Jared Vawter, Kevin Clark, Steven Kalman
Acoustic Positioning System
Jared Vawter
Kevin Clark
Steven Kalman
Acoustic Positioning Device Last Saved: 4/6/2005 12:13 PM
E-File: 050404 Elec499 REPORT.doc Last Printed: 0/0/0000 0:00 AM
Jared Vawter, Kevin Clark, Steven Kalman
Table of Contents
Abstract 3
Summary 4
Glossary 5
1 Introduction 1
1.1 Motivation 1
1.2 Problem definition/Solution 1
2 Theory Behind the Acoustic Positioning System 2
2.1 Position location 2
2.2 Phase Detection 4
3 Implementation 4
3.1 Transmitter/Receiver Transducer 4
3.2 Phase Detector 6
3.3 Small Signal Amplifier 6
3.4 Zero Crossing Detector 7
3.5 DC Output Stage 7
3.6 Software design and considerations 7
3.7 Experimental Apparatus 8
4 Experimental Data 9
4.1 Operating Distance 9
4.2 Distance Measurement Stability 9
4.3 One Dimensional Distance Accuracy 9
4.4 One Dimensional Distance Accuracy Versus Speed 9
4.5 Two Dimensional Distance and Angle Accuracy 9
4.6 Two Dimensional Distance and Angle Accuracy Versus Receiver Separation 9
5 Recommendations/Future Developments 9
6 Conclusions 9
7 References 10
Table of Figures
Figure 1: (a) Phase difference between transmitted and received signals (b) Output voltage characteristic proportional to the difference in this phase 3
Figure 2: 2-D position location arrangement 3
Figure 15 - Transducer Patterns for: (a) the Transmitter [2], (b) the Receiver [3] 4
Figure 16 - Top View of the 16mm (Left) and 12mm (Right) Transducers 5
Figure 3 - Side View of the 16mm (Left) and 12mm (Right) Transducers 5
Figure 3: Small signal amplifier circuit 6
Figure 4: Software block flow diagram 8
Table of Tables
Error! No table of figures entries found.
Table of Appendices
Appendix A: Transducer Datasheets
Appendix B: Small Signal Amplifier Design Calculations
Appendix C: Code for Base Station
Abstract
Summary
Glossary
A/D Analog-to-Digital
LCD Liquid Crystal Display, to display text from the microcontroller
Limacon Software Software company established in 2002 by Kevin Clark and Jared Vawter
Limasonic Team Name for Group 5, ELEC 499
PIC Microcontroller produced by the Microchip corporation
ROM Read Only Memory
Vcc Supply voltage, 5V
USBL Ultra Short Baseline, meaning an acoustic positioning device that has its receivers in a single cluster, opposed to in long lines, or distributed along the surface of a ship’s hull.
AUV Autonomous Underwater Vehicle
Acoustic Positioning Device v Last Saved: 4/6/2005 12:13 PM
E-File: 050404 Elec499 REPORT.doc Last Printed: 0/0/0000 0:00 AM
Jared Vawter, Kevin Clark, Steven Kalman
1 Introduction
1.1 Motivation
Acoustic positioning devices are useful for describing an objects position is space. An application for this technology is underwater navigation for objects such as autonomous underwater vehicles (AUV). For a positioning system to be effective for an AUV, the system needs to be able to describe position in 3 dimensions with accuracy. As with most systems, the more economical the system is the better.
One commercially available acoustic positioning system is the acoustic range finder. There are both complete systems and electronic kits available for sale, for around $40. These inexpensive systems use 40 kHz sound and measure travel time for a pulse to give range. This is very economical, but does not allow for constant tracking and only measures in one dimension.
Another system is the Ultra Short Baseline (USBL) acoustic positioning system. USBL systems use a single group of close-spaced receivers unlike their long baseline (which use long lines of receivers on the ocean floor) and short baseline (several receivers mounted across the hull of the tether vessel) competition. Though USBL systems do not have the long range of the long baseline or short base line systems, they are cheaper to deploy and more accurate for close range measurements than either system. Commercially available USBL systems are generally high priced (in the order of $100,000) and have accuracy of ±10 cm at 100m.
1.2 Problem definition/Solution
An acoustic positioning system that could have high accuracy and low cost would be very desirable. This proposed solution, once modified for underwater use, could have a much higher accuracy and low cost. The accuracy of the system is dependant on the phase detection accuracy, usually on the order of degrees. With a sound wave length in water of 1.5 cm, even 10º phase accuracy the error would be within ±0.41mm.
2 Theory Behind the Acoustic Positioning System
2.1 Position location
The position of an acoustic source can be located in space by comparing the phase shift between the transmitted signal and the signal observed at the receiver. By processing both the transmitted and received signals with the appropriate electronic circuits, an output waveform characteristic can be generated which relates the magnitude of the waveform to the difference in phase between the signals. Therefore, measuring the instantaneous voltage at the output of the receiver circuit will enable the position of the acoustic source to be accurately measured. Figure 1 shows the input/output waveform characteristic for the receiver circuit discussed in sectionXX.
(a)
(b)
Figure 1: (a) Phase difference between transmitted and received signals
(b) Output voltage characteristic proportional to the difference in this phase
Assuming the transmit frequency is exactly 40 kHz and the speed of sound to be 344 m/s in air at room temperature, the wavelength of the transmitted and received signals is 8.6 mm. Thus, as the difference in phase between the transmitted and received signals becomes 2π radians, the acoustic source is considered to have been displaced by one full wavelength (i.e. it has been displaced by 8.6 mm).
Add in a blurb about counting phase crossovers
By using two receivers, the position of the acoustic source can be accurately obtained in 2-dimensional space. By constructing two identical receiver circuits and monitoring the output voltage of both, the distance between the acoustic source and the two different receivers can be determined. Figure 2 shows the configuration for a 2-dimensional position location set-up.
Figure 2: 2-D position location arrangement
As shown in the figure, each receiver will measure a different distance from the acoustic source. The receiver closest to the source will observe the distance ds and the receiver furthest from the source will measure the distance dl. If the two receivers are positioned equal distance from the center line, the true distance from the center, R, can be calculated as well as the angle from the center line Ψ. Calculations are given below to determine the distance R and angle Ψ.
In the calculation of Ψ, angles left of the center line in Figure 2 are considered to be positive, while angles to the right are assumed negative.
2.2 Phase Detection
3 Implementation
3.1 Transmitter/Receiver Transducer
The 40 kHz transducers for this project were chosen with the following considerations in mind. First and foremost was availability. Because the project had a 3 month timeframe, all transducers with long back orders were instantly rejected. The second was cost. 40 kHz is a popular transducer frequency for use in range finders and other hobbyist activities, such as bat detecting, wind speed monitoring, and level measurements, but there are limited selections on the market. The range of transducer pricing went from around five dollars for low power models, to thousands for high power models designed for sonar applications. The third consideration was beam width. A very directional transducer would cause problems when trying to detect position in two or three dimensions, where the transmitters are not necessarily pointed directly at the receivers.
After weighing these considerations and researching what was available on the market, the Kobitone 400ST16/400SR12 series from Mouser Electronics was chosen. These were not the optimal solution, but they were available, inexpensive, and had a fairly wide beam. Unfortunately, there were no 400ST12 transmitter models available, and the 400ST16 was ordered instead. The larger case size increased the directionality of the transmitter as can be seen in Figure 3.
(a) /
(b)
Figure 3 - Transducer Patterns for: (a) the Transmitter [2], (b) the Receiver [3]
In early testing, though the receivers could pick up a signal from many different angles, the transmitters needed to be aimed directly at the receiver. Upon looking inside the transducers, it appears the only difference is the aluminum case is larger, and to make the radiation pattern of the transmitter become wider the only modification would be to make the case smaller. This similarity can be seen in both Figure 4 and Figure 5.
Figure 4 - Top View of the 16mm (Left) and 12mm (Right) Transducers
Figure 5 - Side View of the 16mm (Left) and 12mm (Right) Transducers
Modifications could be made to try to change the beam width of the transmitters, but because of the precision required and the time constraints this was not attempted. Another limitation is the Kobitone transducers have a bandwidth of 40 kHz ±1 kHz. This low bandwidth would limit future developments, such as using spread spectrum techniques to limit multi-path interference. There are transducers available on the market that could be used, such as the Panasonic EFR-TQB40K5 and EFR-RQB40K5 transmitter and receiver pair[4]. These are more expensive than the Kobitone transmitters, but offer wider beam width, for the Q package, or a wider bandwidth for the S or U package.
The full datasheets for the Kobitone transducers used the project and the alternative Panasonic transducers can be found in Appendix A.
3.2 Phase Detector
The phase comparator used in this project uses two D-type flip flops. As described in section xx, the outputs of the flip flops will go high on a leading edge of the input signal, and will remain high until they are reset. They are reset when both signals are high. The output signal will be a signal with pulse width proportional to the time difference in rising edges of the input signals. The circuit diagram for this can be seen in Figure 5.
Figure 5 - Phase Comparator Circuit
3.3 Small Signal Amplifier
In order for the receiver circuit to operate properly, it requires the received signal to be sufficiently large. The small signal amplifier, shown below in Figure 1, serves as the front-end to the receiver circuit, providing large amplification (approximately 150x) of the received signal to achieve the sufficient level needed by the rest of the receiving circuit.
Figure 6: Small signal amplifier circuit
The small signal amplifier was implemented using the common-emitter design because of its high gain factor and relatively low noise component. The circuit is built upon an NPN transistor and consists of a network of resistors and capacitors to provide the appropriate biasing voltages, AC coupling, and desired gain. Calculations for the selected component values are given in Appendix B.
3.4 Zero Crossing Detector
3.5 DC Output Stage
3.6 Software design and considerations
The software component of the project serves two important roles: to process the output of the phase detector circuit and to interface the results to an LCD display. Software for the PIC16F688 microcontroller was developed using the CCS compiler, MPLab, and the PICKit interface program.
The PIC16F688 was chosen because of its availability, low cost, and sufficient internal operating speed. One primary downfall of this choice, however, was that it had a very limited amount of program memory (just over 7 kilobytes of ROM). The entire program, once compiled, accounted for approximately 96% of the total available program memory. If any additional features were to be added to the program, a different microcontroller with a similar memory structure would be needed (such as the PIC18F458 or similar devices).
A decision was made to program all software using the C programming language, as opposed to assembly language. Because of the amount of trigonometry required to determine the position of the acoustic source, C provided a much simpler and efficient platform for development. In addition, several examples were provided with the CCS compiler, significantly reducing the programming effort.
As discussed in previous sections, the output of the phase detector circuit is a DC waveform whose voltage is proportional to the phase difference between the transmitted and received signals. By monitoring the instantaneous output voltage using the PIC microcontroller, it is possible to determine the position of the acoustic source. The analog DC waveform is captured on the microcontroller and converted to a digital signal for processing. Following the theory presented in sectionXX, the difference between successive analog to digital conversions can be readily used to determine the position of the acoustic source.
The PIC microcontroller was initialized to produce analog-to-digital conversion results in the range 0x00 to 0xFF. Given the characteristics relating the output voltage to the phase difference between transmitted and received signals, 0x00 corresponds to the waveforms being identically in phase while 0xFF corresponds to a phase shift of 360 degrees. For any result in between, say 0xjj, the analog-to-digital conversion result corresponds to a phase shift of (jj / FF)*360 degrees. By evaluating the difference between two successive analog-to-digital conversions, the distance traveled by the acoustic source can be accurately determined.
A simple block flow diagram of the software routines and call structure is given on the following page in Figure 7. Source code for the software implementation can be found in Appendix C, attached at the end of the report.
Figure 7: Software block flow diagram
3.7 Experimental Apparatus
4 Experimental Data
4.1 Operating Distance
4.2 Distance Measurement Stability
4.3 One Dimensional Distance Accuracy
4.4 One Dimensional Distance Accuracy Versus Speed
4.5 Two Dimensional Distance and Angle Accuracy
4.6 Two Dimensional Distance and Angle Accuracy Versus Receiver Separation
5 Recommendations/Future Developments
6 Conclusions
7 References
[2] Kobitone 400ST16 Transducer Datasheet, Mouser Electronics, www.mouser.com