Electronic Design of Acousto Optic Tunable All Fiber Spectrometer

Electronic Design of Acousto Optic Tunable All Fiber Spectrometer

Tanuja Ashekin Mouli

University of Arizona, Tucson

Faculty Mentor: Professor Henry P. Lee

Dept. of Electrical Engineering and Computer Science,

Interdisciplinary Material Science and Engineering.

Abstract

This paper discusses the construction of a spectrometer using a fiber acousto-optic tunable filter (AOTF). An AOTF spectrometer was built because it is a low cost portable alternative to a regular optical spectrometer. The AOTF has low insertion power loss, thus giving accurate output, and also is easily tuned and can be electrically controlled. In the proposed spectrometer, a voltage controlled oscillator (VCO) (part #SN54L624) was controlled by a picaxe 18x microcontroller, which applied an RF voltage to the AOTF at a user-controlled frequency. The output of the AOTF is measured by a photo detector (PD) and converted into a voltage, which is then measured by the microcontroller. The microcontroller converts the analog signal to a digital signal, which is viewed on a PC screen, thus giving spectral information about the light source under test. Due to time constraints, the project was not finished completely; however, the microcontroller successfully established an interface between the VCO and the current to voltage converter.

Key terms:

AOTF: Acousto Optic Tunable Filter

PZT: Piezo Electric Transducer

RF: Radio Frequency

VCO: Voltage controlled Oscillator

Introduction

The studies of optical fibers are becoming more significant day by day because of their enormous number of applications, including communication, optic sensors, and medical purposes. Using optical fibers, an acousto-optic tunable filter (AOTF) can be built to create a notch filter to pass wide range of frequencies, while attenuating a narrow band of frequencies (Schers, 217). The AOTF has recently reached technological maturity, moving from the research laboratory to the commercial environment.

AOTFs have been the focus of the research for Bragg fiber grating, multiplexing, etc. To drive the AOTF, a function generator and a RF (radio frequency) amplifier is needed. The function generator could be replaced by a portable chip, which could save space and money.

This research was conducted to create a portable optical spectrometer by replacing a function generator. With this spectrometer, wavelength of any light source can be determined. As the world is advancing with new technologies, it is important to build portable devices so study could be conducted beyond the surface of the Earth. Thus, if a type of light source were detected elsewhere, i.e. Mars, this spectrometer could be used since the device would be portable.

To achieve this goal, a microcontroller and a VCO (voltage controlled oscillator) chip were used to control output frequencies. The basic setup of this project is shown in Figure 1:

Figure 1: Big picture of this project

Theoretical Background:

AOTF spectrometers have very low insertion power loss, broad wavelength, fast scanning speed and manufacturing cost which make them excellent for commercial application of interrogation units in a fiber network (Kazemi, 16).

The setup of an AOTF setup is shown in Figure 2. An all-fiber device creates a wave into an optical fiber that creates a periodic phase grating (Kazemi 4). The wave is generated by a Piezoelectric Transducer (PZT) disk, driven by a RF amplifier with 1Mhz – 5Mhz frequency range. Figure 2 shows the wave created by the PZT:

Figure 2: Acoustic Wave created by the PZT disk

By varying the RF frequency of the PZT, it is possible to scan the wavelength spectrum. Since the RF frequency is electronically controlled within a few Hz of accuracy, the resulting output of the AOTF output is very accurate.

As the fiber optic vibrates, this creates a notch filter, as shown below.

:

This notch filter is used to pass large band of frequencies while attenuating narrow bands. As a result, applying to this project, the actual information about the original wavelength applied can be retrieved.

In this project, the AOTF acts as a sensor to determine wavelength of a light source. Singlemode fiber sensors can be created for measuring an enormous range of physical and chemical variables. Common applications include detecting chemical compounds, particle sizes, vibration, acoustic waves, electric field, fluid flow and temperature (Handerek, 109).

Methods and material:

The main materials used in this project were:

1. AOTF
2. Photo detector
3. Operational Amplifier – used to build the current to voltage converter
4. VCO (SN54LS and MAX038)
5. Picaxe 18X microcontroller
6. Push Buttons

The proposed setup for the AOTF spectrometer is shown in Figure 3:

Figure 3: Setup of the spectrometer

As Figure 3 suggests, a microcontroller sends voltages controlled by a user via push button to the VCO. For each frequency the VCO outputs, the AOTF will attenuate the input light by a different amount. The photodetector outputs a current that is proportional to the total optical power remaining in the fiber. The current-to-voltage converter circuit converts the current to a voltage. The microcontroller then takes this analog voltage value and converts this to a digital data which could be viewed in the computer.

The main parts that the project required to work on were the current to voltage converter circuit, interfacing VCO, the push button with the microcontroller, and writing the code for these interfaces.

Experimental Setup:

Current to voltage converter:

The current to voltage converter was built with an op-amp, according to the design in the book “Practical Electronics for Inventors.” The design is shown in Figure 4:

Figure 4: Current to voltage converter circuit

Different op-amps were tested until the best suitable one was found. The tests were conducted by biasing current using a voltage source with a resistor to provide current. The data is shown in the Results section of the paper.

Interfacing of pushbutton and VCO with the microcontroller:

The VCO used for this project was the Texas Instruments SN74lS624, and this was interfaced with Picaxe18x microcontroller. Since the AOTF can perform in the range of frequencies between 1 MHz to 5 MHz, the microcontroller was programmed to control the voltage of the VCO for that range of frequencies. The frequency range of the VCO is shown in Figure 5.

Figure 5: The frequency control for SN74lS624 VCO

As shown in figure 5, the VCO was controlled with two voltage inputs, VI(rng) and VI(freq). In this project, VI(rng) was set to a constant 5 volts while the VI(freq) was varied using a microcontroller. The microcontroller was a picaxe18X microcontroller, and it was interfaced with a push button to allow user input of the VI(freq). The circuit for the VCO interfacing is shown in Figure 6.

Figure 6: Circuit for VCO

The output ports were connected to the VCO via a resistor and a diode as shown in Figure 6. The resistor values were chosen to give different voltage through each output. The diodes were placed so the current would not flow the opposite direction and create a different voltage than what is expected through the output. The push button was connected to an input port of the microcontroller (in1). The VCO was controlled via the push button. The push button interface is explained in detail in the Code section.

Code

The code was written to interface the push button with the microcontroller to control the VCO. The microcontroller was programmed using basic language. The push button worked as follows: if no button was pressed, it sent out 0V; thus VI(freq) will be 0V and VI(rng) will be constant 5V. Thus, by looking at Figure 5, this would give out a frequency of 1 MHz. If the button was pressed again, VI(freq) would send out 1V, still keeping VI(rng) at 5V. The program allowed the push button to change its value four times, thus increasing up to 4V by each press. When the button was pressed the fifth time, the voltage would go back to zero and start incrementing again by the press of the button. The code written was as follows:

symbol_cond = b0 //name a variable symbol_cond and store this in b0
symbol_new = b1 // name a variable symbol_new and store this in b1
let cond = pin0 //input pin will be triggered by cond

main:
if pin0 = 1 then //If statement: if pin0 is on, (button pressed)
if cond = 1 then//turn on the pin0
new = new+1//5//increment till 4 then go back to beginning
cond = 0//turn the pin0 off so it doesn’t keep incrementing while releasing button
endif//end condition
else//If pin0 is off
cond = 1//keep the same value
endif//end condition

if new = 0 then //if push button not pressed, turn off all the output ports
low 1
low 2
low 3
low 4
low 0

else if new = 1 then//if push button is pressed once, turn off all the output ports but

low 1open output port 0 (out0)//
low 2
low 3
high 0
else if new = 2 then//if button pressed for second time, turn on output port 1, others closed
low 0
low 2
low 3
high 1
else if new = 3 then//if button pressed for third time, turn on output port 2, others closed
low 0
low 1
low 3
high 2
else if new = 4 then//if button pressed for fourth time, turn on output port 3, others closed
low 0
low 1
low 2
high 3
endif//end condition

readadc 1, b2 // read analog value from b2 and send it to input port 1 (in1)
high 4//open output port 4
goto main//go back to main

Results, Problems and Solutions:

Current to voltage converter:

Figure 7: Current to voltage converter test circuit

Figure 7 shows the schematic for the op-amp test circuit. A series of tests were conducted between OPA602, LM741, and OPA27GP, and the op-amp that performed the best was selected. Ideally, Vout should be equal to Iin. The data is shown in Tables 1, 2 and 3.

Table 1: Results for OPA602 op-amp

Vin (V) / Vout (v) / Iin(mA)
1.08 / 0.54 / 0.54
1.56 / 0.78 / 0.78
2.06 / 1.03 / 1.03
3.10 / 1.55 / 1.55
4.15 / 2.08 / 2.075
5.22 / 2.61 / 2.61
6.20 / 3.11 / 3.1
7.27 / 3.64 / 3.635
7.97 / 4.00 / 3.985
10.85 / 5.44 / 5.425
12.98 / 6.51 / 6.49

Table 2: Results for LM741 op-amp

Vin (V) / Vout (v) / Iin(mA)
1 / 0.5 / 0.5
1.98 / 1.14 / 0.99
2.77 / 1.43 / 1.385
3.83 / 1.95 / 0.095
4.75 / 2.39 / 2.375
5.64 / 2.79 / 2.82
6.22 / 3.02 / 3.11
7.12 / 3.02 / 3.56
8.50 / 3.33(unstable) / 4.25
10.07 / Out of control / 5.03

Table 3: Results for OPA27GP

Vin (V) / Vout (v) / Iin(mA)
0.081 / 0.42 / 0.54
1.71 / 0.86 / 0.855
3.47 / 1.74 / 1.735
4.76 / 2.40 / 2.38
5.98 / 3.05 / 2.99
6.92 / 3.57 / 3.46
8.63 / 4.49 / 4.315
10.12 / 5.30 / 5.06
12.34 / 6.45 / 6.17
13.40 / 7.01 / 6.7
17.64 / 9.11 / 8.82 (op-amp hot)

By comparing Tables 1, 2 and 3, it is clear that Table 1 was stable throughout and gave more accurate values, i.e. (Vout = Iin). After the right op-amp was chosen, the circuit was built and a code was written to convert analog signal to a digital signal as described in the previous section.

Push Button and VCO Interfacing

At first, the plan was to interface a keyboard with the microcontroller. However, due to time constraints, a push button was used instead. One of the problems that was encountered was that even though the microcontroller was programmed successfully to control the VCO, the VCO was not giving out a clean sinusoidal; the output contained a lot of noise. Different values of capacitors were used to fix this problem and the signal became clearer. However, it could not be fixed completely. To solve this problem, a different VCO chip was used (MAX038). This gave out a noise-free sinusoidal. Nevertheless, when it was connected to the microcontroller, the chip got hot. One of the assumptions made was that too much current was going into the input of the VCO chip. This could be fixed by putting a Bipolar Junction Transistor (BJT) on the input of the VCO to regulate the current.

Code

A few problems occurred while writing the code. At first, the code consisted of a lot of lines and had a lot of loops. This created delay and, as a result, it took a while for the microcontroller to detect when the button was pressed. This was fixed by writing the code more efficiently and reducing it to 10 lines.

Another problem was that the output port would change twice with a single press. To clarify, the output port changed at a press of the button and changed again after releasing the button. This was unwanted since the output port is supposed to change at each press of the button. This was fixed by adding the extra logic (cond) shown in the code section. The final circuit is shown in Figure 8:

Figure 8: Final Circuit for the AOTF All Fiber Spectrometer

Discussion:

The project went well but, due to time constraint and other obstacles, the project was not finished. The VCO and the push button were controlled by the microcontroller successfully. Even though the VCO was getting hot, this problem could have been solved easily. Sadly, the final data could not be viewed on a PC screen since that required learning LabVIEW—a program that interfaces electrical devices together. If there had been more time, this could have been accomplished. Even though the project was not finished, a lot of knowledge and skills were obtained from this project.

Acknowledgements: I would like to acknowledge Said Shokair for directing IM-SURE and also providing support and encouragement. I would also like to thank NSF and UROP for funding us and giving undergrads and opportunity to conduct research. Furthermore, I would like to acknowledge Dr. Henry P. Lee for giving me an opportunity to work with him. Also, I would like to thank Fares Alhassen, Patrick Chan and Kevin Edmonds for helping out with my project, giving me suggestions and providing support.

Works Cited:

Sun, T, Handerek V. A. and Grattan K.T.V.Kluwer. London:Academic Publishers, 2000.

Scherz, Paul. Practical Electronics for inventors. 2nd Edition. New York: McGraw-Hill Companies, 2007.

Kazemi, Saba. “AOTF”. 2004

1