Week of 10/30/2017 - Progress Report 4

Topics covered: Calibrating the system to test the theory while using a 980nm pulse laser and a SI-Photodetector diode

Materials used: Optical breadboard, Tektronix Osciloscope, FLUKE PM5138 function generator, conducting wires, SI-Photodetector diode, 980nm pulse laser, sturdy metal optical board, external voltage source

The objective for this week was to verify in more detail why the 980nm pulse laser would not be a good choice to use for monitoring the behavior of glucose by observing changes on the Tektronix oscilloscope. The circuit from last week was rebuilt, using a laser, a SI-Photodetector diode, jumper wires and a resistor. The picture of what the circuit looked like is shown below. At first, the sole idea was to recreate the project done in the previous years. However, after some brainstorming, I came to a realization that in order for the circuit to work, it must be biased. The external voltage power supply was used as source voltage.

The resistor used in this circuit is 47kOhms. The decision to use this value of a resistor is made based on the research done by students in previous years. The 980nm laser is a pulse laser, which means that it can only work when it is connected to the function generator. The laser has 3 wires coming out of it:

1) the black one, which was connected to ground,

2) the red one, which was connected to the +3-5V VCC, and

3) the yellow one, which was connected to the function generator.

The function generator was set to 1MHz, so that the waveforms on the oscilloscope can be clearly visible with the bare eye. The wave was set to be a square wave, at a 50% symmetry and at 2.8Volts AC peak-to-peak. The picture of the function generator and its settings used in this experiment is shown below.

The Tektronix Oscilloscope was connected across the biasing resistor and the waveform was observed. I was expecting to see a square wave there, since it is the output, and the input from the function generator was also a square wave. The input of the function generator was connected to another channel of the Tektronix oscilloscope, just to observe the waveform and to make sure that I am inputting the correct wave. The picture below shows a snapshot of the Tektronix oscilloscope with a yellow and a blue output. The yellow line, which is a square, represents the input created by the function generator. The blue one represents the output, across the biasing resistor.

As it can be seen from this picture, I wasn't getting a square output waveform from the output of the circuit. I was inputting the correct thing, but could not get the correct output. It can also be seen that the output and even the input of the function generator are both very noisy.

After some consultations with our project advisor, I decided to do more research on biasing circuits for a photodetector and decided to come up with a new design for a biasing resistor circuit for the experiment. I looked up the data sheet of the photodetector online and found an idea for biasing the circuit, as well as dealing with noise issues. The biasing circuit is shown in the picture below.

I tried building this biasing circuit, as well as using different values of resistors. That wasn't working. I still wasn't getting the correct waveform at the output. Then, I used a potentiometer instead of the resistor and tried to fine tune a circuit that way. I observed no changes in this either.

Then, I moved on to making sure all the components of the circuit were connected correctly. The picture below shows the pinout of the photodetector. All the connections were double checked, and I made sure no mistakes were made.

All of this was a very time demanding process, so unfortunately, by the end of this week, I didn't succeed to show the waveform at the output of the 980nm laser is a square wave. This should continue to be done over the course of next 7 days in order to prove this theory.

For next week, more thorough research about simple biasing circuits for photodetectors should be done and this theory should be proved concluding next week. Also, since the input and the output are very noisy even when using a noise filter, the data obtained on the Tektronix oscilloscope should be recorded on a flash drive, exported and smoothed in Matlab. This way, we will be able to observe the waveforms of glucose solutions of different concentrations more accurately and show why it is better to observe the behavior of glucose at 1550nm instead of 980nm.