EPA-600/3-77-035
April 1977
Ecological Research Series
INSTRUMENTATION TO MONITOR LOCATION OF
FISH CONTINUOUSLY IN EXPERIMENTAL
CHANNELS
Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Duluth, Minnesota 55804
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ECOLOGICAL RESEARCH series. This series
describes research on the effects of pollution on humans, plant and animal spe-
cies, and materials. Problems are assessed for their long- and short-term influ-
ences. Investigations include formation, transport, and pathway studies to deter-
mine the fate of pollutants and their effects. This work provides the technical basis
for setting standards to minimize undesirable changes in living organisms in the
aquatic, terrestrial, and atmospheric environments.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/3-77-035
April 1977
INSTRUMENTATION TO MONITOR LOCATION
OF FISH CONTINUOUSLY IN EXPERIMENTAL CHANNELS
by
Joseph R. Jahoda
Bayshore Systems Corporation
Springfield, Virginia 22151
Contract No. 68-01-0752
Project Officer
Kenneth E.F. Hokanson
Monti cello Ecological Research Station
Environmental Research Laboratory-Duluth
Monti cello, Minnesota 55362
ENVIRONMENTAL RESEARCH LABORATORY-DULUTH
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
DULUTH, MINNESOTA 55804
-------
DISCLAIMER
This report has been reviewed by the Environmental Research Laboratory-
Duluth, U.S. Environmental Protection Agency, and approved for publication.
Approval does not signify that the contents necessarily reflect the views and
policies of the U.S. Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or recommendation
for use.
-------
FOREWORD
Our nation's freshwaters are vital for all animals and plants, yet our
diverse uses of watei for recreation, food, energy, transportation, and in-
dustry—physically and chemically alter lakes, rivers, and streams. Such
alterations threaten terrestrial organisms, as well as those living in water.
The Environmental Research Laboratory in Duluth, Minnesota develops methods,
conducts laboratory and field studies, and extrapolates research findings
--to determine how physical and chemical pollution affects
aquatic life
--to assess the effects of ecosystems on pollutants
--to predict effects of pollutants on large lakes through
the use of models
--to measure bioaccumulation of pollutants in aquatic organ-
isms that are consumed by other animals, including man.
This report describes development of an ultrasonic tracking system used
to monitor position and temperature history of a mobile fish population in
an experimental stream channel, and discusses design problems and solutions
associated with operations near a nuclear power electrical generating facil-
ity.
Donald I. Mount, Ph.D.
Director
Environmental Research Laboratory
Duluth, Minnesota
iii
-------
ABSTRACT
This study resulted in the development and construction of equipment
to monitor continuously the position and the temperature of up to 20 fish
in a water channel 486 meters long, 3 meters wide, and 1 meter deep. The
system utilized miniature sonic transmitters (tags) operating in the 51 kHz
to 366 kHz frequency range which were implanted in 500 gram or heavier
fish. The battery operated tags were pulse modulated and designed for over
1 year operational life. A temperature sensitive thermistor controlled
the repetition rate of the tag providing the temperature of the fish to an
accuracy of 1 degree C. The nominal range of the polyurethane encapsulated
tage was several hundred feet. Nominal tag size was 16 mm OD x 32 mm long
(4.6 - 5.4 g in water). Sixteen hydrophones were located at 30.5 meter
intervals in the water channel.1 A control console contained a manually-
operated, frequency-stepped receiver which could select any i. dividual
hydrophone, thus locating the fish to within +_ 15.25 meters. Up to 20
individual fish could be monitored. Automatic operation and recording of
the data was considered in the design of the system for future equipment.
Severe radio frequency interference problems were encountered, requiring
extensive precautions and modification of the channel equipment and wiring.
Also investigated were passive fish monitoring and tracking of small
fish fry. An experimental system was completed for limited monitoring
applications.
This report was submitted in fulfillment of Contract Number
68-01-0752, by Bayshore Systems Corporation under the sponsorship of the
Environmental Protection Agency. Work was completed as of October 1974.
IV
-------
CONTENTS
Foreward iil
Abstract iv
List of Figures .Yl
List of Tables v1^
Acknowledgements 1X
Sections
I Introduction 1
Large Fish Monitor System 3
Small Fish Monitor System 6
II Conclusions 9
III Recommendations 10
IV Description of the Large Fish Tracking System 12
V Large Fish Tracking Sub-Systems 17
Ultrasonic Temperature Transmitters (Tags) 17
Hydrophone/Preamplifier 33
Receiver 42
Temperature Decoder 45
Receiver-Temperature Decoder Console 49
Installation 51
VI Problems and Solutions 55
Radio Frequency Interference 55
Electromagnetic Interference 56
Stability of Equipment • 57
Tag Leakage and Aging 57
Acoustic Attenuation Problems 57
References 58
Appendices 62
A Electromagnetic Field Measurements 62
B Investigation of Inductively Coupled Circuits for
Use in Fish Tags 87
C Fry Detection System 105
D Supplementary Materials 116
Glossary of Abbreviations 117
-------
FIGURES
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Mounting
kHz, Vertical Mounting
(Tag) Assembly
Test Channel at Monticello Ecological Research Station
Monticello Ecological Research Station Test Site
Large Fish Tracking System, Hydrophone and Channel Conduit Layout
Ultrasonic Temperature Transmitters
Blocking Oscillator
Modified Colpitts Oscillator
Hartley Oscillator
Acoustic Temperature Transmitter Schematic
Beam Pattern - Tag Frequency: 82 kHz, Horizontal
Beam Pattern - Tag Frequency: 82
Acoustic Temperature Transmitter
Dual Sided Receiving Hydrophone
Hydrophone Preamplifier Schematic
Hydrophone Assembly
Hydrophone Assembly
Receiver Schematic
Temperature Decoder
Signal Descriptions of Temperature Decoder
Receiver Decoder Console
Steel Conduit Runs from Blockhouse
Junction Box
Interconnecting Box
Interconnecting Box Wiring Diagram
Electromagnetic Test Sites - Monticello Ecological Research
Station Channels, Monticello, Minnesota
Electromagnetic Field Measurements - Composite Tests
Measurements
Measurements
Measurements
Measurements
Measurements
Measurements
Measurements
Measurements
Measurements
Measurements
Measurements
Measurements
Ground Conditions at the Monticello Ecological Research Station
Ground Loop Circuits
Induction Coupling into Pool Conductor
Electromagnetic Field
Electromagnetic Field
Electromagnetic Field
Electromagnetic Field
Electromagnetic Field
Electromagnetic Field
Electromagnetic
Electromagnetic
Electromagnetic
Electromagnetic
Electromagnetic
Electromagnetic
Field
Field
Field
Field
Field
Field
- Test
- Test
- Test
- Test
- Test
- Test
- Test
Number
Number
Number
Number
Number
Number
Number
- Composite Tests
- Test Number 1
- Test Number 2
- Test Number 3
- Test Number 1
1
2
3
4
5
6
7
2
2
13
14
18
18
18
21
27
28
31
37
38
40
41
43
46
48
50
52
52
53
54
64
65
66
67
68
69
70
71
72
73
74
75
76
77
80
81
83
VI
-------
FIGURES
No. Page
41 Electric and Magnetic Fields Around a Half Nave Vertical
Antenna 88
42 Variations of Q Against Frequency for Various Litz Wound
Solenoid Inductors 90
43 Variations of Q Against Frequency for Various Litz Wound
Solenoid Inductors 91
44 Variations of Q Against Frequency for Various Litz > r
Solenoid Inductors 92
45 Curves Showing the Relative Received Energies for a vuryin^
Source Energy for Q-l and H Type Core Materials 94
46 Curve Showing the Relative Received Energy for Constant
Source Energy 95
47 Schematic Showing the Configuration of a Power Pickup and Con-
verter for Use as a Power Source for an Ultrasonic Tag 96
48 Externally Triggered, Battery Operated Ultrasonic Tag 97
49 Test Equipment Configuration for the Evaluation of High Current
Inductive Field Sources and Receivers 99
50 High Power Test Data Showing Received Signal as a Function of
Drive Current at a Distance of 6 Feet 101
51 High Power Test Data Showing Received Signal as a Function of
Distance from the Transmitter 102
52 Schematic of Externally Powered Ultrasonic Tag 103
53 Photograph of Ultrasonic Receiver for Reception of 35 kHz
Signals 104
54 Instrumentation Sites for Fry Detection System 106
55 Sample Test Data, Equilibrium Potential System 111
56 Acoustic Fry Detection System Block Diagram 113
vn
-------
TABLES
No. Page
1 Transmitter (Tag) Frequencies (kHz) 24
2 Corrmercially Available Mercury Cells 29
3 Overall Dimensions, Weight and Volume of Each Tag 34
4 Tag Repetition Rate and Pulse Width Characteristics 35
5 Tag Cylinder and Coil Specifications 36
vm
-------
ACKNOWLEDGEMENTS
This program was completed based on a team relationship between the
EPA, Monticello Ecological Research Station as directed by Dr. Kenneth Hokan-
son, and his assistant, Mr. Charles Kleiner, and Bayshore Systems Corporation
(BSC). The many contributors at BSC were Don Anderson, Fred Gagnon, Homer
Guerra, Francis Sherwood, Robert Greenwell, Richard Rockwell and Joseph R.
Jahoda. Their contributions include advances in the state-of-the-art in low
current and high efficiency ultrasonic transmitters, RF interference elimina-
tion, tag potting techniques, step tuned sensitive rf receiver, hydrophone
preamplifier development, temperature A/D decoders, and light emitting diode
displays. Note must be made of the days of working in cold/hot weather side-
by-side with EPA personnel installing the equipment.
IX
-------
SECTION I
INTRODUCTION
To solve some of the more complex ecological problems concerning
the discharge of waste heat into the nation's surface waters, the Environmental
Research Laboratory-Duluth constructed a field station at Monti cello,
Minnesota. Eight (8) identical experimental stream channels, each about
486 meters (1,600 feet) long were constructed at the Monticello test site
(Figure 1). Each channel is composed of alternating riffle and pool
areas. Each channel contains a total of eight (8) riffle areas of
2.4 meters (8 feet) wide and .3 meters (1 foot) deep and eight (8) pools
of 3.7 meters (12 feet) wide and .92 meter (3 feet) deep; each about
30.5 meters (100 feet) long. To date, one channel was instrumented,
channel number 1, and placed into operation. A small laboratory provides
office space, sample processing areas, and houses the control and
monitoring equipment (Figure 2).
Raw Mississippi River water is used for the water supply. This
water is heated to the desired temperature in a heat exchanger using the
heated condenser water of Northern State Power (NSP) nuclear power plant.
The condenser water is returned to the NSP discharge where it is monitor-
ed against state water-quality standards. The rate of flow of water in
channel number 1 was .028 cubic meters per second (1 cubic foot per
second) with mean velocity less than .03 meters per second (0.1 foot
per second).
Four different annual temperature regimes in duplicate will be
programmed and controlled at the upstream end of the channels. Thermal
gradients will develop in these channels except for one pair which will
follow ambient Mississippi River temperatures. Temperatures will vary
*
diurnally and seasonally along thermal gradients. It is intended to
monitor water temperature and other water quality and meteorological
parameters in some detail to accurately describe conditions in each
channel.
-------
Figure 1. Test channel at Monticello
Ecological Research Station,
Figure 2. Monticello Ecological Research Station
test site. Building to the right is the
laboratory, the building in the center
is a garage and storage building.
-------
Natural river plankton and insect fauna are expected to colonize
all of the experimental channels. Once an abundance of food organisms
is present, one fish species will be stocked at natural densities. The
distribution of fish is known to be strongly influenced by the environ-
mental temperature and varies with the different age groups. Young fish
can thrive at higher temperatures than adult fish. Larval fish will
most likely be found near the surface in pools and in areas of low
current velocity. Juvenile and adult fish may have a wider distribution
and may be found on the riffles near the substrate. Recognizing the
problem of monitoring temperature of mobile organisms, this contract was
awarded on 30 June 1972 to develop the instrumentation to monitor the
location of large fish on a continuous basis in the experimental channels.
This instrumentation is necessary to permit determination of the thermal
experience of each fish in the channel so that the effects of temperature
can be properly assessed. In addition, a small fish monitor system,
required to monitor continually the location of small fish fry, was
demonstrated. The large fish monitor system was completed and delivered
by October 1974.
As part of the overall program other items which were delivered
included Instruction and Maintenance Manual and Engineering Drawings
of System Prototype.
LARGE FISH MONITOR SYSTEM
r,
A system was required by the EPA to monitor instantaneous positions
and temperature of fish in the water channels. The system requirements
included a temperature monitoring transmitter (tag) which would be affixed
to fish and measure the temperature of the fish or the surrounding water
to the nearest 1 C. Further, the tag must have a minimum life of 1 year
and be small enough to be placed on, or in, a 25.4 cm (10 inch) fish
without significantly interfering with growth or survival. Each fish
position and temperature should be recorded once an hour. The tags will
require a minimum range of 30 to 150 meters (100 to 500 feet). Either
different frequencies or pulse rates should be used for each fish to permit
-------
distinguishing 10 to 20 individual fish in each channel. The final
system design must be capable of interfacing with a real-time, digital
recording system at a later date.
An additional factor which must be considered in the final design
of the system is the presence of two high voltage overhead transmission .
lines (345 and 230 kilovolts) which cross the channels and will unques-
tionably cause radio frequency interference (RFI) problems (see Appendix A
for details).
During the early phases of the program several system approaches
were considered. They all utilized the basic concept of attaching a small
transmitter to the fish and thereby transmitting the location and internal
temperature to pick-up units on shore. Several variations were possible
and accordingly considered.
Radio Frequency System
A significant amount of research has been done in related military
1 St Y\f\ 9
and wildlife programs for developing rf tags for use in fresh
water. Ranges of over 500 meters have been reported. The use of rf tags
would have required rf receivers, multiple frequencies (to identify
individual fish), elaborate direction finding antenna systems (to locate
each fish), and means to eliminate the effects of rf signals present
from local and distant transmitters and interference from the overhead
power line. These problems would prove insurmountable and the rf system
was dropped from consideration early in the program.
Sonic System
The use of sonic tags has been used for well over 25 years for fish/
manual location, to , temperature monitoring, and 26, and other
O"7 i-. *5 O
fish/mammal/water parameter monitoring, . All of the efforts to
date had concentrated on manual tracking over relatively short periods
of time, usually 2 to 4 weeks. The ranges usually involved 100 to
500 meters (110 to 550 yards).
-------
Early in the program it became evident that to meet the one year
operating life requirement of the tag in a compact unit, a highly efficient
tag would have to be developed. Since long range was not required in the
experimental channels, a system was designed which would decrease the
range requirement of the tags to 40 meters (132 feet) and thereby
considerably reduce size and increase battery life .
The minimum fish size of 25.4 cm (10 inches) dictated the use of
internally mounted tags with lengths of less than 3.2 cm (1.26 inches)
and outside diameters of less than 1.6 cm (.63 inches). This size
9 35
limitation was arrived at by researching the literature, , discussions
with fish biologists, and experimentation at the Monticello EPA laboratory.
All of these requirements imposed severe restrictions on the
final tag design. Factors considered were:
(a) Long term (over one year) waterproof sealing.
(b) Ultra-miniature packaging.
(c) Circuitry stabilization to minimize frequency drift
as a function of short and long terms and temperature.
(d) Non-toxic outer coating.
(e) Long life batteries.
(f) Minimum weight (in water).
Passive System
The use of passive systems has found very little application in fish
2
tracking systems . The approach basically utilizes sonic or magnetic
energy beamed to a remote location (fish tag) from a shore station. This
energy is converted to dc and used to power a transmitter which then
responds with an appropriate signal. The limiting factor is the range
between the tag and shore location. After considerable effort, a maximum
range of only 3 meters (9.9 feet) was accomplished. By utilizing larger
power levels at the shore based transmitters this range could be increased.
However, after considering all factors, this system was deemed economically
unfeasible (see Appendix B for details).
-------
SMALL FISH MONITOR SYSTEM
Initially, the EPA required the counting and/or location within
the water channels of freeswimming offspring from natural reproduction
in relation to temprature recording stations. A remote sensing device
was required to locate and/or count (i.e. + 10%) larval and juvenile
fish (i.e. 4 to 40 mm total length) at fixed stations. Fish count or
position recordings were to be made once an hour at each station.
Relative abundance indices at each station were required, rather than
a total fish count in each channel. Such a means of detection would not
require handling or tagging individuals. A design goal was to discriminate
different size fish so that large and small fish could be counted as
individuals regardless of their direction of movement. The system was
to have the design capability of interfacing with a real time, digital
recording system at a later date.
There is very little precedent for a system of this type which
concerns itself with a female fish spawning 15,000 to 40,000 eggs in
the channel bottom. Initially, the survival rate is 60% to 70% and one
should be prepared to monitor several thousand 4 to 12 mm (.16 to .47 inches)
fish fry. In the early stages the fry may school together or assume a
uniform distribution. As they approach 10 mm (0.4 inches) they start to
school, covering an area of approximately one cubic meter (1.31 cubic
yards).
The following are typical of the methods proposed and/or actually
utilized by researchers, fisheries, universities, commercial fisherman
and industrial, NASA, military and non-profit agencies for tracking
schools of fish:
1. Passive - acoustic
a. Swimming noise.
b. Vocal sounds.
c. Strumming, beating or scratching noises.
d. Noise produced by contact with external objects.
-------
2. Passive - general
a. Water-color spectrometer - color/luminescence.
b. Luminescing fish detection.
c. Fish oil slick detection.
d. Infra--red detection of fish swimming near the water
surfaces.
e. Visual-by air and shipborne observers.
f. Photographic - IR and color.
g. RF signal emmissions from fish.
3. Active
a. Sonar - transmission and reception of acoustic signals.
b. Laser - transmission and reception of coherent light.
c. Microwave/millimeter radiometers.
After considering all of these techniques, those which offered
promise for locating fish fry in the narrow channels were sonar, laser,
and infra-red.
Sonar
This is probably the most often taken approach. Sonar systems for
fish location have been built over the frequency range of 10 kHz to
600 kHz. They have been used to: count salmon going upstream, measur-
ing biomass (kilograms) of fish in rivers, lakes, and portions of the
ocean, etc. with some success. However, the transmission of an acoustic
signal in a small narrow channel offers additional problems. These are
due to such factors as multiple reverberations, pickup: from the main
and side lobes of the transducers, unpredictable variation in reflection
from the fish as a function of its position, reflection from the water
surface and irregular contours, the unpredictable variation in signal
strength returned from fish of different sizes, and doppler shifts due
to rapidly moving fish. Many techniques have been proposed or attempted
to obviate these disadvantages. Such factors as swim bladder, vertebral
column and body tissue resonances can be used to isolate fish
-------
acoustically, ° , especially small fish, from others of different
sizes. To properly utilize resonance effects would require transducers
operating at multi-frequencies. The problem of reflections, echos,
false responses, etc., can be solved by time gating, FM techniques
and multiple transducers. One could use specially built sonar arrays,
funnels, baffles, etc., to assist the sonar system, however, the small
size of the fish and the desire to avoid unusual mechanical intrusions
into the water further complicate the use of sonar techniques. The
final system chosen, after careful consideration of all factors and
extensive experimentation, utilized a modified multi-frequency sonar
technique, see Appendix C for system details.
Laser
This would prove expensive due to the complexity of the ancillary
equipment required. In addition, the data processing would prove too
complex and expensive and it would suffer from shading effects from
fishes nearest the laser. The same comments would apply to the use of
similar techniques based on light source sonar.
Infra-Red
Limited to fish at surfaces of the water. Shading effects and
difficulty of implementation prevented serious consideration.
8
-------
SECTION II
CONCLUSIONS
The final system, which accomplished this objective was composed of
16 hydrophone/preamplifiers, sonic fish tags, shielding cabling and conduits,
and a receiver/decoder console. The 16 hydrophone/preamplifiers were
located at 30.5 meter (100-foot) intervals along the channel. The operator
can select the operating frequency of the desired tag (unique frequency in
the 51 kHz to 366 kHz range for each fish) and the appropriate hydrophone
which has the strongest response. Knowing which hydrophone received the
strongest signal would locate the fish to within an accuracy of +_ 15.25 meters
(50 feet). The decoder converted the repetition rate to a numeric display
of the fish temperature.
The extrapolated life of the tag meets the one (1) year objective.
With the advent of the new lithium battery, the life can be further
increased by almost a factor of two.
The seriousness of the radio frequency interference (RFI) from the
overhead high voltage power lines cannot be overstated. Signals were
present at unpredictable times on frequencies very close to those being
used. At times, the frequencies in use had to be changed. Based on the
RFI difficulties experienced with one instrumented channel, additional
shielding, filtering and isolation will be required for the instrumentation
to be installed in the future for the remaining 7 channels. This is mandatory
because the RFI problem can be expected to be cumulative in its impact,
Monitoring of the small fish fry was a most difficult task. Their
small size and their physical properties, being very similar to that of
water, precluded the use of conventional sonar, laser, and infra-red
techniques to monitor their presence. Drifting algae also presented
background interference problems during the initial system development.
After considerable experimentation a modified sonar technique (utilizing
critical frequencies and the transmission of sonic energy through a mass
of material having characteristics different from that of the water) was
developed which was capable of monitoring small fish fry under conditions
of limited floating algae and in areas of minimum turbulence.
-------
SECTION III
RECOMMENDATIONS
Based on the data and results of this program, recommendations
for future effort and system expansion are;
(a) The availability of advanced shielding material, improved
coupling and balancing circuits, and a more selective phase-lock receiver
will greatly improve system operation. The RF interference problem, which
inherently is unpredictable, will necessitate these modifications. In
particular, the use of a phase-lock receiver should be stressed, since it
will permit automatic frequency lock-on to each tag. Thus, should the tag
frequency drift or the fish be successful in hiding behind growth for
a short time, the equipment would not require manual retuning and loss
of data would not result. Specifically, the following should be implemented
during the next program phase:
(1) Hydrophone Preamplifiers - Modify output circuitry
to provide improved sensitivity and decrease ground
loop and RF interference from power station.
(2) Improved shielding of console and receiver.
(3) Automatic phase lock receiver modification.
(4) Improved tags, through the use of temperature
compensated components, to minimize frequency drift.
(5) As additional instrumentation is added for the
remaining 7 channels, extreme care must be exercised
in proper utilization of interference filters, isolation
and shielding to avoid mutual interference problems.
(b) The present receiver-decoder system control panel is
manually operated. This greatly limits the data available to the normal
working hours of available personnel and increases the system operating
costs. The present design was predicated upon ease of conversion to
automatic operation with a computer and recorder. This should be imple-
10
-------
mented after the initial phases of system check-out are completed and
the biological experiment is satisfactorily structured.
(c) Lithium batteries should be used when they become
available. They will permit as much as a 10% decrease in length and
weight for the same life expectancy or a longer life expectancy for the
same size.
(d) It is recommended that a small fish fry detector be
implemented in biologically critical areas of the channel. An efficient
and simplified system can be completed and calibrated especially with
the availability of fish fry on site. They were not available during
much of the original experimentation. As a minimum such a system
would provide position location and approximate volume of the fish fry
school.
(e) Initiate a small program to study the impact of the
projected automated system on the computer software requirements. A
computer with sufficient capacity will be required and a study made of
the specific details of modifying the fish location system. Specifically,
it is important that the computer have sufficient capacity to handle the
data from eight channels. Typical of the critical factors which need to
be considered is the switching rate of the receiver and hydrophones. This
will be limited by reverberation rate of the channels, time required to
locate each fish, etc.. This effort will interface with the biological
work in progress at the Monticello test site and paragraph (a) above.
11
-------
SECTION IV
DESCRIPTION OF THE LARGE FISH TRACKING SYSTEM
The large fish tracking system installed at the Monticello Field
Station consists of a number of component equipments arranged as shown in
Figure 3. The system is comprised of three main subsystems/components;
a number of ultrasonic temperature transmitters (tags, inserted within
the fish), an array of hydrophones, and a receiver-decoder control console.
Ultrasonic temperature transmitters (or fish tags), transmit a
pulsed, acoustical signal through the water (Figure 4). The repetition
rate at which the tags transmit is determined by a temperature sensor
within the tag. At cold temperatures the repetition rate is slow and
for warm temperatures the repetition rate is faster. This form of trans-
mission is called pulse interval modulation (PIM). This type of modulation
permits location and temperature monitoring with minimum hardware. Complex
telemetry modulation techniques would have required more electronics in
the tags as well as in the receiver. As an additional advantage, the system
is much less susceptible to reverberation effects than multi-carrier,
phase or constant wave modulation techniques. The tags are designed to
be pulsed on for a relatively short interval of time in order to obtain
a low duty cycle which minimizes the drain on the internal batteries and
increases the effective life of the tag.
The hydrophones are used to pick up pulsed acoustical-signals
transmitted by the tags and convert them to electrical signals that can
be detected by a receiver. A total of 16 hydrophones, located at 30.5 m
(100 ft) intervals are used. By observing which of the hydrophones receives
the strongest signal, one can determine the location of the fish within
^ 15.75 m (50 ft). Due to the wide range of frequencies covered by the
tags, 51 to 366 kHz, a special wide-band hydrophone was designed with
special filters and preamplifiers. Due to the extremely high radio
frequency interference level, rf shielding was used in the hydrophones/
preamplifiers and at all junctions and conduits. Each preamplifier was
12
-------
4" CONDUIT
" CONDUIT
N3
BLOCKHOUSE
O HYDROPHONE
D JUNCTION BOX
RECEIVER/ DECODER CONSOLE
Figure 3. Large fish tracking system,
hydrophone and channel conduit layout.
-------
Figure 4. Ultrasonic temperature transmitters.
protected against high rf levels and high induced voltage from
electro-fishing instruments and possible lighting discharges by the
use of diodes at their inputs.
The receiver selects the one desired frequency from among the
20 present frequencies at the output of the hydrophone/preamplifier and
converts the pulsed electrical ultrasonic signals to a low frequency
pulsed signal to the decoder. The decoder then measures the time between
pulses and gives a corresponding readout of temperature on a digital dis-
play meter.
Technical specifications of each of the major subsystems are as
follows: The value of the parameters are average or nominal values.
All tags delivered were within 10% to 15% of these values. The dimensions
varied more widely since at the higher frequencies smaller ceramic elements
were used. Specific specifications are discussed in greater detail in
Section V.
14
-------
(1) Ultrasonic Transmitters
Frequency:
Pulse Width:
Power Output:
Repetition Rate:
Duty Cycle:
Expected Life:
Size:
Weight:
Frequency Stability:
SI to 366 kHz (18 Tags)
2.5 milliseconds
+40 dB re 1 JLIbar/meter
0.8 pps at 0 C
1.6 pps at 30 C
.2% at 0 C
.4% at 30 C
450 days at 0 C
275 days at 30 C
(OD) 15 to 17 mm
(L) 30 to 39 mm
(Air) 10.7 to 12.6 grams
(Water) 4.6 to 5.4 grams
0. 15% (0 C to 30 C)
(2) Hydrophone/Preampl
Frequency:
Gain:
Output Impedance:
Input Power:
Shielding:
(RF)
(Electrostatic)
Beamwidth:
(Each Section)
Sensitivity:
(3) Receiver (Located
Frequency:
Sensitivity:
IF Bandwidth (3dB)
Input Impedance:
ifier (16 per channel)
51 to 366 kHz
20 dB (Minimum)
10 ohms (Maximum)
12 V DC
Special Housing/Cable
Coating for Sonic Window
180° Horizontal
40° Vertical
-88 dB re 1 nbar/meter
in Receiver/Decoder Console)
51 to 366 kHz (20 steps)
0.1 microvolts (S/S+N = 1)
: 1.2 kHz or 4.0 kHz
400 ohms
15
-------
(4) Receiver/Decoder Control Console
Temperature Readout: 0 to 30 C
VCO Readout:
Controls:
Jacks:
Meter:
Voltage to VCO (Frequency
analog voltage)
Audio Gain
RF Gain
Tag Selector (20 Frequencies)
Freq. Adj. Fine Tuning
(20 Frequencies)
Temp. Calibration (20 tags)
Hydrophone Selector (16 stations)
Oscilloscope Output
Audio Output
AGC Level
Signal Strength
16
-------
SECTION V
LARGE FISH TRACKING SUB-SYSTEMS
ULTRASONIC TEMPERATURE TRANSMITTERS (TAGS)
Oscillator Circuits
During the design phase of the contract three oscillator circuits
were selected to be evaluated: squegging, modified Colpitts and Hartley.
These oscillators were designed to serve dual purposes: frequency generator
and transmitter, in order to reduce the number of components used in the
transmitter circuit. The size of the tag was'of primary concern, there-
fore emphasis was placed on the use of minimum parts along with the
selection of miniature sized components. Since the electronics would
only occupy less than 25% of the final tag (the battery and ceramic
element are the largest size contributor and cannot readily be miniaturized),
the possible size and weight saving on the final tag of using micro-
circuitry do not justify their high cost and accordingly were not considered.
The squegging or blocking oscillator was found to be the simplest
circuit as far as the number of components used (Figure 5). However,
the circuit was very unstable with changes in battery voltage and resulted
in a 25% change in repetition rate over the life of the battery. A small
tag was assembled, potted and evaluated. The size of the tag was 10 mm
in diameter and 20 mm in length. Test results showed an interaction
between the repetition rate and pulse width with load changes in the
oscillator circuit. As the tag was lowered into the water the repetition
rate would increase and the pulse width would decrease. This circuit,
although simple and small, could not be used for accurately determining
temperature of the water due to large changes in the repetition rate with
battery voltage decay and water loading.
The modified Colpitts oscillator (Figure 6) was breadboarded and
evaluated. The circuit had good frequency stability with changes in
battery voltage (1.4 to 1.0 volts), less than 0.15% frequency drift
17
-------
-A/W
Figure 5. Blocking oscillator.
V
V
T
Figure 6. Modified colpitts oscillator.
Figure 7. Hartley oscillator.
18
-------
resulted. The tuning of the circuit was critical because the ratio of the
two feedback capacitors had to be changed with different frequencies.
The circuit required four capacitors to tune to the operating frequency
and two of the four were large value capacitors which required larger
sized components. The power output of the Colpitts oscillator was
limited due to the required emitter resistor in the oscillator circuit.
The Hartley oscillator (figure 7) was breadboarded and evaluated.
The frequency stability of the circuit was less than 0.12% with the
change in battery voltage. The frequency of the oscillator could be
changed by changing the value of a single capacitor. The voltage required
across the cylinder could be varied by changing the turns ratio of the
transformer. The power output of the oscillator could be adjusted by
inserting a low value resistor in series with the emitter of the oscillator
transistor. Because of the simple circuit with few components the
Hartley oscillator was chosen to be used as the transmitter in the tags.
Timer Circuits
Due to the limited battery power the tag should operate at the
minimum duty cycle consistent with satisfactory reception at the receiver.
The duty cycle is merely the ratio of transmitter "ON" (drawing battery
power) to transmitter "OFF" (drawing no battery power).. To obtain low
duty cycles the transmitter must be turned ON and OFF at certain periods
of time. To do this a timer circuit is used that required little
current drain from the battery. Two timer circuits were evaluated;
astable multivibrator and a complementary astable multivibrator.
The astable multivibrator is a two transistor oscillator that
periodically produces a square or rectangular pulse at its output. The
frequency at which this pulse is produced can be controlled by changing
the value of a resistor in one of the transistor stages. By using
a thermistor, which is a device that has a negative temperature coefficient,
in place of the resistor, large changes in resistance occur wtth changes
in temperature. This provided a means of turning the transmitter ON
19
-------
and OFF at different time intervals that corresponded with the temperature
of the thermistor. The astable circuit proved to be stable with voltage
changes with less than a 5% change in the repetition rate over the life
of a single battery. The use of two batteries reduced the change in
repetition rate to 2.5%. The one disadvantage of the astable multivibrator
was that at all times one of the two stages was in an ON condition,
thereby continually drawing current from the battery. The average current
drawn was reduced to less than 20 micro-amps but this was considerable
as the total average current for both the transmitter and timer had to be
kept below 25 micro-amps to obtain a one year life. This is based on an
available battery capacity of .18 amperes hours. A year has almost
9000 hours, thus at 20 microamperes (or 20 x 10 amperes) consumption
per hour, the battery would require a capacity of 9000 x 20 x 10" or
0.18 ampere hours.
The complementary astable multivibrator uses a complementary pair
of transistors (Ql and Q2, Figure 8) so that both transistors are either
ON or OFF at the same time. The only time current is drawn from the
battery is during the pulse on time which is in the order of a few
milliseconds. With the use of high beta transistors and high value
resistors the average current can be kept below two micro-amps for the
timer circuit. This circuit was selected for the timer because of its
good stability (less than 5% variation in repetition rate over the
voltage variation which results over the life of the battery), low
power consumption and the simplicity of the circuit. It should be noted
that an integrated circuit (1C) commercially available version of the
complementary astable was also evaluated and found to have excellent
stability, less than 1% change in repetition rate with 20% changes in
battery voltage. The 1C required a minimum of 3.0 volts for its
operation and its overall size was larger than that of the discrete
components. Because of the additional batteries required and the larger
size, the 1C was not selected as the timer in the final design.
20
-------
Battery
ro
R - Resistors
Q - Transistors
Y - Ceramic PZT Cylinder
C - Capacitor
T - Transformer
RT - Thermistor
- Battery
Figure 8. Acoustic temperature transmitter schematic.
-------
Transmitting and Receiving Elements
The transmission and reception of underwater sonic signals is made
possible by piezoelectric devices. They have the property of vibrating
mechanically when a voltage is applied to them. Piezoelectric cylinders
and discs were evaluated to be used as acoustic transmitters and receivers
for the tags and the land-mounted hydrophones. The requirements of the
receiving hydrophone were that it be directional to reduce noise pickup
and be highly sensitive so as to reduce the power output requirements
of the acoustic transmitters. As an example, since the water is shallow,
there is no need to have the hydrophone receive signals from the bottom
and top of the channel. Rather it should be most sensitive up and down
stream. Thus extraneous signals and noises coming from the top of the
channel would not be received strongly and would not interfere with the
desired (tag) signals. Lead zirconate titanate (PZT) discs were
selected as the receiving elements because of their directional capabilities
and their high sensitivities.
Three PZT cylinders of different dimensions were selected as the
acoustic transmitters. Initially, one size of cylinder was used. However,
the sensitivity of the unit was poor at the low and high frequencies.
This is due to a quality of resonance which all cylinders have. The
cylinder vibrates best at its resonance and this is primarily due to its
size and shape. To cover the large frequency range several units with
different resonant frequencies were required. The size of the cylinders
and their resonant frequencies are as stated below. Note that the high-
est frequency is 185 kHz. Since this unit operated satisfactorily up to
365 kHz, higher resonant frequency cylinders were not necessary. This
is fortunate since higher frequency units are not readily available
and are fragile.
(1) 12.9 mm OD x 6.4 ram L and 11.4 ram ID
Frequency-Circumference Mode 80 kHz.
(2) 9.6 mm 00 x 6.4 mm L and 8.0 ram ID
Frequency-Circumference Mode 125 kHz.
22
-------
(3) 6.4 mm OD x 6.4 mm L and 4.8 mm ID
Frequency-Circumference Mode 185 kHz.
Thermistor Element
In order to monitor the temperature of the tag environment a
small temperature sensitive element (thermistor) was used in the tag.
Its resistance varies with the temperature. Thus, by using the thermistor
element in a circuit whose repetition rate varies with the variation in the
thermistors, an , its temperature can be monitored at a remote
location by measuring the repetition rate of the received signal.
The thermistor has a resistance of one megohm with a tolerance of
+_ 20%. With a +_ 20% tolerance the resistance variation is 800K ohms
to 1.2 megohms. Such a wide range will result in wide variation from
tag to tag. To compensate for the different repetition rates to be
expected, a gain and slope adjustment was required in the decoder unit which
converts repetition rate to temperature.
Final Design Circuit
The schematic of the final tag circuit is shown in Figure 8. The
timer circuit consists of transistors Ql and Q2 and their associated
components. Transistors Q3, Q4 and Q5 act as switches and buffer amplifiers.
The oscillator and transmitter circuit is transistor Q6 and its components.
The PZT cylinder is represented by the symbol Yl and the thermistor by the
symbol RT1. The frequency determining components are the PZT cylinder Yl,
capacitor C3 and the secondary inductance of Tl. Thus for each tag, these
3 parameters must be carefully evaluated to assure that the correct
final operating frequency results. The thermistor RT1 along with
capacitor Cl determine the repetition rate of the transmitter, with the
thermistor being the variable with temperature changes. The operating
frequencies of the transmitters were selected so as to maintain a minimum
of 6.5 kHz i 2% of the frequency between each transmitter (Table 1).
23
-------
Table 1. TRANSMITTER (TAG) FREQUENCIES
(kHz)
Tag
number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
20
Fundamental
frequency
50.0
59.35
69.2
77.2
88.2
102.0
113.6
123.9
133.0
143.0
160.25
170.2
184.1
194.4
218.4
239.7
365.0
345.0
2nd
Harmon.ic
100.0
118.7
138.4
154.4
176.4'
204.0
227.2
247.8
266.0
286.0
320.5
340.4
368.2
3rd
Harmonic
150.0
178.05
207.6
231.6
264.6
306.0
340.8
4th
Harmonic
200.0
237.4
276.8
308.8
352.8
24
-------
The pulsing ON and OFF of the transmitter creates harmonically
related spurious signals that are of sufficient power levels to interfere
with certain selected operating frequencies. Therefore the second,
third and forth harmonics of each transmitting frequency must be considered
when selecting the operating frequencies, in order that they are different
from the fundamental frequency of all of the tags. This is particularly
important in selecting the lower frequencies. These will have harmonics
which could fall in the range of higher frequency tags. See Table 1,
and note the final selection of frequencies and how the harmonics are
different from all of the fundamental frequencies.
Another factor in determining the operating frequency is the
cylinder. Each cylinder has three modes of operation: Circumference,
length and thickness; each mode resonates at a different frequency.
At these mode frequencies large changes in cylinder impedance occurs,
resulting in critical tuning at, or near these frequencies. These
frequencies must also be omitted in the final section of operating
frequencies.
Another consideration for the selecton of the tag frequencies, and
one which was not considered until the problem surfaced during
final systems testing at Monticello, was interference from signals being
transmitted through the overhead high voltage wires. These signals
are unusually strong and are picked up through the water and into the
hydrophones, at the junction boxes and cables, and at the blockhouse
in the receiver.
The frequency stability of the oscillator at a constant temperature
is determined by the aging process of the ceramic material used in the
manufacture of the cylinder. After the manufacture of the cylinder a
continuing change in the dielectric constant occurs for a period of
several weeks. This dielectric change results in a change in the
capacitance of the cylinder which results in a change in the oscillator
frequency. The aging rate decreases with time and can be accelerated by
placing the ceramic in a heated oven for a period of time.
25
-------
The frequency drift of the oscillator with changes in temperature
is caused by the temperature coefficients of the cylinder, the transformer
and the tuning capacitor. This change in frequency over temperature
was compensated in the final tags to a marked degree with temperature
compensating capacitors. They are larger in size than the miniature
tuning capacitors used in the present tags but do not affect the final
size significantly.
The acoustic power output of the tags average 1 milliwatt with
an average of 7 milliwatts input power for an efficiency of approximately
14%. The range of most tags exceeded 46 meters (150 feet) in clear
stable water, with the design goal set for a minimum of 15.25 meters
(50 feet).
The temperature sensing thermistor was placed inside the tag to
reduce external leads. External leads were kept to a minimum to prevent
any weak points in the tag that might be susceptible to water leakage,
Directional patterns of an,82 kHz tag are shown in Fiaures 9
and 10. The horizontal pattern, Figure 9, clearly indicates that
maximum power is radiated from the sides of the cylinder, as expected.
The vertical pattern, Figure 10, is symmetrical and virtually
omnidirectional. All of the tags will exhibit similar patterns.
Batteries
The mercury and the silver oxide batteries are at this time the
only available miniature high capacity batteries on the market, with
a relatively constant voltage over the life of the battery (Table 2).
The silver oxide battery MS76 has an initial voltage of 1.5 v and a maximum
capacity of 160 milliampere hour for the required size used in the present
tags, 11.4 mm OD, 5.2 mm long, and a weight of 2.8 grams. The mercury
battery MP675 was selected for the tags because of its higher capacity
of 220 milliampere hour. Its size is 11.4 mm OD, 5.1 mm long, and it
weighs 2.8 grams.
26
-------
210°
ISO0
Figure 9. Beam pattern-tag frequency: 82 kHz,
horizontal mounting.
27
-------
Figure 10. Beam pattern-tag frequency: 83 kHz,
vertical mounting.
28
-------
Table 2. COMMERCIALLY AVAILABLE MERCURY CELLS
ro
vo
Type3
RM 212
RM 312
E 312
Hg 312
W-l
Mai lory
RM136H
MS13H
RM 400
E 400
Hg 400
Chemistry
Mercury
Mercury
Mercury
Mercury
Silver Oxide
Mercury
Voltage,
volts
1.4
1.4
1.4
1.4
1.5
1.4
Service
capacity,
ma -hour
16
36
36
60
60
80
Weight,
grams
0.5
0.64
0.64
0.98
0.98
1.15
Diameter/
height, cm
0.33/0.55
0.36/0.79
0.36/0.79
0.535/0.79
0.535/0.79
0.345/1.16
^Letter prefixes in "type" show mfgr. RM, Ms Mai lory;
e, Everready; Hg Burgess
-------
The battery is the major limiting factor in the life of the tags.
The capacity of the batteries vary depending on when they were manufact-
ured and where they have been stored since their manufacturing date.
Batteries of the same type but made by different manufacturers
vary in their capacity. The batteries used in the present tags were
made by Mallory because of their higher capacity rating 220 ma/hr as
compared to Everready with 210 ma/hr and Burgess with a rating of
180 ma/hr. Batteries purchased from Mallory were ordered directly from
the manufacturer and were placed in plastic bags and put into a refrigera-
tor when received. The temperature of the refrigerator was maintained
at approximately 5 C.
A higher capacity lithium battery has been developed and is being
manufactured in the larger battery sizes. The lithium battery has
approximately twice the capacity of an equivalent volume mercury battery.
There are several manufacturers of the lithium battery, however, at
this time there are no miniature batteries of this type being sold.
They should be considered for future tags, however.
Tag Assembly
The tag as shown in Figure 11 is assembled in four steps. First,
the printed circuit board containing the timer is assembled and tested
for operation. The oscillator circuit is next assembled on the back side
of the printed circuit board.
The second step of assembly is the glueing (Eastman 910 Glue) of
the transformer pot core with the cylinder. The pot core is an iron core
upon which the transformer is wound, providing an efficient and smaller
transformer than if an air core were used, witn the use of the 125 m
(0.5 inch) OD cylinders the pot core is inserted into the cylinder to
reduce the length. With the smaller high frequency cylinders the
pot core is attached to the end of the cylinders.
The third step glues the cylinder and pot core to the printed
circuit board. At this point the tag can be tested for operation and
30
-------
BATTERY
Wire connected to
activate power
3
.XL.
'A- StAIML£:SS STEEL
SCREW
Figure 11. Acoustic temperature transmitter (tag) assembly.
-------
can be tuned to its operating frequency. This involves changing the value
of capacitor C3 until the desired frequency is achieved. The last step
consists of attaching the battery, applying epoxy to all components except
the cylinder and the potting of the completed tag.
Potting Compounds and Molds
Potting compounds from several manufacturers (Dupont and Products
Research and Chemical Corporation) were evaluated. The compounds evaluated
were epoxy resins, silicone rubber, and urethane rubber. Parameters used
in the evaluation were: Workability, presence of air bubbles, odor,
hardness, fish toxicity, and moisture absorption.
The two-part epoxy resins tested had a durometer hardness of 35 to
80 on the Shore D scale. The moisture absorption rate of the resins was
dependant to an extent on the hardness of the cured epoxy. The harder
the epoxy the lower the absorption rate. The moisture absorption rate
varied from .1% to 1.2% after ten days with a relative humidity of 96%
as quoted by the manufacturers.
The silicone rubber tested had a durometer hardness of 20 to 35 on
the Shore A scale and was very susceptible to water absorption. This
single part room temperature curing, Silicone Rubber Sealant (RTV) com-
pound was available in tubes and when squeezed from the tube was in a
thin paste form. Two or more coatings of RTV were required to prevent
water leakage into the test circuits.
The urethane rubber, a two-part compound, required an evacuation
process to remove trapped air bubbles from the compound. With a duro-
meter hardness of between 50 to 80 on the Shore A scale, the urethane
rubber proved to be satisfactory for the potting of the tags. The
semi-flexible urethane adhered well to the cylinders and presented a
good acoustical match between the cylinder and the water.
The manufacturer of the urethane selected was Products Research
and Chemical Corporation (PRC). The PRC number of the first urethane
used was PR 1524. This compound was used for over one year before it
32
-------
was known that it contained a cancer suspect agent, 4, 4'-methylene-bis-
(2-chloroaniline) or "Moca" as named by Dupont. At that time PRC came
out with a "Moca" free compound PR 1564. This new compound was used
in the potting of the tags and the hydrophones delivered to EPA.
Plastic molds were used initially for the potting of the tags.
With the use of the higher viscosity compound PR 1524, trapped air
bubbles remained at the bottom of the mold aften evacuation. This
resulted in weak spots in the cured compound which made the tags
susceptible to water leakage. Attempts were made to dip the tags in
the potting compounds, but resulted in an uneven coating around the tags
which after a short period of time in water resulted in water leakage
into the tags.
With the use of the lower viscosity compound PR 1564 and the use
of the teflon molds the majority of the trapped air could be removed by
evacuation for about 30 minutes. This resulted in a uniform coating
around the tag with less chance of water leakage.
Several types of molds were evaluated. The final mold used was a
"split" mold made from 2 sheets of Teflon (each 2.54 cm x 12 cm x 8 cm)
held together in a vise to form a 5.08 cm x 12 cm x 8 cm sandwich.
A 1.5 cm diameter hole was drilled at the joint between the 2 sheets.
The depth of the holes was approximately 6 cm. After the potting was
completed the tags were easily removed by separating the teflon sheets.
Tables 3, 4, and 5 summarize the individual tag overall dimensions,
weight, volume, repetition rate, pulse width, cylinder dimensions, and
coil winding information. Section IV also provides technical specifications
for the tags.
HYDROPHONE/PREAMPLIFIER
Piezoelectric Elements
Three different sized PZT disc were selected for use in the receiving
hydrophone to construct a wide band hydrophone with good sensitivity. The
radial and thickness resonances are used to obtain overlapping frequency
33
-------
Table 3. OVERALL DIMENSIONS, WEIGHT AND
VOLUME OF EACH TAG
Tag
number
2
4
5
6
6*
7
8
10
12
13
14
15
16
Length,
mm
32.4
36.4
34.6
33.3
32.5
33.7
30.4
30.8
32.4
31.8
38.9
37.0
38.8
Diameter}
mm
15.5
17.4
15.7
15.5
16.1
15.5
16.1
16.2
15.4
15.3
16.2
16.1
15.8
Weight
in water,
g
4.77
4.72
4.64
5.12
5.17
5.47
4.98
4.69
4.84
4.91
4.77
4.60
4.79
Weight
in air9
g
11.08
12.58
11.50
11.71
11.68
11.98
11.05
10.67
11.18
11.17
12.44
12.01
12.62
Volume
mg/1
6.3
7.8
6.8
6.7
6.4
6.4
6.1
5.9
6.3
6.2
7.7
7.4
7.9
*No longer in use.
34
-------
Table 4. TAG REPETITION RATE AND PULSE WIDTH CHARACTERISTICS
Tag number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
20
Repetition rate,
PPS
.75
.74
.74
.87
.84
.75
.77
.78
.80
.80
.78 •
.76
.84
.78
.84
.84
.82
.74
Pulse width,
msec
2.5
2.6
2.8
2.6
2.6
2.6
2.8
2.8
3.0
2.8
3.0
2.6
3.5
2.4
3.0
2.6
3.4
2.0
Frequency
air
49
58.5
68.4
78.5
89.2
103
108
123
133
145
161
171
184
195
219
242
365
340
water"
51
62
69
80
91
96
no
122
134
145
162
173
183
198
221
236
366
341
35
-------
Table 5. TAG CYLINDER AND COIL SPECIFICATIONS
Cylinder
Tag diameter x length Coil (number of turns)
number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
20
1.3 x
1.3 x
.97 x
.97 x
.97 x
1.3 x
1.3 x
1.3 x
1.3 x
1.3 x
1.3 x
1.3 x
1.3 x
.97 x
.97 x
.64 x
.64 x
.64 x
cm
.64
.64
.64
.64
.64
.64
.64
.64
.64
.64
.64
.64
.64
.64
.64
.64
.64
.64
primary
15
15
15
15*
15
13
13
13
13
11
11
11
10
8
8
8
8
8
secondary
90
81
80
64
90
80
71
64
62
60
50
45
40
45
40
50
45
30
tap
20
20
20
18
18
18
18
18
18
15
15
15
15
15
15
15
15
15
Nominal
frequency in water,
kHz
51
62
69
80
91
96
110
122
134
145
162
173
183
198
221
236
366
341
36
-------
ranges to cover the frequency range of 51 to 366 kHz.
The single low frequency disc 31.75 mm OD x 15.4 mm thick has
a radial resonance of 70 kHz and a thickness resonance of 150 kHz.
The single medium frequency disc 19.1 mm OD x 6.35 mm thick has a
radial resonance of 125 kHz and a thickness resonance of 300 kHz.
The three high frequency discs (10.7 mm OD x 4.7 mm thick) have a
radial resonance of 220 kHz and a thickness resonance of 400 kHz.
Preamplifier
The dual sided receiving hydrophone (Figure 12), consists of three
sets of PZT discs on each side with each set of discs having its own
bandpass filter and buffer amplifier to isolate the different sets of
discs from each other to prevent interaction between the discs.
The acoustic signals received by the discs are filtered by the bandpass
filters (LI L2/C1C7 and L3L4/C8C14 etc.) and are sent to the buffer
amplifiers (Ql, Q2, Q3...Q6) (Figure 13) The sinnals from each of
the buffer amplifiers are added together and amplified by the low output
impedance power amplifier. The output of the power amplifier (Q12 and
Figure 12. Dual sided receiving hydrophone.
-------
CO
CO
HYDROPHONE
FROKT SIDE
LI4
LI
CERAMIC
1.25" QD. __
xO,50"THK T]' C7-T
^-^r^-j-^m^
• •• "-^ cisr:
CERAMIC
0.75"OD.
1 1
1 1
L3
CAPs800pf
CERAMIC
OA Z" 0,0.
C"
r
i
L5
L6
nnrrir^
CERAMIC
1.25" QD.
i
C"
r
L7
.r^
cie
r
t
CERAMIC
0.75"O.D. ..
x0.25"THK
CAp-.800pf
CERAMIC
0.4 2" 0.0.
11
,, _ U9_
LIO
Tnnrr
C"
r
I ^T
C"
Lit
LI2
3
f
CIS
HrCROPHONE
I REAR SIDE
HYDROPHONE PHEAMP HOUSING
J L
Figure 13, Hydrophone preamplifier schematic.
-------
Q13) is coupled into a low impedance balanced transformer (Tl). The
output of the balanced transformer is connected to the signal cable
through contacts of a relay located inside the junction box at the
selected station.
Power to the preamplifier and the relay is obtained from the
hydrophone selector switch on the receiver-decoder console. The necessary
dc power supply is included in the console.
Hydrophone Assembly
The hydrophone consists of two aluminum housings bolted together
as shown in Figure 14 and Figure 15.
The rear housing contains three sets of discs mounted on rubber
spacers to insure their vertical centering within the housing. Cork
rubber is glued to the bottom and the side of the discs.
The front housing contains the preamplifier printed circuit
board and three sets of discs that are also mounted on rubber spacers
to provide centering.
The two housings are filled, one at a time, with a urethane potting
compound to within 3 mm of the top edge of each housing. The housing
is placed in an evacuation chamber for 60 minutes to remove air bubbles
\,
from the urethane compound. The urethane is allowed to cure for
approximately 24 hours.
Special care was taken to protect against rf interference (RFI).
Any rf signal present in the vicinity of the channel will be picked up
by the hydrophone assembly. Some signals will of course be picked up
more readily than others; such factors as wavelength, power level,
depth of the hydrophone under water, direction from which the signal is
generated, etc., will determine the final intensity of the signal which
is received. Thus, to eliminate false readings or even mask the
signals from the fish tags, the RFI must be greatly diminished.
39
-------
41 ELECTROSTATIC-
Cr Cg, etc.
PZT Cylinders
Figure 14. Hydrophone assembly.
40
-------
Since the hydrophone ceramic elements must be exposed to the
sonic signals, a solid metal shield cannot be used to eliminate the
RFI. What is required is a material which will inhibit RF signals
and allow sonic signals to pass unattenuated. Screen materials
(20 x 20 mm to 100 x 100 mm meshes) have been used with varying success
to solve similar problems associated with military sonar systems.
Unfortunately, when tried on the present hydrophones, the attenuation
to power line frequencies and the RF signals present on them, was not
sufficient to prevent masking of several tag signals. The final
solution to most of the RFI was the use of liquid silver paint
(Acheson-Electrodag # 5041) which was brushed on the outer surface
of the hydrophone (Figure 14).
Figure 15. Hydrophone assembly,
-------
RECEIVER
General
The sonic receiver is designed to monitor various sonic sources
within a frequency range of 50-365 kHz. It is a sensitive and selective
receiver using super heterodyne circuitry built around modern integrated
circuits and discrete components. Tuning is accomplished by varying the
voltage controlled local oscillator thus giving the receiver the feature
of optional remote control tuning. The modification for EPA application
provides a 20 step frequency selection feature with vernier adjustment
for each step, rather than continuous tuning and a mechanical mounting
arrangement which is compatible with the other units of the entire EPA
fish tracking system. The final receiver is mounted in the console,
Description
The selected signal source (described elsewhere in this document)
is connected to the 600 ohm winding of the input transformer (Figure 16).
A band pass filter consisting of capacitors Cl, C2, C3, and C4
and inductors LI, 12 and L3 is included to reject signals outside the
tuning range of the receiver. The signal input level to the first mixer,
Yl, is adjustable by front panel control Rl - RF Gain, The output of Yl
is the difference frequency between the input signal and the local
oscillator, Y6.
The local oscillator frequency range is determined by the
components, C13, C15, L4 and the voltage across CRT, the tuning diode.
The diode tuning controls are connected to pin 9 on the PC board.
Rotating TAG SELECTOR'switch connects resistor networks to the tunina
diode which sets the receiver channel frequencies. Each channel frequency
may be adjusted over a limited range by the TAG TUNI FIG CONTROLS.
The dual outputs of the mixer are connected to two band pass filters
at the IF frequency, 455 kHz. These filters establish the selectivity
42
-------
«{-
CO
C2 LI
Figure 16. Receiver schematic.
-------
characteristics of the receiver. The choice of bandwidths is made by
selector switch BANDWIDTH. The 455 kHz signal is amplified by two
transformer coupled integrated circuit amplifiers, Y2 and Y3.
Signal detection occurs by the product mixer action of Ql, a dual
gate transistor. Transistor Q2 and its associated circuitry form a
beat frequency oscillator (BFO). The output siqnal of the product
detector is the difference frequency between the IF frequency and the
BFO frequency. The BFO is adjusted for approximtely 100 Hz.
The audio signal is amplified to earphone level and speaker level
by 1C U4. The audio level is adjustable by the AUDIO GAIN control (AGO).
The output of the product detector is also fed into the AGC
amplifier consisting of transistors Q3, Q4, and Q5. The AGC output, pin 5,
is connected to the IF amplifiers, integrated components (1C) U2 and LI3,
pin 4, through the SIGNAL STRENGTH meter.
Jacks are provided on the front panel for earphone, loudspeaker,
and oscilloscope monitoring of the audio signal and a jack is provided
to monitor the AGC level.
Operation
The receiver is made operative by applying power to the system.
To monitor a specific tag, rotate TAG SELECTOR switch to the desired
channel. Controls AUDIO GAIN, RF GAIN and the appropriate control
of the TAG TUNING CONTROLS group may be adjusted to produce a satisfact-
ory signal. The relative signal strength of the received tag signal
may be observed by the deflection on the SIGNAL STRENGTH meter. This
is a relative reading only, any deflection above the steady state level
can be compared with other signals.
Adjustments
There are only two adjustments to be made on the receiver, the local
oscillator and the beat frequency oscillator (BFO) frequencies. These
adjustments are to be found on the back of the enclosure holding the
44
-------
receiver electronics. There are two access holes in the compartment cover,
marked |_0 level oscillators and BFO beat frequency oscillator. To
adjust the local oxcillator, rotate the TAG SELECTOR switch to Tag 1, and
set the Tag Tuning Control to mid scale, and the BANDWIDTH switch to 1.5
kHz position. Connect a signal generator to the input of the receiver and
adjust its frequency to 50 kHz. Using an insulated hexigon alignment tool,
adjust the case of L4, maximum signal strength, as indicated in the SIGNAL
STRENGTH meter. The output level of the signal generator should be about
11 microvolt for this adjustment. The beat oscillator may be adjusted to
produce a beat of about 1000 Hz.
TEMPERATURE DECODER
The fish tags used in this system produce a continuous series of
energy bursts whose period varies as a function of temperature. The
temperature information is contained in the period of the tag, and the
temperature decoder extracts this information by measuring the time
between the signals from the tag and applying to that quantity a
multiplication constant.
The measurement of time is accomplished by counting internally
generated clock pulses during the interval between tag signals. From
this point the decoding process involves a scaling operation which is
performed by first converting the contents of the counter to an analog
voltage and then modifying that voltage by means of two operational
amplifiers whose function it is to compensate for the variations in the
period-versus temperature characteristics of the tags. The output of
the operational amplifiers is then displayed as temperature on a digital
(voltmeter) display on the front panel.
Operation
The temperature decoder operates as follows: A timing module (U3),
shown in Figure 17, is used as a source of clock pulses (approximately
900 Hertz) which are accumulated in a counter (U4). Each time a signal
45
-------
ONE SHOT
ASTABLE
CT»
TEMPERATURE DECODER
AUDIO
INPUT
FROM
RECEIVER
C!
R44
2.2K
D4IN4I48
KJOTE 3
R36220K
&AIW/OFFSEJ
OUTPUT
Figure 17. Temperature decoder.
-------
is received by the decoder, the contests of the counter are loaded into
a shift register, (U5, U6 and U7), and the counter is reset. This is shown
in Figure 18. Immediately after the counter is reset, it begins a new
counting cycle which continues until the subsequent audio pulse is
received. If another pulse is not received, the counter will accumulate
to maximum capacity and load "all ones" into the registers, corresponding
to "mininjum"temperature.
Monostable 111, which is triggered by the initial signal from the
receiver, has a pulse width long enough to prevent the monostable from
being re-triggered by echo pulses of the tag^generated acoustic signal.
The signal which loads the register and resets the counter is generated
by the monostable U2 which was fired by the leading edge of a pulse
generated by the monostable Ul (Figure 18).
A digital to analog converter, 1)8, provides an output voltage
which is proportional to the contents of the register. The resistance
versus temperature curve of the fish tags are to a large extent linear.
Non-linearities occur, however, at the higher and lower temperature
extremes. These nonlinearities are corrected by the Op Amp U9 and the
feedback circuitry composed of R39, R40, R41, and D4 at one extreme and
by R36, R37, R38 and D5 at the other extreme. Offset and gain adjust-
ments of the Op Amp translate into calibration of 0 C and at 25£C.
These are performed by potentiometers on the temperature decoder unit
which are accessed through holes in the front panel. Each tag has a
corresponding set of potentiometers for high and low temperature
calibration.
Calibration
Calibration is performed as follows: The TAG SELECTOR is set to
select the desired tag. The tag is immersed in a water bath at 0 C
for about-10 minutes, allowing: it to~ stabilize. & portable
hydrophone is also inserted in the water and connected to the hydrophone
input on the rear of the console. The potentiometer, identified under
the zero degree column by the appropriate tag number, is adjusted until
47
-------
LOCATION
Audio
Input
FUNCTION
Signal From Tag
Filter
Output
Output
Ul (Pin 3)
t^SOmsA inhibit Re-Trigger
Ar-
Output —
U2 (Pin 3)
150/JS
Load Register
(Leading Edge)
Base
of Q4
U4 Pin 11
t * 50 ns
A/—
Reset Counter
NOTE: Not To Scale
Figure 18. Signal descriptions of temperature decoder,
48
-------
the temperature display meter indicates zero. The tag is then immersed
in water at 25 C for 10 minutes and the corresponding potentiometer
in the 25 C column is adjusted until the display reads 25 C.
Ordinarily there is very little inter-relationship between the
zero and 25 degree potentiometers; however.it is recommended that
the cycle be repeated at least once.
RECEIVER-TEMPERATURE DECODER CONSOLE
The receiver-temperature decoder console provides the controls
and displays by means of which fish are located in a channel and their
temperature is determined. The. console contains three elemental assemblies.
They are: The receiver, the temperature decoder and the power supply.
Operation
Up to a maximum of twenty tags may be addressed using the TAG
SELECTOR switch on the panel of the console .(Figure-19). Each tag is
individually selected by one of 20 TAG TUNING CONTROLS. A separate
selector, VCO VOLTAGE switch, is available to display the VCO voltage
associated with each tag. The VCO voltage is directly related to frequency.
This is an excellent check on the tag selected for temperature monitoring.
This feature was readily available due to the circuitry chosen to tune
the receiver. By using voltage tuned diodes to accomplish frequency
tuning, one need only read the voltage on the tuning diode to have a
frequency analog. By using the digital display on the panel to indicate
VCO voltage, a precise analog measurement of frequency is available.
This measurement is representative of the frequency to which the receiver
is tuned and can be recorded for reference purposes.
The frequency bandwidth of the receiver may be set to either
1.5 kHz or to 4 kHz by a selector switch located on the panel. The 4 kHz
bandwidth permits a broader spectrum of signals to be monitored by the
receiver, while the 1.5 kHz bandwidth restricts the reception to a smaller
spectrum of frequency and at the same time reducing the amount of noise
picked up by the receiver. This has the effect of increasing the signal-
49
-------
Figure 19. Receiver decoder console.
to-noise ratio of the reception by approximately the ratio of the band-
widths. Thus, by switching from 4 kHz to 1.5 kHz the signal reception
is improved by approximately 2.6. This is an approximately 50% increase
in reception range.
The panel contains controls for adjusting both the RF and audio
gains of the receiver. It also provides a meter to monitor signal
strength, and phone jacks for additional monitoring of the receiver
output and AGC voltage and for connecting external phones or an external
speaker. Since the tag temperature and tag tuning parameters can be
represented as analog voltages, it has been convenient to display both
on the same display. In order to do this the appropriate scaling circuit-
ry has been designed into both the temperature decoder and the receiver.
The display is, in fact, a digital voltmeter. The desired parameter may
be displayed by selectively throwing a toggle switch.
Calibration
Calibration of the temperature tags is performed at 0 C and at
25 C. Twenty sets of calibration potentiometers are provided on the
console panel for this purpose. These are related to the 'TAG TUNING
-------
CONTROLS-so that a tag is selected by the TAG SELECTOR, the appropriate
set of calibration pots is applied to the temperature decoder.
The power assembly is located in the console and is attached to
the bottom frame. The assembly houses two commercial power supplies,
and a terminal strip. A New Jersey Electronics Model RL 1-5 provides
3 amps at 5 volts. The plus and minus 15 volt source may be either
of the following units: Zeltex Model Z 15AT100DP, Analog Devices
Model 902, or Semiconductor Circuits Model P741-1015-
INSTALLATION
System installation was completed during the period of 25 to
28 June 1974 at the Monticello Ecological Research Station (Figure 1).
The receiver-decoder console was located inside the blockhouse
adjacent to the conduit box provided for the cable runs to the channels.
A ground strap was connected from the console to the conduit.
The power and signal cables from the console to the junction
boxes at the channels were run through a 10.2 cm (4 inch) steel conduit
(Figure 20) to provide additional shielding from the power lines. The
conduit is strapped to each junction box along the channel to insure that
a low resistance path is provided for better shielding of the cables
from the power lines. Both ends of the conduit were connected to 4 meter
rods driven completely into the ground. A cable inter-connecting box
was installed within the junction boxes at each station to provide
additional shielding and to protect the exposed cable wires from moisture
(Figure 21). These grounding and shielding precautions are all essential
to minimize ground loop currents within the overall hydrophone, preamplifier,
cables and console system.
The inter-connecting boxes have romex connectors mounted on them
for the entry of the signal and power cables. The boxes contain a
terminal strip, provided for the inter-connection of the cable wires,
and a relay. This relay is used to activate the appropriately selected
preamplifier.
51
-------
Figure 20. Steel conduit runs from blockhouse.
Figure 21. Juction box.
!
-------
The shield of each power and signal cable entering the box is
connected to the box by an adjustable clamp on the romex connector
(Figure 22). The shields are connected together again on the terminal
strip.
The output cable from each hydrophone is run through a 19 mm (3/4 in.)
conduit to the junction boxes. The conduit is connected to the hydrophone
housing by a compression type connector that is mounted through one
wall of the hydrophone housing (Figure 23).
Figure 22. Interconnecting box.
-------
r
COiOSOLf
leiug
VFLLOVJ
S. Hi £1.0
SHISL&
0 0
0^0
-0*0
WHlTf
=D-tO
2)-^0
3_
0^<
0 0
WHITE
SREY
BLOC
NEU.OW
QLUE
WHITE
tt
-^
Figiire 23. Interconnecting box wiring diagram.
NOTE: Each cable feeds through 1.27 on (1/2 inch) romex electrical
connector mounted through side of box.
54
-------
SECTION VI
PROBLEMS AND SOLUTIONS
Tests conducted at the Monti cello Ecological Research Station
revealed two interference problems that resulted in loss of reception
of signals transmitted by ultrasonic transmitters.
The two problems were electromagnetic interference (EMI) and
radio frequency interference (RFI), both of which were being generated
by the overhead power lines.
RADIO FREQUENCY INTERFERENCE
During the early phases of the program an RFI evaluation of the
area was made, see Appendix A, over the frequency range of DC to 50 kHz.
A quick search at higher frequency was also made; however, no signals
were evident up to, and beyond, 200 kHz. The signal frequencies below
50 kHz appeared to be of a level which would not cause serious RFI
problems. Subsequently, the power companies started to transmit signals
which are now causing problems. Telemetry signals used for relay switch-
ing are transmitted along the power lines. The frequency of the signals
105 kHz, 146 kHz, and 192 kHz fall within the receiving bandwidth of
the ultrasonic receiver, and are close to the frequencies of tags
numbers 6, 10, and 14. These signals were being picked up at low levels
by the signal and power cables and at higher levels by the hydrophones.
The signals and, to some extent, their harmonics, could also be detected
at low levels within the blockhouse. The level of these signals were
two or three times higher with the hydrophone located in channel 1 as
compared to the hydrophone being located in channel 2.
Testing of electrostatic shielding was conducted to determine to
what degree the interference could be reduced. Shielding materials
consisting of mu-metal, aluminum screen and steel screen, were evaluated
for their attenuation properties. A frequency generator was used to
produce the RFI signals. The use of aluminum screen placed over the
55-
-------
hydrophone and connected to the housing resulted in attenuation of
RFI signals by 40 dB where as the steel screen resulted in an attenuation
of only 27 dB.
A solid mu-metal shield placed over the hydrophone and connected
to the hydrophone housing resulted in an attenuation of 58 dB of the
RFI signals. Additional testing using silver paint resulted in RFI
attenuation of 67 dB. Two hydrophones were built and shielded with the
silver paint. These hydrophones were tested at the test site and
resulted in the attenuation of the RFI signals to a near acceptable
level, that is one can reasonably isolate and recognize tag and RFI signals.
Additional attenuation of the undesired RFI signals was gained
when the hydrophone cable was inserted in the 19 mm (3/4 in.) conduit and
when the signal and power cables were connected inside the inter-connecting box
within the junction boxes. Further attenuation was noticed when the
front panel of the receiver was cleaned on the inside with steel wool
allowing better contact of the front panel to the receiver housing.
The RFI signals at this point had been reduced to an acceptable
level and resulted in little, if any, interference with the reception
of the tag signals.
ELECTROMAGNETIC INTERFERENCE
The EMI signals were picked up by the signal and power cables
and were induced into the receiver at levels high enough to block
the reception of the lower level acoustical signals. A common mode inter-
ference problem resulted due to these high level signals that required
changes to be made in the hydrophone preamplifiers and the receiver.
To reduce the common mode interference, a balanced transformer was
placed at the receiver input and at the hydrophone preamplifier output.
With the addition of the balanced transformers and additional filtering
within the preamplifier circuits the EMI problem was reduced to an
acceptable level. See Appendix A on electromagnetic testing for additional
detailed information.
56
-------
STABILITY OF EQUIPMENT
The reciever-decoder components were designed to be operated
within a controlled temperature environment. Temperature variations
of more than a few degrees centigrade will cause variations in the
frequency of the local oscillator within the receiver. This oscillator
has to be capable of being continuously tuned over a wide frequency
range. Crystals could not be used in the receiver unless the tags
also contained crystals to maintain a constant frequency at each of
the tag frequencies.
The hydrophone preamplifiers were designed to operate over a
wide temperature range due to their location within the channels. The
operation of the preamplifiers is constant over the temperature range
of 0 C to 40 C and over a voltage range of 10 to 18 volts.
TAG LEAKAGE AND AGING
Some problems were experienced with water leaking into the
electronics of the tags. The problem was solved by covering the external
2 wires which are used to activate the tag with GE Silicone Rubber Sealant,
RTV-102. To minimize aging effects of tags, all units were stored under
refrigeration.
ACOUSTIC ATTENUATION PROBLEMS
Serious tag signal attenuation was observed which severely limited
the system range. Under some conditions, the range was limited to 3 meters,
The problem was caused by vegetation, drifting algae, and, in some cases,
bubbles in the vicinity (within a 3 meter area) of the channel inlet.
To solve the bubble problem, the area was screened off and the fish
were excluded from the inlet area.
57
-------
REFERENCES
1. Lonsdale, E.M. and G.T. Baxter. Design and Field Tests of a Radio-Wave
Transmitter for Fish Tagging. Prog. Fish-Cult. 30_:47-52, 1968.
2. Mackay, R.S. Bio-Medical Telemetry: Sensing and Transmitting Biological
Information From Animals and Man. New York, John Wiley and Sons, Inc.,
1968. 388 p.
3. Adams, L. Progress in Ecological Biotelemetry. Bioscience. 15:83-86,
1965.
4. Albers, V.M. Underwater Acoustics Handbook. University Park, Pennsyl-
vania State University Press, 1960. 290 p.
5. N.W. Fisheries Centre. Electronic Tags Used to Study Adult Salmonid
Movements in the Columbia River Basin. National Marine Fisheries Ser-
vice, Seattle, Wash. 1974. p. 1-7.
6. Baldwin, H. Marine Biotelemetry. Bioscience. l_5_:95-97, 1965.
7. Chapman, C.J., A.D.F. Johnstone, and G.6. Urquhart. Preliminary Acous-
tic Tracking Studies of Nephrops norvegicus. Department of Agriculture
and Fisheries for Scotland, Marine Laboratory Internal Report IR 74-2
(Unpublished typescript), 1974. 14 p.
8. Hart, L.G. and R.C. Summerfelt. Homing Behavior of Flathead Catfish
Tagged with Ultrasonic Transmitters. Proc. 27th Ann. Conf. S.E. Assoc.
Game & Fish Comm., Oct. 14-17, 1973, pp. 520-531.
9. and H.F. Henderson. Instrumentation Problems in the Study of
Homing in Fish. In: Biotelemetry, Slater, I.E. (Ed.). New York, Per-
magon Press, 1963.
10. and I. Mohus. 1973 Equipment. Fish Telemetry Report 2. SINTEF
and Reguleringsteknikk, 7034 NTH - Trondheim, Norway. Report STF48
A73051, 1973. 29 p.
11. Horrall, R.M., H.F. Henderson, and A.D. Hasler. Ultrasonic Tracking of
Migratory Fishes with an Internal Tag. Presented at Amer. Fish. Soc.
Meeting. .Portland, Oregon. 1965. 2 p.
12. Ichihara, T., M. Soma, K. Yoshida, and K. Susuki. An Ultrasonic Device
in Biotelemetry and Its Application to Tracking a Yellowtail. Bull.
Far Seas Fish. Res. Lab. 7:27-48, 1972.
58
-------
13. Johnson, J.H. Sonic Tracking of Adult Salmon at Bonneville Dam, 1957.
U.S.D.I., Fish and Wildl. Serv., Bur. Comm. Fish. Fishery Bull. 60:471-
485, 1960. ~"~
14. Johnson, R.K. Acoustic Tracking of Individual Fish. Oregon State Uni-
versity, School of Oceanography,Corvallis, Oregon, 97331.
15. Koo, T.S.Y and J.S. Wilson. Sonic Tracking Striped Bass in the Chesa-
peake and Delaware Canal. Trans. Amer. Fish. Soc. 1_0]_:453-462, 1972.
16. Li scon, K.L. Sonic Tags in Sockeye Salmon, Oncorhynchus nerka, Give
Travel Time Through Metropolitan Waters. Marine Fisheries Review (U.S.).
35:38-41, 1973.
17. Malinin, L.K. Use of Ultrasonic Transmitters for Tagging Bream and
Pike. Report I. Reaction of Fish to Net Webbing. Biologiya Vnutren-
nikh Vod (USSR) Informatsionnyi Byulleten (Biology of Inland Waters. In-
formation Bulletin). 7;64-69, 1970.
18. Monan, 6.E. and D.L. Thome. Sonic Tags Attached to Alaska King Crab.
Marine Fish. Rev. 35:18-21, 1973.
19. Novotny, J. and G.F. Esterberg. A 132-Kilocycle Sonic Fish Tag. Prog.
Fish-Cult. 24:139-141, 162.
20. Poddubnyi, A.G., L.K. Malinin and V.V. Gaiduk. Experiment in Telemetric
Observations Under Ice of the Behaviour of Fish. Biologiya Vnutrennikh
Vod (USSR), (Biology of Inland Waters). 6:65-70. Fish. Res. Board Can.
Transl. Ser. 1817, 1970. ~"
21. Steadman, J.W. A Short Range Underwater Biotelemetry System. Trans.
Int. Telemetering Conf. 3;316-324, 1967.
22. Tesch, F.W. Experiments on Telemetric Tracking of Spawning of Migra-
tions of Eels (Anguilla anguilla) in the North Sea. Fish. Res. Board
Can. Transl. Ser. 2724: 29 p. 1972.
23. Trefethen, P.S. Sonic Equipment for Tracking Individual Fish. U.S.D.I.
Fish and Wildl. Serv., Spec. Sci. Report (Fish) No. 179, 1956. 11 p.
24. Ziebell, D.C. Ultrasonic Transmitters For Tracking Channel Catfish.
Prog. Fish-Cult. 35_:28-32, 1973.
25. Lawson, K.D. and F.G. Carey. An Acoustic Telemetry System for Transmit-
ting Body and Water Temperature From Free-Swimming Fish. Woods Hole
Oceanographic Institution, Woods Hole, Massachusetts, May 1972.
26. Rochelle, J.M. and C.C. Coutant. Temperature Sensitive Ultrasonic Fish
Tag, Q-5099. Oak Ridge National Laboratory, Oak Ridge, Tenn., Report
ORNL-TM-4438. 1973. 26 p.
59
-------
27. Ferrel, D.W., D.R. Nelson, T.C. Sciarrotta, E.A. Standora and H.C. Car-
ter. A Multichannel Ultrasonic Marine Bio-Telemetry System for Monitor-
ing Marine Animal Behaviour at Sea. In: Instrumentation in the Aero-
Space Industry, Washburn, B. (Ed.). Pittsburg, Instrument Soc. Amer.,
J9_:71-84, 1973.
28. Gaiduk, V.V. and L.K. Malinin. An Informational Ultra-Sonic Transmit-
ter for Biotelemetric Investigations. Akademiya Nauk SSSR, Institut
Biologiya Vnutrennikh Vod, Informatsionnyi Byulleten (The Academy of
Sciences, USSR, Institute of Biology of Inland Waters, Information Bul-
letin). ]_2:74-78, 1971.
29. Roland, B. 1973 Experiments. Fish Telemetry, Report 3. SINTEF and
Reguleringsteknikk, 7034 NTH - Trondheim, Norway. Report STF48 A73052.
1973. 22 p.
30. Luke, D.McG., D.G. Pincock and A.B. Stasko. Pressure-Sensing Ultrasonic
Transmitter for Tracking Aquatic Animals. J. Fish. Res. Board Can. 30:
1402-1404, 1973.
31. Malinin, L.K. and A.M. Svirskii. Application of Biotelemetry to Ich-
thyology. TsNIITEI of Fisheries Economic Utilization of the World Ocean
Fish Resources. J_:17-39 (in Russian). English Translation: Fish Res.
Board Can. Transl. Ser. 2707. 1972. 29 p.
32. Slater, A., S. Bellet and D.G. Kilpatrick. Instrumentation for Teleme-
tering the Electro-Cardiogram From Scuba Divers. IEEE Trans. Bio-Med.
Engng. BBE-16:148-151, 1969.
33. Summerfelt, R.C. and L.G. Hart. Performance Evaluation of a 74 Kilocy-
cle/Second Transmitter for Behavioral Studies on Reservoir Fishes. Proc.
25th Ann. Conf. S.E. Assoc Game & Fish Comm., Oct. 1971, pp. 607-622.
34. Lin, W.C. and W.H. Ko. A Study of Microwatt-Power Pulsed Carrier Trans-
mitter Circuits. Med. Biol. Engng. 6;309-317, 1968.
35. Shirahata, S. Effect of Externally Attached Sonic Tags on Fish Behav-
ior. (Freshwater Fisheries Research Laboratory, Nikko) Bulletin of
Marine Biotelemetry Research Group. No. 4, p. 3-12, Dec. 1971.
36. Andreeva, I.B. Scattering of Sound by Air Bladders of Fish in Deep
Sound-Scattering Ocean Layers. Soviet Phys.-Acoust. 1_0:17-20, 1964.
37. Batzler, W.E. Acoustic Target Strengths of Some Marine Animals. Pa-
per UW3, 73rd Meeting of the Acoustical Society of America, 19 April
1967.
38. Batzler, W.E., R.M. Regan and G.V. Pickwell. Resonant Acoustic Scatter-
ing from Air-Bladder Fishes. Paper D7, 75th Meeting of Acoustical So-
ciety of America, 21 May 1968. Abstract J. Acoust. Soc. Amer. 44:356,
July 1968. ~
60
-------
39. Hasler, A.D. Underwater Guldeposts. Madison, University of Wisconsin
Press. 1966. 155 p.
40. and W.J. Wisby. The Return of Displaced Largemouth Bass and
Green Sunfish to a 'Home1 Area. Ecology. 34_:289-293, 1963.
41. Mersey, J.B. and R.H. Backus. Sound Scattering by Marine Organisms.
In: The Sea, Vol. I. New York, John Wiley & Sons, Inc., 1962. p. 499-
507.
42. Love, R.H. Maximum Side-Aspect Target Strength of an Individual Fish.
J. Acoust. Soc. Amer. 46(3-Part 2):746-752, September 1969.
43. Nickles, J.C. and R.K. Johnson. A Digital System for Volume Reverbera-
tion Studies. J. Acoust. Soc. Amer. J50_:314-320, 1971.
44. Noerager, J.A., E.J. Rice and C.E. Feiler. A Method For Reducing
Groud Reflection Effects From Aoustic Measurements. NASA Technical TN
D-6666. March 1972. p. 1-67.
\
45. Raitt, R.W. Sound Scatterers in the Sea. J. Mar. Res. 7^:393-409, 1948.
46. Rochelle, J.M. Design of Gateable Transmitter for Acoustic Telemeter-
ing Tags. IEEE Trans. Bio-med. Engng. BME-21(L):63-66, 1974.
47. Weston, D.E. Sound Propagation in the Presence of Bladder Fish. In:
Underwater Acoustics, Vol. 2, Albers, V.A. (Ed.). New York, Plenum
Press, 1967. p. 56-88.
61
-------
APPENDIX A
ELECTROMAGNETIC FIELD MEASUREMENTS
TEST OBJECTIVES
The objectives of these tests are to determine the design criteria
for the instrumentation system with respect to the possibility of
interference from the power lines that are located overhead or from
the power generators themselves.
TEST SPECIFICATIONS
The electromagnetic fields present at the experimental channels
will be measured at various points along the channels in order to establish
a profile of the field at the site.
In order to relate the on-site test to a known set of conditions
a set of measurements will be made in Virginia on power lines of
similar construction and power capacity. By comparing the two sets of
data the on-site radiation contributed by the generators may be
determined,
A third set of measurements will be made in the Bayshore Systems
laboratory area where the fish tanks are located and where the in-house
experimental work is undertaken. The level of interference in the lab
i
can then be compared to that at the actual site and establish a bench
marker to work from in the design of the instrumentation.
The above mentioned tests are to be conducted in the frequency
range of from DC to 50 kHz, with equipment having a sensitivity of
1 microvolt, a sensitivity equal to, or better than, that of the propos-
ed test instrumentation. A search from 50 kHz to 200 kHz will also be
made to observe possible spurious signals.
62
-------
TEST RESULTS
The electromagnetic radiation tests were run &t the EPA test
channel at the Monticello Nuclear Generating Plant at Monticello,
Minnesota on October 5, 1972.
The Monti cello test site is 40 miles northwest of Minneapolis on
the Mississippi River. The soil condition at the site is that of sandy
loam to the channel bottom level. The overall site is flat and 8 to 10
feet above the river level.
Conditions of the test were with heavy fog during the first half
of the day and heavy overcast during the afternoon. The temperature was
in the high 40's and low 50's. Some water was evident on the ground
with surface mud indicating a rainfall earlier in the week.
The power plant was in operation during the test. Two or more
pieces of construction equipment were in operation during most of the
test. During the noon break of the construction equipment a test was
run to establish the effect of the equipment. There was no apparent
effect, with the radiation from the power line masking out any of the
interference from the construction equipment.
The map in Figure 24 illustrates the positions of each of the
tests that were run.
CONCLUSIONS
From the test fcesults, Figures 25 to 37, it is evident that the
signal levels present on the site, resulting from power line radiation,
is of a magnitude that will cause interference to the fish tracking
systems. The tests have indicated that the design of any system
instrumentation to be used on this site must be designed for operation
in a high nedsc level environment.
From the various tests that were run, both on site and local, the
radiation originates from the power lines rather than from the power
generating equipment or the nearby distribution station. This point
63
-------
CT>
SCALE
1" = 200'
Figure 24. Electromagnetic test sites - Monti cello Ecological Research Station channels,
Monti cello, Minnesota.
-------
200
S-
O)
S-
O
S_
o
•=c
H-
—I
O
Q
LoJ
UJ
Qi
20
TEST LOCATION: MONTICELLO
TEST DATE: 10/5/72
10 100 1000
FREQUENCY, hertz
Figure 25. Electromagnetic field measurements - composite tests.
-------
en
Ol
O)
Q.
200
180
160
•5 140
OJ
>
O
JD
"3
-------
200
180
160
140
120
100
80
60
40
20
TEST LOCATION: MONTICELLO
TEST DATE: 10/5/72
cu
4->
o>
s_
a)
o.
1/1
o
s_
O>
>
o
CD
H-
O
O
LU
tJ
10
T"
100 1000
FREQUENCY, hertz
Figure 27. Electromagnetic field measurements - test number 2.
-------
CT>
OC
O)
O
s_
O
'i
O
o>
>
O
JO
(O
CD
-------
XQ
s_
o>
-M
0)
o>
Q.
I/)
O
S_
(J
01
>
O
.Q
iTJ
cs
I—
O
Q
UJ
UJ
UJ
o:
200
180
160
140
120
100
80
60
40
20
10
TEST LOCATION: MONTICELLO
TEST DATE: 10/5/72
100
1000
HI
IT
II
FREQUENCY, hertz
Figure 29. Electromagnetic field measurements - test number 4.
-------
'—1
o
200
Ol
-»->
CD
E
s_
OJ
o.
1/1
o
i-
c_)
'i
o
O
JD
fO
LiJ
CS
O
UJ
o
UJ
DC.
20
10
TEST LOCATION: MONTICELLO
TEST DATE: 10/5/72
100 1000
FREQUENCY, hertz
Figure 30. Electromagnetic field measurements - test,
5,
-------
200
180
160
140
120
100
80
60
40
20
TEST LOCATION: MONTICELLO
TEST DATE: 10/5/72
s_
o>
s_
O)
o.
o
LU
LU
LU
o:
10
100
1000
FREQUENCY, hertz
Figure 31. Electromagnetic field measurements - test number 6.
-------
ro
200
£ 180
O)
E
O
>
O
160
140
u
e 120
O
JD
(O
100
80
< 60
o
>
o
UJ
Qi
40
20
10
TEST LOCATION: MONTICELLO
TEST DATE: 10/5/72
100
1000
FREQUENCY, hertz
Figure 32. Electromagnetic field measurements - test number 7.
-------
CO
i-
CD
O>
O)
Q_
1/1
O
>
O
s_
o
O)
>
o
-Q
fC
.a
T3
C3
LU
LU
40
20
10
TEST LOCATION: VA. TEST
TEST DATE: 10/5/72
TOO 1000
FREQUENCY, hertz
Figure 33. Electromagnetic field measurements - comnosite tests.
-------
Ol
•t-1
O)
Ol
CL
200
180
160
o 140
>
O
S-
u
0)
>
O
_Q
(O
CD
I—
O
>
O
UJ
•>
I—I
UJ
UJ
120
100
80
60
40
20
TEST LOCATION: FAUQUIER COUNTY, VIRGINIA
TEST DATE: 9/15/72
1000
FREQUENCY, hertz
Figure 34. Electromagnetic field
- test, numhpr 1.
-------
CJ1
s_
OJ
+->
o>
s_
Q.
l/l
O
S-
u
O
-Q
(T3
UJ
UJ
a:
20
10
TEST LOCATION: FAUQUIER COUNTY, VIRGINIA
TEST DATE: 9/15/72
100 1000
FREQUENCY, hertz
Fiaure 35. Electrnmaanetir. fipld measurements - test mimhpr ?,
-------
200
O)
CD
E
S-
Q-
O
S-
o
QJ
>
O
J3
(O
CD
t—«
LU
LU
o:
60
40
20
10
TEST LOCATION: FAIRFAX CO, VA.
TEST DATE: 9/20/72
100 1000
FREQUENCY, hertz
Fiqure 36. Electromaqnetic field measurements - test number 3,
-------
TEST LOCATION: BAYSHORE LAB AREA
TEST DATE: 10/7/72
i-
OJ
4->
QJ
S-
OJ
Q.
O
>
O
s_
O
01
>
O
-Q
fC
T3
LU
CD
Q
LU
CJ
LU
a:
40
20
1000
FREQUENCY, hertz
Figure 37. Electromagnetic field measurements - test numhpr 1.
-------
is made obvious when matching the curves from the different test
locations.
There are two distinct areas of interference:
1. Those of power lines related frequencies up to about
10 kilohertz generating levels that will interfere
with the sensor signals.
2. Above 10 kilohertz there are those carriers that are
imposed on the power lines. The carriers are used for
the relaying of switching and distribution information
by the Northern States Power Company. These signals
usually fall in the range from 14 to 50 kHz.
It should be noted that none of these type of signals
were detected. It is anticipated that when the power
plant is fully operational their presence will be
apparent. Normally, the measured level of the signals
are not as high as the power line frequencies but are
well above the ambient noise level on site. Since the
signals are in the range of the sensor information
they are not easily filtered out of the system.
As expected the signal levels vary over the site and are directly
related to the distance from the power lines. The area of high test
levels are those that are directly under and run parallel to the power
lines. Unfortunately the laboratory building itself is set in a high
level area.
The most severe problem to face the system designer is the poor
ground that is present on the site.
The soil condition at the pond is a fairly high grade of sand,
with the top 15 - 25 cm mixed with an organic base. The sandy soil,
which showed good drainage characteristics during the test, will not
hold moisture to any great extent. This point is also illustrated
by the need of an additional sealer on the bottom of the channel to
prevent leakage and errosion.
78
-------
The problem presented by the sand is similar to many instrumentation
and communication systems installed on islands or in soil that does
not exhibit good ground characteristics. Although ground conductivity
tests were not made, a good earth ground should be present at the
water table, which in this soil should be at the river level, about
3.0 - 3.7 meters below the surface.' This would place the reference ground
at about 2.4 meters below the bottom of the pond. In order to maintain
a reliable earth/ground the ground plane should be somewhat below
this in order to compensate for a shift in the water table. It is
conceivable that the ground level may be somewhat higher if there
are deposits of minerals. In observing the soil in the channels this
does not seem the case.
As stated the sandy soil will present a variable ground condition
to the system designer. During or after a rainfall the ground conditions
will be good with a low ground resistance. When the soil dries out, the
ground resistance becomes higher, the sand becomes an insulator and the
so-called ground is now above the earth/ground reference and a resistance
is presented between the instrumentation and the ground reference.
The sketch in Figure 38 illustrates the channel in the sand base
and the ground resistance (Rg). -;As shown, the channel when full of water
is a rather large, long electrical conductor.
As shown in the site plan the four channels are located parallel to
a power line for a distance of about 183 meters. The four channels are
spiced out from the power lines to a distance of about 45.7 met'ers (150 feet),
As shown in Figure 39, the channels act as four isolated conductors
spaced progressively away from the 345KV power line. The channels, like
any other conductor under the same conditions, will act as the secondary
of a transformer with a voltage generated in the channels due to the
CQM&ling to the power lines. Where the instrumentation lines are run
above electrical ground, a voltage will also be induced into the
conductors. A greater problem will exist where a sensor is "grounded"
79
-------
POWER LINES
SENSOR
immiimiiii [minimi
GROUND
RESISTANCE >
WATER TABLE OR ELECTRICAL GROUND
38, Ground con#iti«fls at the
KeltttteCTe' Ecological Re-
search Station.
80
-------
—55—
345 KY
00
•e
r
OJ
I,
r
r
x \\ \
INDUCTIVE COUPLING
Channel No. 1
Channel No. 2
Channel No. 3
Channel No. 4
KL
Re i
WV—|
Figure 39. Ground loop circuits,
-------
at the channel but will actually conduct the channel voltage (V ) back to
the instrumentation at the laboratory.
Figure 40, illustrates the methods in which the various interfering
voltages are introduced into the system. There are sources of interfering
voltage on both the sensor end and the load end of the system. On the
sensor end a voltage is generated when the induced current from the shield
flows through the ground resistance (Rg) with a voltage drop I Rg. An
addition to the ground loss is the voltage developed from the signal
current in the sensor. Since the two generators are in series both of
the voltages will be present on the center conductor of the line. In the
case of the channels nearest the power lines, the voltage induced may be
greater than the voltage generated by the sensor. The induced voltage
(or noise) from the power line may be of such a magnitude that makes the
detection of the sensor signal impossible. Some relief from the ground
loop may be had by using twin conductor cables with a floating ground
(common) to the sensor. Using the two conductor cables will keep the
line induced signal off the signal line but will not protect against
the noise from being induced on the shield, and conducted back to the
load end.
On the load end there are two.-possible sQueees ofggwo^d loop
voltages. One of the sources is the same as in the sensor end where the
loop voltage is developed across the ground resistance which is in series
with the load. Again, like the sensor this condition may be relieved
if two conductor cable are used. If the two conductor cable or balanced
system is used then some type of transformer or other balanced type of
transducer input circuitry must be used such as differential amplifiers.
In the case of the balanced input the ground loop components will be
dependent upon the balance of the system, at best the ability to hold
balance is not better than about 0.1%. In the differential amplifiers
or op amps this is referred to as common mode rejection.
Another problem associated with a pper ground system is cross
talk that is generated by the common sensor signal currents flowing
82
-------
(fa 345 KV
03
co
PI
JL
Dt
K,
Channel No. 1
~~AAACL
T1TR
Channel No. 2
LVW-L
Channel No. 3
—WV-L
j-jAAAr—
Channel No. 4
—AA/V-L
Figure 40. Induction coupling into pool conductor.
-------
through a common ground resistance. In this case a portion of each
of the sensors will show up at the input of every other sensor load. This
means that in every channel there will be the desired signal and three
undesired signals at various levels. In some cases this can drastically
reduce the value of the received information causing a low signal/noise
ratio.
A serious problem can also arise from a poor ground with lightning
or similar phenomena. With the channels setting above ground it is
conceivable that the channel could build up a high potential thhough the
various discharge phenomena. For this reason alone the channels should
be well grounded, in the same way that a building or other structure
is grounded.
RECOMMENDATIONS
As a result of the testlthe following recommendations are made:
a. Receiver Sensitivity:
The sensor receiver sensitivity should be maintained as
high as possible (insensitivity) while retaining the
necessary system performance. A receiver with high
sensitivity will be prone to over load (blocking) and
generation of intermodulation products and spurious
outputs.
b. Frequency Selection:
The operating frequencies of the system should be main-
tained as high as possible while retaining the other
performance factors. By all means the operating frequency
should be above 50 kHz.
c. Signal Levels:
The signal level present in all areas of the system must
be maintained at a relatively high level with high
signal to noise ratios. The use of local preamplifiers
84
-------
c. Signal Levels: (Continued)
at the output of the sensors and intermediate amplifiers
along the line is recommended.
d. Shielding:
Shielded two conductor cables should be used for all
signal lines. In addition to the braided shield on the
cable, the cables should be routed through conduit.
The conduit should be of a type affording-magnetic shield-
ing (cold rolled or drawn steel). At all joints the
conduit should be bonded with straps or welded. Each
of the sensor inputs should be terminated in a shielded
box. The termination boxes should be located along the
line and like the conduit should be well bonded to the
conduit itsalf. In addition to the shielding capability
the boxes should be weatherproof.
e. Filtering:
High pass filters will have to be iRCOroc?rated into the
design of the sensor and the receiver. The cutoff of the
filters should be as close to the operating frequency as
possible, preferably near 50 kHz. The filters at the
sensor end of the system may have to reject the line
frequencies as much as 40 to 50 dB which will require
5 or 6 section filter.
f. Earth/Ground:
The grounding problem that is present on the site must
tje surmounted. The twin conductor will help. For
the purpose of instrumentation the channel should be
paired, the lines running up the dike between the
channels or ponds. By pairing, a common ground can be
established for the ponds. A series of ground rods
must be driven into the dikes to a level that will provide
85
-------
f. Earth/Ground: (Continued)
a good ground system. At this time it is not known to
what depth or at what intervals the rod will be set.
An estimate would be that the rods should be driven not
less than 3.7 meters (12-feet) deep and should be spaced
every 15.2 meters (50-feet) along the line or at each of
the junction boxes. At best this will require a great
number of ground rods to be installed. The rods should
be one inch in diameter and made of copper clad steel.
The junction boxes should be bonded to the rod or better
yet mounted on the rods themselves.
86
-------
APPENDIX B
INVESTIGATION OF INDUCTIVELY COUPLED CIRCUITS
FOR USE IN FISH TAGS
Magnetically coupled transmitters and receivers are not new .
Contemporary portable broadcast receivers use a Ferrite antenna rod
which uniquely responds to the magnetic field of electromagnetic
radiation. As shown in Figure 41, a typical vertical antenna radiates
electric flux lines vertically with closed horizontal magnetic flux lines.
Hence, the electrical line is vertical and at right angles to the
magnetic lines. Since magnetic lines generate an electromagnetic fiild
when they cut electrical conductors, a multi-turned inductor resonant at
the traasmitted signal generates voltage at points a and b. The unique
sensitivity of the inductor to the magnetic field alone, causes the
broadcast portable receiver to be directional. Maximum signal is
received when the Ferrite rod is orthogonal to the radios:l1ne?f#Qffl"-tbee
transmitter and minimum when the end of the rod is pointed at the station.
The expression for E., the induced electromagnetic field (emf) is:
E1 = wn HA /Mr = n d0/dt
(The generator equation)
where H = received magnetic field intensity,
A = cross sectional area of the rod,
M = effective rod permeability,
The actual output voltage is a function of the electrical loss parameters
of the tuned circuit. "Q" is used to express these losses. Generally,
"Q" is expressed as a bandwidth function, that is f/Af. ButAf is
actually constrained due to the losses. In an air core coil these
losses are due to fringe effects, parasitic capacitances and leakage
87
-------
VOLTAGE
STANDING
WAVE
MAGNETIC
FLUX LINES
Figure 41, Electric and magnetic fields around
a half wave vertical antenna.
88
-------
flux. For a cored inductor the fringe effects, parasitic capacitances,
eddy losses and frequency constraints (loss tangents) of the core material
are the significant loss factors. Leakage flux is significant dti the
reduction of "Q" in air core inductors and certain ferrites reduce this
leakage and thereby, significantly increase the "Q".
Therefore, the significant requirements for a miniaturized induction
link system may be generalized. First, it should be noted that the general
electromagnetic radiation equation show the electromagnetic field radiation
p
is reduced as the inverse of the distance squared (1/r ), and the induction,
or near field, falls off as the inverse of the cube^of the distance (1/r3)
for distances greater than 10 coil diameters. The range of the induction
field communication is limited then to very short distances. The exact
distances are ultimately dependent of the size of the transmitter and
receiver inductance loops, although specific coil parameters seriously
affect the received signal strength. The coil parameters must be developed
empirically, and the range must be experimentally determined. For these
reasons, it is extremely difficult to calculate many of the performance
chanacteristies. Hence, when the receiving loop size is constrained,
various inductors must be made within these constraints and measured
electrically. From these data trends may be shown, trends which lead to
the final design figures. Such data are described in the next section.
EXPERIMENTAL DATA
Several experiments conducted to evaluate the qualities of solonoid
inductors are discussed in the following section. For these inductors,
selected core materials representing optimal choices from off-the-shelf
components were used. Various sizes of litz wire were used. These results
are shown in Figures 42, 43, and 44. Once having tabulated the basic
properties of these inductors with regard to core material, core size,
wire and inductance, further tests evaluting thefr performance as receiver
and transmitters were conducted. The data shown in these curves evaluate
the "Q" against frequency for Indiana General's Q-l and H material. The
89
-------
500 r
400
300
200
100
- HOT 30/46 Q
HOT 7/41 Q
HOT 30/46 H
HOT 7/41 H
CORE SIZE
_i
10 100 1000
FREQUENCY, kHz
10000
Figure 42. Variations of Q against frequency for various
litz wound solenoid inductors. Core size
as shown.
90
-------
500 -I
400 -
300 -
200 -
100-
54T 18/42 Q HAT
54T 7/41 Q MAT
54T 18/46 M MAT
54T 7/41 M MAT
10 100 1000
FREQUENCY, kHz
10000
Figure 43. Variations of Q against frequency for
various litz wound solenoid inductors.
Core size as shown.
91
-------
CORE SIZE
200 _
150 -
TOO -
50 -
140 T 7/46 Q MAT
150 T 7/41 Q MAT
HOT 30/46 H MAT
140 T T/41 H MAT
~1 1
10 100
FREQUENCY, kHz
1000
10000
Figure 44. Variations of Q against frequency for various
litz wound solonoid inductors. Core shape and
size as shown.
92
-------
litz wire size is shown. Since the highest output voltage is a function
of figure of merit "Q", then the higher "Q" values are desirable. Note
that "Q" is variable with material, number of windings, and wire type.
"Q" peaking is at various frequencies and is possible through judicious
selection of core material and wire type. "Q" was measured with a Boonton
Model 260 A Q-meter. It should be noted that the Boonton "Q" meter measures
"Q" with 8 ma. rms through the inductor at a scale multiplication of 1.
It does not measure the actual "Q" at low signal levels such as would
be experienced when the inductors are used in receiving systems as antennas.
In these cases, performance can only be measured and, in general, curves
are a good rational fit with the "Q" data.
For the actual use data, the test setup in Figure 45 was used for
the accompanying data in both Figure 46 and Figure 47. Power into the
transmitting inductor is provided by a current source with a maximum out-
put voltage of 25 volts rms. Since litz wire is not easily available in
larger sizes, some data is shown which obviously conceals some of the
2
I R losses in the transmitter inductance. Additional data for higher
powered sources was obtained by using a larger solid copper wire. The
"Q" of the transmitter coil is not especially important. Lower "Q" coils
merely require more voltage to drive the same amount of current as would
be obtained by a higher "Q" coil. This does not affect transmitter
efficiency, an efficiency which is not important for these tests. As
stated before, a high "Q" for the receiver inductor is essential as the
signal varies directly as the "Q".
EXTERNALLY POWERED AND EXTERNALLY TRIGGERED TAGS
To externally power or trigger a tag requires adequate energy
to be transmitted to the tag receiver to either be rectified and hence
provide power to the tag, Figure 48; or to be applied to a sensitive
93
-------
10
35 khz
78 inches
2.3 MILLIHENRY
Q-l INDIANA GENERAL
CORE MATERIAL
LENGTH 1 INCH
220 TURNS 25/44 LITZ
20
40 60 80
TRANSMITTER CURRENT, mi Hi amperes
Figure 45. Curves showing the relative received energies for a varying source
energy for Q-l and H type core materials.
-------
1.0,—
VO
en
0.8
0.6
CIS
i—i
co
o
LU
CJ
0.4
0.2
RECEIVER COIL
Ql CORE L=l" D= .125
400T 7/41 LITZ
f= 35khz
(XMTR COIL = 10.2 mh)
L = 7" D= .5"
INPUT I = 57ma.nms
10
DISTANCE, feet
12
14
Figure 46. Curve showing the relative received energy for constant
source energy.
-------
1/8 inch diameter core Q-l material
150 turns 25/44 litz
f 35 khz
BRIDGE
RECTIFIERS
X
D. C. OUT TO
POWER TAG
Figure 47. Schematic showing the configuration of a
power pickup and converter for use as a
power source for an ultrasonic tag.
96
-------
LIB
MATERIAL, "T5KOID
uc
AT
-f
Figure 48. Externally triggered, battery operated
ultrasonic tag.
97
-------
switch which would turn-on a battery powered tag, see Figure 48, and
thereby save battery life. Each of these choices required an evaluation
of the useable power one could transmit to the tag and over what kind
of distances this was practical. To externally trigger a battery operated
on, 100 millivolts rms had to be available at the output of the tag receiver
antenna and power (standby) for the switch activation is all that is required
during off times. Since bridge rectifiers are used, the forward drop of the
diodes is lost and this contributes to the inefficiency of an externally
powered tag. In addition, the signal being pulled off the receiving inductor
must be developed across a resistance and this spoils the circuit "Q";
therefore, the current that can be drawn from the antenna is limited. The
load of the rectifiers and filter can be approximated by a 1000 ohm load.
Any configuration can be evaluated for sufficient voltage by this technique.
Germamium diodes, if available in miniature form, have a forward drop of
0.3 volts. For the two used in the bridge circuit, 0.6 V is lost. Earlier
tests at less than 100 ma drive on the transmitting antenna indicated that
higher drive currents would be required to reach the necessary voltage.
Initial estimates were approximately 500 ma.
Preliminary experiments with higher power were discouraging. With
the early transmitting inductors which were with litz wire, losses became
excessive above 100 ma. rms. Increasing the drive power provided little
increase in inductor current. The effect seemed to be due to the limited
current handling ability of the small litz wire. When another inductor
was wound with #21 enamel wire, the'current limit effect was not observed
and the inductor coil could be driven to the limit of the power amplifier.
The test set-up is illustrated in Figure 49. The operating frequency
is determined by the signal genertor, an Hp 3310B. This signal is fed
to a power amplifier, an Hp 467A,which in turn drives the inductor. The
-------
vo
VO
HP 331 OE>
35 Kite.
A A
y u
^ X
HP %7A
A fi (.-P
u 1 1 y\ \
\
SIGNAL.
GENERATOR.
FbWER.
AMPLIFIED
Figure 49. Test equipment configuration for the evaluation of high current
inductive field sources and receivers.
-------
inductor is series resonated at the generator frequency with a variable
capacitor. A current probe, Hp 456A, was connected around one of the
leads from the power amplifier to the inductor. The inductor consisted
of 80 turns of #21 Formwar insulated wire on a 1.27 cm (0.5 in.) diameter
by 17.7 cm (7 in.) long Indiana General Q-l ferrfte rod. The inductor
and capacitor were mounted on a non-metallic stand which allowed a variety
of positions and angular adjustments.
The receiver inductor was wound with a 25/44 litz wire on a
3.2 mm (1/8 in.) diameter ferrite rod of Indiana General Q-l ferrite rod
2.54 cm (1 in.) long. It was parallel resonant with a capacitor at the
transmitter frequency. The voltage across the resonant circuit was carried
by coax to a switch which allowed the drive current and received signal
to alternately be measured with a Hp 3400A rms voltmeter.
The received signal versus drive current and distances are shown
in Figures 50 and 51, respectively. Since the ultra-sonic tag requires
^
1.2 volts for minimum operation, over 2 volts rms must be generated by
the receiver inductor. From the earlier data showing the received signal
strength, Figure 50, over 10 amps must be driven through the transmitter
inductor to effectively be operational over a distance of 2.125 M (five ft).
The ultrasonic tag used for this test was optimized for low voltage low
current operation (Figure 52). Some frequency and pulse duration instabil-
ities exist because of the variable voltages generated at the receiver
due to distance and angle variations. When the antenna is nearly orthogonal,
the signals are so low that the tag will not be activated. These are
statistically small occurrences and should not seriously ramify the applica-
tion.
ULTRASONIC RECIEVER
To receive the ultrasonic signal in a laboratory environment a
versatile receiver is required. Such a receiver was fabricated for this
100
-------
250 _
0
200 -
in
E
150 _
2 TOO
oo
Q
UJ
O
UJ
a:
50 -
T
T
500 1000 1500 2000
SOURCE DRIVE CURRENT, mi Hi amperes
2500
Figure 50. High power test data showing received signal as a
function of drive current at a distance of 1.82
meters (6 feet).
101
-------
500-^
400 _
2000 ma P-P drive
300 -
CL.
D.
-------
Figure ^. Schematic of externally powered ultrasonic tag.
103
-------
program, and it is shown in Figure 53. It is similar to the receiver
of Figure 16. The input is a PZT microphone. The received signal is
amplified and hetrodyned down to video for audio detection. No attempt
was made to filter out frequencies beyond the tuning range and should
they be present, they will be heard as clicks in the loudspeaker.
Basically, it is a tuned frequency receiver and uses one integrated
circuit amplifier which serves also as the mixer for audio conversion
Figure 53. Photograph of ultrasonic receiver
for reception of 35 kHz signals.
104
-------
APPENDIX C
FRY DETECTION SYSTEM
GENERAL
The investigation of thermal pollution effects at the Monticello
Ecological Research Station requires a system to detect and count free-
swimming fish fry entering and leaving a test channel. This condition
must be monitored if meaningful results offish production are to be
measured. Consultation with fish hatcheries has provided a better
understanding of the habits of fish fry, their size, quantity, physical
characteristics and general behavior during the early stages of their
life.
Basically, one is concerned with a fish spawning several thousand
eggs in the river bottom. Initially, the survival rate is 60% to 70%
and size varies from 4 to 12 mm long and 1 to 2 mm in diameter. In the
early stages the young fish will be clumped together on the river bottom.
As they approach a 100 mm size they start to school, covering an area of
approximately one cubic meter. The schools range further afield as they
grow and generally follow the flow of the current. Detection of single
small fry would be exceedingly difficult but in schools the fry are detect-
able.
The physical layout of the Monticello test station provides a
restricted area at both ends of each channel. River water is introduced
at the head of each channel through a well 30.5 cm (1 ft) x 46 cm (18 in.)
and at the lower end through a spillway with a V-notch weir gate (Figure 54)
This means that a school of fry entering or leaving the test channel would
be forced through a confined area providing an ideal point to establish
a fry monitoring station.
105
-------
Test channel entrance.
liiiii
III
111
Exit gate.
Figure 54. Instrumentation sites for
fry detection system.
106
-------
BACKGROUND
Much effort has been expended over the last 30 years in attempts
to locate, track, classify and count fish. A literature search was
conducted to determine what methods had been tried and the results of
research in the fish counting area. The following areas were found to
be worthy of further development for our specific problem:
a. Magnetic field.
b. Impedance technique.
c. Electrostatic method.
d. Equilibrium potential electrodes.
e. Sonar pulsing.
The scope of investigation was limited to a detection method
without consideration of target classification or data acquisition and
processing.
INVESTIGATION OF FISH FRY DETECTION SYSTEMS
Magnetic Field Techniques
The detection of fry as a result of perturbation of an establish
magnetic field was investigated using a series of flat wound coils
bottom-mounted in the channel. The electrical Q of the coils could
not be maintained in water and the system was insensitive to fry activity.
Impedance Techniques
The impedance of a section of the channel can be measured as a
function of frequency. This impedance, R + JX, as a function of frequency
would take into consideration the type of fish, its size, and quantity.
After sufficient empirical data was obtained, extraneous objects,
pollutants, etc., could be isolated due to their different composition.
To increase the sensitivity of the method, the impedance would have to
107
-------
be normalized for varying temperatures. This will offer no problem since
the temperature will be readily available for each channel section.
Because this technique is parallel to the electrostatic method most
effort was devoted to the capacitive approaches.
Electrostatic
The capacitance of a section of the channel can also be measured
as a function of frequency. This method is similar to the impedance
technique except it can be implemented more easily. An oscillator, using
the capacity of the channel section as one element of the frequency deter-
mining LC network was investigated. A typical impedance or capacitance
plate configuration to be used for either the capacitance or impedance
techniques would in its simplest form be a 2.5 cm (1 in.) pipe laid on
both sides of the channel and one on the channel bottom. The pipe would
be composed of 3.05 m (10 ft) sections connected by a flexible insulated
coupling. Inside the pipes each section would have an isolating electronic
circuit to permit making the measurement at each section and allowing the
data to be connected by relatively long wires to the on shore electronics
and control center without these wires introducing stray impedances.
The use of pipe on the side and bottom of the channel allow more
accurate measurements to be made on a school of small fishes located in
the small depth areas near the shore.
Obviously, other impedance plate configurations are possible.
Other plate configurations could be fiber glass mesh with silver painted
surface, conventional wire mesh fence material, etc.. In all cases the
metal surfaces must be covered to insulate the plates from the water.
Two types of plates were fabricated for this series of experiments.
Aluminum plates 30.48 x 30.48 cm and type G10 copper-clad glass filled
108
-------
epoxy circuit board 15.24 cm x 30.48 cm were used. It was necessary to
completely insulate the plates from the water and the G10 material with
the copper clad sandwiched between glass insulators proved easiest to
seal. The G10 boards were mounted on four wooden dowels separating the
plates by 15.24 cm.
Direct current (dc) insulation resistance was checked with the plates
in position in the filled fish tank and resistance was greater than 1 megohm.
A 1.5 m (5 ft ) long pair of test leads, separated by at least 15.24 measured
7 picofarads capacity. The complete system capacity with plates in air
was 10 picofarads and the same system when placed in water exhibited a
capacity of 1200 picofarads.
Because of the small variation in capacity expected as a result of
fish located in the dielectric field, a second set of capacitor plates
were constructed and placed in the water. The use of two sets of plates
allows us to use a bridge circuit. The second set of plates are isolated
from the fish and form the reference leg of the bridge. The capacity
change caused by the fish is the differential capacity (AC) of the bridge
circuit.
The passage of a 3.8 cm (1.5 in.) goldfish through the test area
produces changes in capacity of 0.25 to 3.2 picofarad depending on the
aspect of the fish during transit. Maximum capacity change occurs when
the fishes length is perpendicular to the capacitor plates. Testing
neo-tetras in place of goldfish under the same conditions produced
variations of 2.5 to 3.0 picofarads capacity.
This system proved to be insensitive and adjustment very critical.
Stray capacity due to lead dressing and shifting caused considerable
problems in reproducing test results.
109
-------
Equilibrium Potential Method
The equilibrium potential approach to fry detection was investigated
following the outline of the paper presented by Spoor and Drummond of
ERL-Duluth. Only one set of electrodes were used and one channel was
recorded.
After the potentials in the test tank had stabilized and in the
absence of electrical interference, which would affect the normal chemical
equilibrium between electrodes, significant readings of repeatable data
were recorded, using goldfish, as shown in Figure 55. Swimming movements
could be identified, but no indication of breathing activity was observed.
We seemed to be limited in ambient level due to AC power line pickup.
It was necessary for the fish to pass very close to the electrode
for detection (2cm or less). Any activity in the tank other than fish
movements produced responses similar to fish movement. It appears that
a fairly elaborate electrode grid would be required to effectively monitor
a small segment of the channel and would represent a definite impediment
to the flow of large fish and debris.
Sonar Method
Since the simpler systems originally proposed did appear feasible
from preliminary testing a sonar system concept was set up and tested at
the Monti cello facility. This is probably the most often taken approach.
Sonar systems for fish location have been built over the frequency range
of 10 kHz to 600 kHz. They have been used to: Count salmon going upstream,
measuring the biomass (kilograms) of fish in rivers, lakes, and portions of
the ocean, etc. with some success. However, the transmission of an acous-
tic signal in a small area (such as the narrow channel of the Monti cello E-
cological Research Station) presents many problems. This is due to such
factors as multiple reverberations, side lobes of the transducers,
110
-------
END OF TEST
RUN
SMALL GOLDFISH PASSED
CLOSE TO ELECTRODE
ATTEMPTED TO MOVE
FISH INTO TEST ZONE
SMALL GOLDFISH PASSED CLOSE
TO POTENTIAL PROBE WIRE
1 GOLDFISH PLACED
IN TEST AREA
1 MARCH 1973
START OF
TEST RUN
RECORDER:
BECKMAN RP
Figure 55. Sample test data, equilibrium
potential system.
Ill
-------
unpredictable variation in reflection from the fish as a function of
its position (can vary as much as 30 dB), reflection from the water
surface and irregular contours, the unpredictable variation in signal
strength returned from fish of different sizes, and doppler shifts due to
rapidly moving fish. Air in the water and plant life also add to the
problem. Many techniques have been proposed or attempted to obviate
these disadvantages, such factors as swim bladder, vertebral column
and body tissue resonances can be used to isolate fish acoustically,
especially small fish, from others of different sizes.
The sonar system tested at Monti cello is shown in block diagram
in Figure 56. A tone burst pulse generator provides the signal to the
transmitting transducer at one side of the channel. The signal travels
the acoustic path to two receiving transducers separated 15.24 cm on the
receiving side of the channel, where the two receiver outputs are compared.
A variation in either path is an indication of an abnormality present.
The direction of movement can be determined from phase information.
Target classification is possible by examination of the received waveform
and period of perturbation.
DEMONSTRATION OF FRY SYSTEM AT MONTICELLO.
The fry detector system test program was conducted to demonstrate
the feasibility of detecting small fry as they enter or leave the test
channels at the Monticello Ecological Research Station, using an ultrasonic
detector system.
The system will detect fry, and in addition, grass clippings,
clumps of algae, and air bubbles. The system will not count the number
of fry, but will indicate the time period fry are in the sonic beam
allowing an estimate of the number of fry in the school. (There is
112
-------
co
V &ATE AT
LOWER E.ND
TEST CHAMKIEL
Figure 56. Acoi
VERT.
OSCILLATORS
FULSE GENERATOR.
INTERSTATE
HOD.
90
stic fry detection system block diagram.
-------
error here unless it is determined that the system is indeed looking
at fry).
The detector transducers were placed in the outlet at the end of
Channel No. 1 which is 92 cm (3 ft) wide with a water depth of 76 cm
(30 in.). The sound field was established 46 cm (18 in.) upstream from
the "V" slot weir and 38 cm (15 in.) below the water surface.
There is a source transducer on one side of the channel and two
receiving transducers on the opposite side. The signal received is
applied to the plates of an oscilloscope and then differences in the
scope pattern indicate one interruption of the sound field. Two receivers
in a balanced configuration are the most sensitive to changes in the
field.
A group of approximately 25 fry (identified as red horse, Moxostoma spj
20 to 40 mm in length were swimming in the current before the "V" notch
weir in the end of the channel and would drift back and forth through the
sound field. These transients were easily detectable in the oscilloscope
display. There was considerable floating algae in the area and flowing
through the sound field. Large clumps of algae extending a foot or more
below the surface, disturbed the scope display in the same manner as the
fry. Algae at the surface did not interfere with the system. Agitation
of the surface of the water did not interfere with data displays.
The system was operated at frequencies from 40 kHz to 400 kHz and
it was found that most satisfactory results were obtained at resonance of
the transducers - 90 kHz. There was no indication of air bladder resonance
or improved target acquisition as a function of frequency.
No attempt was made during this demonstration to optimize test zone
acoustic path length, receiver spacing or acoustic window placement in
relation to the bottom or distance from the "V" notch.
114
-------
The tests demonstrated a viable sonic fry detection method,
although further work would be required to properly implement a final
system for enumerating fish fry without interference.
115
-------
APPENDIX D
SUPPLEMENTARY MATERIALS
Oversize blueprints and other specific information of limited
interest are available only from the project officer. Those materials
are:
(1) Engineering blueprints
—System block diagram
—Receiver/temperature decoder
—Acoustic temperature transmitter assembly
— Interconnecting box wiring diagram
—Hydrophone assembly
(2) Parts lists for some of the five preceding components
(3) Instruction and Maintenance Manual for Large Fish Tracking
System, 27 p.
Readers who want to review these materials may borrow them by writing
to Ken Hokanson, Ph.D., Chief, Monticello Ecological Research Station,
P.O. Box 500, Monti cello MN 55362.
116
-------
GLOSSARY OF ABBREVIATIONS
AGC Automatic Gain Control
cm Centimeters
cw Constant Wave
dB Decibels
dc Direct Current
emf Electromagnetic Field
1C Integrated Circuits
ID Inside Diameter
kHz Kilohertz
mm Millimeter
OD Outside Diameter
pm Pulse Modulation
pps Pulses Per Second
PZT Lead Zirconate Titanate
Mbar Microbar
rf Radio Frequency
Shore A Scale of Relative Hardness
Shore D Scale of Relative Hardness
VCO Voltage Controlled Oscillator
117
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse I if ore completing)
1. REPORT NO.
EPA-600/3-77-035
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Instrumentation to Monitor Location of
Continuously in Experimental Channels
Fish
S. REPORT DATE
April 1977
issuing date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
Joseph R. Jahoda
9. PERFORMING ORGANIZA1 ION NAME AND ADDRESS
Bayshore Systems Corporation
5404 B Port Royal Road
Springfield VA 22151
10. PROGRAM ELEMENT NO.
1BA608
11. CONTRACT/GRANT NO.
Contract 68-01-0752
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Research Laboratory-Duluth,
Office of Research and Development
U.S. Environmental Protection Agency
Duluth, Minnesota 55804
13. TYPE OF REPORT AND PERIOD COVERED
Final. Completed 10/74
14. SPONSORING AGENCY CODE
EPA/600/03
15. SUPPLEMENTARY NOTES Q
isfartorily under fie
cost1*'. Supplementary materials,
., the large fish tracking unit was delivered ana performed sat-
d conditions. However, the technology was not continued as too
ints and Maintenanct
Tf are aval Table
including blueprints and~Maintenance and Instruction
Manual, are aval Table from the Pro.iect
This study resulted in the development and construction of equipment to Officer.
continuously monitor the position and temperature of up to 20 fish in a water channel
486 meters long, 3 meters wide, and 1 meter deep. The system utilized miniature sonic
transmitters (tags) operating in the 51 kHz to 366 kHz frequency range which were
implanted in 500 gram or heavier fish. The battery operated tags were pulse modulated
and designed for over 1 year operational life. A temperature sensitive thermistor con-
trolled "the repetition rate of the tag providing tne temperature of the fish to an ac-
curacy of 1 degree C. The nominal range of the polyurethane encapsulated tag was sev-
eral hundred feet. Nominal tag size was 16 mm OD x 32 mm long (4.6 - 5.4 g in water).
Sixteen hydrophones were located at 30.5 meter intervals in the water channel, A con-
trol console contained a manually-operated, frequency-stepped receiver which could
select any individual hydrophone, thus locating the fish to within +_ 15.25 meters. Up
to 20 individual fish could be monitored. Automatic operation and recording of the
data was considered in the design of the system for future equipment. Severe radio
frequency interference problems were encountered, requiring extensive precautions and
modification of the channel equipment and wiring. Also investigated were passive fish
monitoring and tracking of small fish fry. An experimental system was completed for
limited monitoring applications.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
*Biotelemetry, *Temperature measuring in-
struments, *Acoustic receivers, *Monitors,
*Electromagnetic noise, *Ultrasonic fre-
quencies, *Pulse modulation, *Miniaturiza-
tion, ^Hydrophones, *Preamp1ifiers", *$onar
*Inductive reactance, *Impedance :
fish location, control
console, polyurethane
encapsulated, mercury
batteries, .temperature
decoder, Hartley oscilla-
tor.,...complementary
•astable multivibrator
06B
09F
8. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)'
UNCLASSIFIED
21. NO. OF PAGES
128
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
118
*
------- |