to measure the incandescent light source, it still picks up a significant amount of infrared. This has an overall effect of washing out the rest of the visible spectrum. Therefore, we digitally filtered out the IR spectrum data and limited measurements to 650nm wavelength. This data was then compared to the spectrum obtained from the inexpensive device for the same wavelength range. The plot below overlays the spectrum from the inexpensive and PRS650 using an incandescent light source with a blue color filter. The intensities follow a similar trend across the range of wavelengths. There especially seems to be a good fit in the blue light range (420 to 480nm).

Compared spectra of blue filtered light from incandescent source

Conclusions

Two spectrometer designs were built using materials under $10. The apparatus in conjunction with the spectral workbench software were used to evaluate spectra.These spectrometers were characterized and compared to the lab-grade PRS650. The cereal box design that used reflective spectrometry did not work due to errors in calibration of the spectral workbench software. The black box design gave spectrum data that is comparable to the PRS650. The following table compares the 3 designs for different parameters.

In the future, a light source with variable intensity could be used in order to test sensitive (PRS650) and less sensitive spectrometer designs without changing the test setup. In addition, the designed spectrometer could also be tested with actual application of spectroscopy such as identify liquids through spectral absorption.