ISSN 1055- 1425
November 2008
This work was performed as part of the California PATH Program of the
University of California, in cooperation with the State of California Business,
Transportation, and Housing Agency, Department of Transportation, and the
United States Department of Transportation, Federal Highway Administration.
The contents of this report reflect the views of the authors who are responsible
for the facts and the accuracy of the data presented herein. The contents do not
necessarily reflect the official views or policies of the State of California. This
report does not constitute a standard, specification, or regulation.
Final Report for Task Order 5606
CALIFORNIA PATH PROGRAM
INSTITUTE OF TRANSPORTATION STUDIES
UNIVERSITY OF CALIFORNIA, BERKELEY
Development and Field Testing of Laser
Photodiode Array- Based Vehicle Detection
System
UCB- ITS- PRR- 2008- 28
California PATH Research Report
Harry H. Cheng, Ben Shaw, Joe Palen, Hong Duan,
Stephen S. Netinger, Bo Chen, Ping Feng
CALIFORNIA PARTNERS FOR ADVANCED TRANSIT AND HIGHWAYS
Development and Field Testing of
Laser Photodiode Array- Based
Vehicle Detection Systems
Harry H. Cheng
Ben Shaw
Joe Palen
Hong Duan
Stephen S. Nestinger
Bo Chen
Integration Engineering Laboratory
Department of Mechanical and Aeronautical Engineering
University of California, Davis
Davis, CA 95616
1/ 72
Contents
Abstract ...................................................................................................................... 4
Executive Summary ................................................................................................... 5
1 Overview of the Laser- photodiode Based Detection System ( LBDS) ............. 7
1.1 Architecture of the LBDS................................................................................. 9
1.2 Opto- Mechanical System................................................................................ 10
1.2.1 New Structure of the Optics System...................................................................... 10
1.2.2 New Laser Module.................................................................................................. 11
1.2.3 Using Laser Pointer ................................................................................................ 12
2.3 Redesign of the Electronics System .................................................................. 13
2.3.1 Microcontroller Based Control System.................................................................... 13
2.3.2 Laser Diode Module and the Trigger Circuit.......................................................... 14
2.3.3 APD array and its bias voltage supply..................................................................... 15
2.3.4 Signal Pre- Processing Circuitry ............................................................................... 17
2.3.4.1 Three- stage Pre- amplifier .................................................................................... 18
2.3.4.2 Comparator and Monostable Multivibrator ...................................................... 18
2.3.4.3 Hysteresis to Overcome Noise and Signal Amplitude Fluctuation ................... 19
2.3.4.4 Digitally Controlled Potentiometer ..................................................................... 19
2.3.4.5 Timing Window..................................................................................................... 20
2.3.4.6 Pre- board Noise Data Analsys and Data Process............................................... 20
2.3.4.7 Arrangement of Printed Circuit Boards in Signal Box...................................... 22
2.3.5 Power Supply Circuit and DC/ DC Converter......................................................... 22
2.3.6 Temperature Compensation for Responsivity of APD........................................... 23
2.3.7 Data Acquisition and Output.................................................................................... 25
2.3.7.1 Rabbit Data Acquisition ....................................................................................... 26
2.3.7.2 Output Data Through Ethernet Communication .............................................. 26
2.3.7.3 Output Data Through Wireless Ethernet ........................................................... 27
2.3.7.4 Parallel Port Output ............................................................................................. 27
2.3.8 Optimal Self- Calibration........................................................................................... 28
2.3.9 Remote Control for Trolley through Wireless Ethernet ........................................ 32
2.3.10 Software Design for Rabbit Microprocessor ........................................................... 33
2.4 New Design of the Mechanicle System ............................................................. 34
2.4.1 Breadboard for opticle system.................................................................................. 34
2.4.2 Acrylic Box for Whole LBDS.................................................................................... 35
2.5 Software on remote computer........................................................................... 35
3 Comparison Study with the Previous System.................................................... 36
3.1 Longer length of the Detection Area ................................................................ 36
3.2 24 channels.......................................................................................................... 36
3.3 Stronger Anti- noise Ability ............................................................................... 36
3.4 Flexibility for Different Environment.............................................................. 36
3.5 Communication.................................................................................................. 36
3.6 laser Alignment .................................................................................................. 36
4 Testing ................................................................................................................ 37
4.1 Indoor Lab Testing ............................................................................................ 37
2/ 72
4.2 Roof Testing........................................................................................................ 38
4.3 Mobile Truss Testing ......................................................................................... 39
4.4 Outdoor Highway Testing................................................................................. 41
5 Conclusions........................................................................................................ 43
References................................................................................................................. 43
Appendix A: Circuit and Board Diagrams .............................................................. 45
Appendix B: Source Code ........................................................................................ 49
Source Code in rabbit 3200 for control and communication....................................... 49
Source Code of TCP/ IP communication in Linux ........................................................ 69
3/ 72
Abstract
Over the past year we have enhanced the performance of the Laser- photodiode Based vehicle
Detection System ( LBDS) which is sponsored and funded by the California Department of
Transportation ( Caltrans) through the PATH Program. This project incorporates the development a
vehicle detection system for highway use that can implement correct, reliable, and accurate vehicle
detection and has the ability to provide high quality traffic information such as traffic density and
vehicle speed, length and profile to the traffic management center for surveillance. After several years
of research, the performance of the LBDS has been enhanced and is almost ready for use in practical
applications. In the last year, the focus of our research has been on the reliability, accuracy and anti-disturbance
capabilities of the LBDS. By using new laser diode modules, the laser line projected on
the road surface has been enlarged to 4 m, which almost covers the full width of a lane on the
highway, and now have a uniform intensity distribution along the line increasing accuracy at the
edges of the detection zone. A new optical system has been designed and manufactured that allows
the avalanche photodiode ( APD) array in the LBDS to “ view” the full width of a lane on the highway.
Accordingly, the electronic systems have been redesigned. The number of detection channels has
been increased from the previous 8 channels to the present 24 channels, increasing the detection
resolution. Use of a hysteresis comparator and a timing window has reduced the effect of backdrop
noise, enhancing the reliability and stability of the LBDS. Temperature compensation for the
responsivity of the APD has been implemented which stabilizes the amplitude of analog signal from
the APD. Temperature feedback improves system performance over a wide range of temperatures,
which afflicted previous designs. The optimal self- calibration capability of the LBDS makes the
system more flexible to different operating environments. The preprocessing circuit board has been
redesigned based on the analysis of the effect of sunlight and oscillating noise on the system. A high-filter
circuit was used to increases the S/ N ratio which allows for stable long- term use. Outputting the
detection data through wireless Ethernet communication makes the LBDS mobile and remote
accessible. With the above improvements on its performance, the LBDS has is now more suitable for
the practical application of vehicle detection on the highway. This report provides a detailed
description of the design and implementation of the enhanced LBDS. The field- test results of the
LBDS prototype at different test sites, including Old Hutchison testing at UC Davis and highway
testing in Sacramento, are documented in the report.
Keywords: highway vehicle detection system, laser detector,
photodiode arrays
4/ 72
Executive Summary
Highway traffic information gathered such as the number of vehicles passing through a detection
zone, vehicle speeds, and vehicle profiles used for vehicle classification, are fundamental to advanced
traffic management systems, e. g. Intelligent Transportation System ( ITS) [ 1]- [ 2]. Therefore, the
authenticity and accuracy of the information has an adverse affect on the behavior of the ITS. High
quality information relies on the performance of field surveillance devices on the highway, i. e.
vehicle detectors. Traditionally, highway vehicle detection systems were based around inductive loop
detectors [ 3]- [ 4]. Although they are still widely used, loop detectors can hardly provide high quality
information required by an ITS [ 5]. The Laser- photodiode Based vehicle Detection System ( LBDS) is
a new kind of advanced vehicle detector based on different working principles from the loop detector
[ 6]- [ 10], which has ability to provide more traffic information with much higher quality compared to
loop detectors. A previous publication on the LBDS discussed [ 10] the salient features provided by
the LBDS in comparison to other available detector systems.
Over the past year we have enhanced the performance of the Laser- photodiode Based vehicle
Detection System ( LBDS) which is sponsored and funded by the California Department of
Transportation ( Caltrans) through the PATH Program. This project incorporates the development a
vehicle detection system for highway use that can implement correct, reliable, and accurate vehicle
detection and has the ability to provide high quality traffic information such as traffic density and
vehicle speed, length and profile to the traffic management center for surveillance. After several years
of research, the performance of the LBDS has been enhanced and is almost ready for use in practical
applications. In the last year, the focus of our research has been on the reliability, accuracy and anti-disturbance
capabilities of the LBDS. By using new laser diode modules, the laser line projected on
the road surface has been enlarged to 4 m, which almost covers the full width of a lane on the
highway, and now have a uniform intensity distribution along the line increasing accuracy at the
edges of the detection zone. A new optical system has been designed and manufactured that allows
the avalanche photodiode ( APD) array in the LBDS to “ view” the full width of a lane on the highway.
Accordingly, the electronic systems have been redesigned. The number of detection channels has
been increased from the previous 8 channels to the present 24 channels, increasing the detection
resolution. Use of a hysteresis comparator and a timing window has reduced the effect of backdrop
noise, enhancing the reliability and stability of the LBDS. Temperature compensation for the
responsivity of the APD has been implemented which stabilizes the amplitude of analog signal from
the APD. Temperature feedback improves system performance over a wide range of temperatures,
which afflicted previous designs. The optimal self- calibration capability of the LBDS makes the
system more flexible to different operating environments. The preprocessing circuit board has been
redesigned based on the analysis of the effect of sunlight and oscillating noise on the system. A high-filter
circuit was used to increases the S/ N ratio which allows for stable long- term use. Outputting the
detection data through wireless Ethernet communication makes the LBDS mobile and remote
accessible. With the above improvements on its performance, the LBDS has is now more suitable for
the practical application of vehicle detection on the highway. A prototype of the newly designed
LBDS is shown in Figure 0- 1.
5/ 72
Figure 0- 1. Top view of the 24- Channel prototype of the LBDS.
6/ 72
1 Overview of the Laser- photodiode Based Detection System
( LBDS)
The LBDS is an overhead vehicle detector that provides the speed, length, width and profile of
vehicles on the Highway. A diagram of the working principles behind the LBDS is shown in Figure
1- 1. The LBDS contains two sets of laser diodes, avalanche photodiode ( APD) arrays and optical
systems installed in a crosswise fashion. The laser diodes, which are triggered at 10 KHz, project
pulsed laser beams onto the road surface. The APD arrays act as laser- sensitive sensors and are used
to detect the reflected laser light from the road surface. The optical systems focus the reflected laser
light, which creates an image of the projected laser line onto a APD array. When no vehicle is present
beneath the LBDS, laser light from the laser diode on left side of the LBDS is reflected from road
surface and focused onto the APD array on right side of the LBDS through the right side optical
system. Similarly, laser beam from right side will be focused onto the left side APD array. Once
activated by the reflected laser beam, the APD sensors will output an analog signal to represent the
absence of vehicle. Therefore, when the APD arrays receive a laser pulse, the projected laser line is
determined to be “ unblocked”, i. e. no car is present. When a vehicle is passing under the LBDS, as
shown in Figure 1- 1, the reflected laser light be “ blocked” by the vehicle, i. e. no light will reach the
APD arrays. Therefore, the APD arrays will not output a signal, which represents the presence of a
vehicle. By monitoring the APD output every clock cycle, the presence of a vehicle in the detection
zone can be determined along with the times associated with the vehicle entering and leaving the
zone. The detection data is transmitted to a remote computer that displays or records the desired
traffic information. For example, when a vehicle passes through the detection zone, the front of the
vehicle first blocks laser beam 1 at time t1 and then blocks laser beam 2 at time t2. Then, its rear will
unblock laser beam 1at t3 and laser beam 2 at t4. Based on these time stamps, the front and rear
velocities of the vehicle as well as the acceleration and length of the vehicle can be calculated [ 8].
7/ 72
Figure 1- 1. Working principles of the Laser- photodiode Based Detection System.
The two linear laser beams that are used as laser sources are shown in Figure 1- 2 ( a), which were
imaged indoors through the use of a laser viewer. Figure 1- 2 ( b) shows the two laser beams projected
across a lane of a highway when a vehicle is at the edge entering the field of view of the first
laser/ senor pair. The laser beam is reflected from the ground and is focused onto a 7.5 mm linear APD
array through an optical system. The APD is an array of 25 photodiode elements, of which 24 are
used. There are 24 signals for Laser1 and another 24 signals for Laser2. Each pair of element signals
of Laser1 and Laser2 will turn out a set of vehicle parameters, giving a total of eight sets of vehicle
parameters. The purpose of acquiring multiple vehicle parameters is to improve the reliability of the
parameters and to acquire the outline profile of a vehicle.
8/ 72
( a) ( b)
Figure 1- 2. Laser- based detection system overview.
When the Rabbit3200 reads the signal data, it packs data into TCP Packages and sends them to a
remote computer through an Ethernet port. Using a wireless router, the LBDS is a fully un- tethered
system allowing mobile systems to connect wirelessly to it in order to obtain the desired data. The
vehicle parameters are calculated and displayed in a remote computer.
1.1 Architecture of the LBDS
The architecture of the LBDS is shown in Figure 1- 3. It is composed of an opto- mechanical
subsystem for laser- beam collection and adjustment, signal amplification and shaping circuitry to pick
up the laser signals for conversion to digital signals, a Rabbit 3200 system for data acquisition,
processing and data communication broker, and a computer for parameter calculation and
record/ replay/ reuse of data.
The opto- mechanical laser receiver collects the laser signal reflected from the ground into the APD
linear array. The APD linear array generates signals when the laser beam is present. The signals are
then amplified by the signal amplification and shaping circuitry. The opto- mechanical laser receiver
has a very narrow field of view. When the system is moved to different heights, the laser sources must
be adjusted so that the projected linear laser beam stays within the systems field of view. The opto-mechanical
system is very sensitive to the signal output from the APD that is mounted to it. Our
system should be adjusted accordingly relative to the height of the detector from the ground.
The signal amplification and shaping circuitry is used to pick up the signals generated from the APD
array, and then amplify and shape them. It outputs digital signals that are read by the Rabbit 3200
core module. Considering that 48 signals will be read in the future instead of the 16 signals that are
currently read, we use bus technology to connect the Rabbit core module to the signal amplification
and shaping circuitry.
The Rabbit acts as a PWM generator to trigger the laser diode modules at 10 kHz. The laser diode
module produces a linear laser beam with a full fan angle of 15 degrees that is projected towards the
ground. Some of the reflected laser beam from the ground will fall onto the front of the receiving lens
of the opt- mechanical system, which then falls onto the APD array causing it to output a pulse signal.
The Rabbit core is triggered at the same frequency as the laser diodes to read the digital signals
coming from the signal amplification and shaping circuitry. Every 10ms, the Rabbit module sends one
TCP package containing 100 sets of data to a computer over a network.
9/ 72
The computer calculates the parameters of vehicles as they pass underneath the detector, and then
displays, records, or replays the results.
Figure 1- 3 The architecture of LBDS
1.2 Opto- Mechanical System
The optical system of the LBDS is one of the most important subsystems since it directly and
adversely affects the overall performance of the LBDS.
1.2.1 New Structure of the Optics System
A new optical system has been designed in conjunction with JML Optical Industries, inc. The initial
prototype is shown in Figure Figure 1- 4.
10/ 72
Figure 1- 4 The prototype of the new optical system.
Figure 1- 5 Assembled optics- sensor pair.
1.2.2 New Laser Module
The laser module we used in the previous design could only provide a laser line 1.5m in length at the
road surface 8m far away from the module. This is not long enough to cover the lane width of 3.8m of
highway. Two laser modules are therefore used on one side to combine the two laser lines into one
line, hoping to solve the problem. But the combined laser line is only 2.5m long since an overlapping
area is needed for the combination. So the problem still existed. In addition, it is difficult to position
the two laser line in one straight line.
11/ 72
With the effort on the development of a new laser module that meets our requirements, a laser module
producing company developed a laser module for our application. The laser module has 60W peak
power with the pulse frequency up to 10kHz. The length of the laser line on road surface 8m away
from the module can be longer than 4m, and the intensity distribution along the laser line is close to
uniform. The performance of the new laser module is helpful to the correct vehicle detection. Figure
1- 6 shows the comparison between the new laser line and the old laser line. The upper line is the new
laser, and the lower line is the old laser which is combined by two laser lines.
Figure 1- 6. Laser line
The second batch sample laser modules from Power Technology Inc. However, they exhibit
overheating issues cannot be used for extended LBDS testing. The new laser modules operate at 24
volts as compared to the old models at 12 volts. A new power supply circuit was designed and
fabricated to provide the necessary 24 volts. Figure 1- 7 shows the laser module profile.
In order to do a comparison of the new laser modules with the older ones, outdoor testing, roof testing
and highway testing were conducted. New laser module mounts were fabricated for the tests. Testing
results showed that the new lasers are brighter benefiting signal acquisition due to an increase in
signal strength. However, electrical noise from the powerful drivers of the new laser modules have a
strong effect on the front- end circuits and result in increased noise compared to the previous setup.
Opto- electrical coupling components were added to take steps in avoiding the noise from new laser
modules. The design has illustrated in the next section.
Figure 1- 7 Laser module
1.2.3 Using Laser Pointers
The laser light we are now using in LBDS is invisible infrared laser. It is difficult to align the laser
line to the correct position on roadway. To solve this problem, we use a pair of green laser pointers to
help the alignment of the infrared laser. The laser pointers are mounted on the laser mounts for
housing the infrared laser module. One laser pointer is used on one side. The laser pointer points the
same direction as the infrared laser on the same mount points, and the spot of the laser pointer will be
in the center of the infrared laser line on the road. In this way, the infrared laser can be aligned easily
to the required detection area for correct vehicle detection.
12/ 72
1.3 Redesign of the Electronics System
1.3.1 Microcontroller Based Control System
Based on the working principle of LBDS, the main tasks that the electronics system should implement
include follows: triggering a pair of laser diodes in a determined high frequency; preprocessing the
analog pulsed signals output from a pair of APDs and converting them to a digital signal that can
represent the absence/ presence of vehicle clearly; outputting the raw data to a remote computer which
is tens meters away from the LBDS for calculating the traffic information. Moreover, some special
technologies are applied to the electronics system to let it have high reliability and stability and strong
anti- noise ability for actual application. The block diagram of the system is shown in Figure 1- 8.
Figure 1- 8. Block diagram of the electronics system.
The control core of the electronics system is a microcontroller “ RabbitCore RCM3200 Module”. It
generates a timing clock signal to synchronize the operation of all other parts of the system. The clock
signal drives a trigger circuitry to trigger the laser diodes on both sides separately and it also drives an
interrupt generator to generate an interrupt application to request the microcontroller readout all the
data from both sides of signal pre- processing circuitries. The signal pre- processings on side A and on
side B are two similar circuitries which are used to amplify the signal output from relative APD array
and convert it to binary digital data. The data buffer is used as bus- oriented receiver to get the data
from signal pre- processing circuitry on both sides and transmit the data to the microcontroller through
an 8- bit I/ O port. There are two kinds’ ports for detection data output, one of them is an Ethernet port
which is already integrated onboard in the microcontroller, and the other one is a parallel output port.
The Ethernet port is a 10/ 100Base- T Ethernet port through which the LBDS can connect the remote
computers far away via TCP/ IP communication, while the parallel output port enables the LBDS
output the data almost no delay. Both kinds of outputs are available at the same time. The digitally
controlled potentiometers ( DCP) which are used in each channel of the signal preprocessing
circuitries for optimal self- calibration are controlled by the microcontroller. The timing window is
used to reduce the effect of noise to the detection data, and the temperature compensation is designed
for high reliability of detection in a wide range of temperature.
The operation timing sequence of the electronics system is shown in Figure 1- 9. The top line is the
waveform of clock signal. Its frequency is 10.8 KHz and duty cycle is 10%, so the period between t1
and t7 is 92.5μs and the period between t1 and t3 is 9.25μs . The second line and the third line are
waveforms of trigger signals for laser diodes on side A and on side B. Since the laser diode module is
triggered at the rising edge of the trigger pulse, the laser diode on side B will be triggered at t1, and
that on side A will be triggered at t3. The advantage of triggering the laser diodes on both sides at
different time is that the crosstalk between side A and side B can be avoided easily by using timing
13/ 72
window. The fourth line is the waveform of interrupt signal which will generate interrupt application
at t5, and the fifth line is the waveform of latch signal.
Figure 1- 9. Operation timing sequence of the LBDS.
In every clock cycle, the laser diode on side B is triggered at t1 to project a narrow pulsed laser to
road. This laser will then be reflected to APD array on side A ( A2) if there is no vehicle to block the
laser. Having received the reflected laser, A2 will output a narrow and small analog pulsed signal to
the signal pre- processing circuitry on side A. This circuitry will amplify the signal and compare it
with a reference voltage to convert the amplified analog signal to a binary digital signal which will be
kept until t8. Similarly, after the laser diode on side A is triggered at t3, the APD array on side B will
output pulsed signal to the signal pre- processing circuitry on side B. The circuitry will amplify the
signal and convert it to a binary digital signal which will be kept until t10. After the binary digital
signals of both side A and side B have been ready to readout, an interrupt application to the
microcontroller will be generated at t5 by the interrupt generator. Once the microcontroller responses
the interruption, it will readout all the binary digital signals on both sides through the data buffer
before t7. Since the microcontroller has an Ethernet communication port integrated on board, the
acquired and processed data will be transmitted to the remote computer through the port. At the same
time when the interrupt application is generated, the latch signal is also generated for the parallel
output port. The output of the parallel port will be refreshed at t6 by the binary digital signals from
both signal pre- processing circuitries. By this way, the LBDS is able to transmit the detection data in
a high speed. Figure 1- 8 shows a prototype of LBDS with the high performance electronics system.
1.3.2 Laser Diode Module and the Trigger Circuit
The laser source used in LBDS is ML series of pulsed laser diode system which is an integrated
module. All components necessary for generating and projecting a laser line have been included in the
module. Only a DC power supply and a TTL timing trigger signal are needed for its operation. The
laser module will be triggered at the rising edge of the trigger signal to project a linear pulsed laser
whose wavelength is typically 905nm and the pulse width is 20ns.
The peak power of the new laser module is 60W, it is stronger than the old one which is 30 W, in the
meantime, the noise from the drive circuit is increased, and effects accuracy of the laser pulse signal
14/ 72
acquisition. The optically coupled isolators are used to avoid the interferential signal from laser drive
circuit.
The trigger circuitry is shown in Figure 1- 10. The monostable multivibrators U3A and U3B are
triggered at the rising edge and the falling edge of the clock signal respectively. When U3A or U3B is
triggered, its “ Q” pin will output a negative pulse with a pulse width of t R C p = 0.45 × × , where
is the pulse width in ns,
p t
5 6 R = R = R is the resistor in kΩ, 1 2 C = C = C is the capacitor in pF. This
negative pulse leads the transistor Q1 or Q2 output a positive pulse which will trigger the laser
module at the rising edge of the positive pulse. Since the trigger terminals of the laser diode modules
have a low input resistance and the laser modules are triggered only at the rising edge of the
triggering signal, we use power transistors Q1 and Q2 to have more power to drive the module, and
we set t s p = 8μ to reduce the power consumption of laser modules. Another advantage of using
monostable multivibrator as the triggering signal generator is that the triggering of the laser module
can be controlled. U3A and U3B can be enabled or disabled by pulling their “ CLR” pins to high level
or low level. If the monostable multivibrator is disabled, its output Q will always stay high and the
laser module will not be triggered whatever the clock signal changes. In our design, the laser
projecting can be switched on or off just by setting high or low level on “ CLR” pins of U3A and U3B
while the clock signal is still running. This function is necessary in the optimal self- calibration
procedure where the laser projecting is required to stop several times while the timing clock should be
still running.
Figure 1- 10. Laser diode modules trigger circuit.
1.3.3 APD array and its bias voltage supply
The APD array used in LBDS is a linear APD array which has 25 photodiodes located in a line.
Figure 1- 11 is its internal connection. The anodes of APDs D1 to D25 are common connected and
15/ 72
16/ 72
should be applied with a negative bias voltage ranged between - 200V to - 400V to let them work. G1
and G2 are guard ring diodes for protecting the chip. When the cathode of an APD is connected to
signal pre- processing circuitry and the reflected laser falls on the APD, a small signal current will
flow through the APD to the bias, which is the output signal of APD. The value of this signal will
change as the laser intensity on the photodiode changes, and it will also change when the bias voltage
changes. The stronger laser light on APD and/ or the more negative bias voltage will lead to a bigger
signal.
D1 D2 D3 D4 D22 D23 D24 D25
G1 G2
bias voltage
1 2 3 4 22 23 24 25
GND
Figure 1- 11. Internal connection of APD array.
The bias voltage supply circuitry is shown in Figure 1- 12 U2 is a DC/ HVDC ( High- Voltage DC)
converter which can generate the bias voltage that we needed. When the input of U2 varies from - 5V
to - 12V, the output of U2 varies from - 150V to - 460V. U1 is an adjustable voltage regulator whose
output can be adjusted by W1, so that the output of U2 can reach an expected value. Since the optimal
value of bias voltage for different APD array is different, two potentiometers W2 and W3 are used to
adjust the bias voltage for both sides of APD arrays respectively. Two Zener diodes D2 and D3 with
the breakdown voltage of 400V are used to limit the maximum negative value of the bias voltage.
Thermistors R1 and R2 are used for temperature compensation. The R1 and R2 are connected to the
surface of APD array, when the temperature on APD changed, the R1 and R2 changed, than
LM337MP voltage will change to control the output voltage of the DC/ HVDC. R1 and R2 are
negative thermisters, the R1 and R2 will be lower when temperature of APD is higher.
The stability of bias voltage on both side A and B is very important for the signal acquisition, the
signal is depended on the bias voltage and amplifiers. We have changed the C3, C4 from 0.1uF to
10uF, so as to let the bias voltage steadily.
W1
C3
W2 W3
R4 R5
E1
E3
C1
C2
C4 C5
OUT+ 3
OUT- 4
1 IN+
2 IN-DC/
DC
Trans.
U2
EMCO_ G05
2 Vin
ADJ 1
- Vout 3
U1
LM337MP
t
R1
D2 D3
GND
- 12V input
GND
bias voltage for side A
bias voltage for side B
t
R2
R3
E2
R6 R7
D1
Figure 1- 12. Bias voltage supply with temperature compensation for APD.
1.3.4 Signal Pre- Processing Circuitry
The signal pre- processing circuitry plays an important role in the operation of LBDS. It is used to
amplify the detection signal output from APD and convert this amplified analog signal to a digital
signal. The input signal comes from an APD sensor and its output is a binary logic signal which
indicates the present/ absence of vehicle. Since the amplitude of the pulsed signal output from an APD
is only a few μA and its width is about 100 ns, the circuitry is required to have large gain and wide
bandwidth to convert the weak pulsed current signal from APD to a pulsed voltage signal with a
comparable level and with a high signal/ noise ratio. Considering that the signal from an APD has a
heavy noise background ( from sunlight and circuitry) and its amplitude fluctuates, the circuitry should
have the ability to greatly reduce the effects from noise and amplitude fluctuation. Finally, the
amplitude of the pulsed signal will vary as the environment changes, which may lead to incorrect
vehicle detection. This situation may occur because the LBDS is to be used outdoors 24 hours a day
and 7 days a week with a lifespan of several years. Therefore, the circuitry is also expected to have
the ability to automatically adjust to an optimal state when the operating environment changes. As to
the output of the circuitry, a binary logic signal is a better choice to represent the absence/ presence of
vehicles. This digital output can be easily realized by hardware to ensure high processing speed and
reduce computational requirements for determining the absence/ presence of vehicles, and it also
significantly enhance the noise rejection ability of the system.
A schematic diagram of the circuitry for one element of an APD array is shown in Figure 1- 13. The
circuitry mainly consists of a three- stage preamplifier, a comparator, a monostable multivibrator and a
digitally controlled potentiometer ( DCP). Since total 48 elements of two APD arrays, 24 elements in
each, are used for detection in LBDS, there are 48 signal pre- processing circuitries similar to Figure
1- 13 24 for one side.
Figure 1- 13. Signal preprocessing circuitry.
17/ 72
1.3.4.1 Three- stage Pre- amplifier
This pre- amplifier is used to convert the pulsed weak current signal output from an APD to a
voltage signal and also to amplify this voltage signal to a comparable level. It is required to have large
gain and a fast response speed. Therefore, this pre- amplifier is designed to be a three- stage
preamplifier using two high quality operational amplifiers to make the signal amplification higher
quality. The first stage is Q1 and the second stage is U1. Q1 is Silicon bipolar transistor
AT41486, which is 8.0Ghz high gain- bandwidth up to 6GHz low noise , amplifier TL952, which are
large scale integrated circuits with small size, And the amplifier Opa 2694. All amplifiers offer low
noise and high frequency performance over 150MHz. The current signal from an APD is amplified
and converted to a considerable voltage signal by Q1, which will be coupled to U1 for a second
amplification and then to the third U2A. The third amplifier has a potentiometer W1 in its feed back
loop to adjust its gain manually, which can compensate the non- uniformity of the APD gain between
the elements of an APD array.
In the previous design, this pre- amplifier used the combination of the AD8039 and AD8054.
Although both of them have higher frequency performance, they are in smaller integration. But there
is a lot of noise from the background, they are not suitable to use in the pulse signal amplifier.
According to the analysis of data acquisition, the stronger noise from the sunshine is on the lower
frequency from 0Hz to 5MHz, and the pulse signal that we used is on the higher frequency. The low
filter has been used in the amplifier.
1.3.4.2 Comparator and Monostable Multivibrator
The comparator U2, which is a high speed comparator AD8564, is used to convert the analog signal
to a binary logic signal by comparing the amplitude of the analog signal with a reference voltage that
forms a threshold. Using a hardware comparator to convert the analog signal to a binary logic signal
is much simpler and faster than using A/ D conversion and comparison. The non- inverting input of U2
is connected to the reference voltage through resistor R11, and its inverting input is connected to the
output of the preamplifier. If the level of the preamplifier output is higher or lower than the reference
voltage, U2 will output either a low or high level to indicate the absence or presence of a vehicle in
the detection zone. Since the output of the preamplifier is normally at a minimum level lower than the
threshold, only the peak of the pulsed signal whose duration is about 200ns will be higher than the
threshold. The output of the comparator will be a series of negative pulses with durations less than
200ns if there are pulsed signals in the preamplifier output. Conversely, the output of the comparator
will stay at a high level if there are no pulsed signals in the preamplifier output.
Since the output of U2 is a narrow negative pulse when a pulsed signal occurs in the preamplifier
output, it is difficult to sample such narrow pulses for detecting the absence or presence of a vehicle
in the detection zone. To remedy this, a monostable multivibrator, component U3A in Figure 1- 13 is
used to keep the comparison result of U2 to the next cycle of the timing clock. In this way, the
comparison result of U2 can be sampled easily. The output of U2 is coupled to the input of U3A by
C4 and an “ OR gate” U5A. The negative pulse output from U2 will trigger U3A to output a positive
pulse from its output Q if U5A is controlled open by the low level of the timing signal. The pulse
width of the output of U3A is determined by parameters of the external RC circuitry ( R10 and C5) of
U3A, which can be calculated as: 10 5 t 0.45 R C w = × × , where is the pulse width in ns, is the
resistor in kΩ and is the capacitor in pF. In the circuitry we set
w t 10 R
5 C t s w = 110μ , which is slightly
longer than the timing clock cycle T s clock = 95μ . If U3A is triggered continually in every , its
output will always be at a high level. Once the trigger stops, the output of U3A will jump to a low
level
clock T
110μs after it was last triggered. Since the laser diodes are triggered at the beginning of every
cycle of the timing clock and an APD will output a pulsed signal very near to the time it receives the
reflected laser, the output of U3A will go to a low level about 15μs after the beginning of the first
timing clock cycle from when the trigger to U3A stops.
18/ 72
The output of U3A is related to vehicle detection. A high level means the laser beam is unblocked and
a low level means the laser beam is blocked. Vehicle detection can be performed by sampling the
output of U3A. This signal is sampled 25μs after the beginning of every timing clock cycle to get
correct and timely vehicle detection information.
1.3.4.3 Hysteresis to Overcome Noise and Signal Amplitude Fluctuation
The amplitude fluctuation of the preamplifier output will cause the output of U3A to toggle when the
amplitude is close to the threshold, leading to incorrect vehicle detection. This problem often happens
in the transition period when the laser beam is being blocked or unblocked, e. g., when a vehicle is
moving into the detection area or leaving the area, the laser light can be blocked or unblocked
gradually so the amplitude of the pulsed signal in the preamplifier output will go downward or go
upward gradually through the threshold during the transition period. Using hysteresis on the
comparator is a simple and effective way to solve this problem. With hysteresis, the comparator has
two thresholds, an upper threshold and a lower threshold. When the peak- to- peak value of the
amplitude fluctuation is in the range between the upper and lower thresholds, the effects of
fluctuations can be eliminated.
Resistor R12 is used to configure comparator U2 with hysteresis. It connects the complementary
output Q of monostable multivibrator U3A and the non- inverting input of comparator U2 to form a
positive feedback. The output Q of U3A will then contribute to switch the threshold of U2 together
with the reference voltage. The voltage on the non- inverting input of U2 can be calculated as:
a ref Q V
R R
V R
R R
V R
11 12
11
11 12
12
+
+
+
= , where is the wiper output of U4, and ref V Q V is the output Q of U3A
which can be either 0V or 5V.
When monostable multivibrator U3A is triggered, its output Q will go to low level that will be kept
to the next timing clock cycle. This status of Q through R12 will drive the level of the non- inverting
input of U2 lower than when R12 is not connected. This level is “ low threshold”. As long as the
amplitudes of the sequential pulsed signals are higher than the “ low threshold”, U2 will output
negative pulses reliably and keep the output Q of U3A at low level. However, if the amplitude of the
pulsed signal is lower than the “ low threshold”, U2 will not output negative pulses any longer, so the
output Q of U3A will jump to a high level 15μs after the beginning of the timing clock cycle, which
will lead to a higher level on the non- inverting input of U2. This higher level is “ high threshold”.
During the next clock cycle, only when the amplitude of the pulsed signal is higher than the “ high
threshold”, can U2 output a negative pulse to trigger U3A and force the output Q of U3A to a low
level, which will drag the non- inverting input of U2 to “ low threshold” again. If this condition
happens, U2 will output negative pulses over the sequential clock cycles until the signal amplitude
becomes lower than “ low threshold”. In this way, the effect of the amplitude fluctuation of the pulsed
signal is greatly reduced.
1.3.4.4 Digitally Controlled Potentiometer
Correctly setting the threshold of the comparator, i. e.“ calibration”, is important to the proper
operation of the circuitry because the amplitude of the pulsed signal output from the preamplifier will
vary as the operation environment changes. To obtain correct detection results in different
environments, the circuitry should have the ability to automatically adjust the threshold to proper
values.
Component U4 in Figure 1- 13 is a nonvolatile digitally controlled potentiometer ( DCP), which is used
to adjust the variable reference voltage for the threshold of the comparator to realize optimal self-calibration.
Its wiper can be moved in 128 steps between its two terminals under the control of 3- wire
19/ 72
digital signalsCS , INC and . When the wiper is moved between the bottom and the top
terminals of the internal resistor array, the wiper output will vary between 0V and 5V. Since the
Rabbit microprocessor is used in the LBDS, the variable reference voltage can be adjusted by the
DCP automatically under the control of the microprocessor.
U / D
1.3.4.5 Timing Window
A timing window is an efficient method to reduce disturbances of external linear noise. It cannot
reduce the random noise such as sunlight. Only signals within timing window can pass while all
signals ( e. g., noise) outside the window will be cut off. The timing window is important to these
electronics. Since the monostable multivibrator U3A will be triggered as long as a negative pulse is
exerted on its input pin “ A”, it is easily disturbed. Component U5A is an “ OR gate” that is used to set
up the timing window. It is inserted in the connection between the output of U2 and the input pin “ A”
of U3A, and acts as a strobe to let only the real trigger pulses pass. One of its inputs is connected to a
“ timing” control signal whose waveform is as the upper line in Figure 1- 14. The “ timing” signal will
fall to a low level when the laser diode is triggered and rise to a high level about 4μs after. In the
period when “ timing” is at a low level, U5A is opened and the negative pulse from U2 can pass
through U5A to trigger U3A. This period is the “ timing window”. However, when “ timing” is at a
high level, no signal can pass through U5A, so U3A is isolated from noise and disturbances. The
lower line in Figure 1- 14 is the pulsed signal in the preamplifier output. The “ timing” signal is
generated by a monostable multivibrator, which is triggered by the timing clock. The width of the
“ timing window” can be adjusted by changing the external RC parameters of the monostable
multivibrator.
Figure 1- 14. Timing Window.
1.3.4.6 Pre- board Noise Data Analsys and Data Process
The pre- board is in the front of acquisition system, and it is important to the system, the sunlight and
oscillating circuit noise effect the signal acquisition. According to the analysis of the signal oscillating
spectrum Figure 1- 15, Signal analysis experiments were carried out using a spectrum analyzer. It was
found that a high frequency range provided the best S/ N ratio. the sunlight noise signal is lie on the
lower frequency range, the pulse signal is lie on the higher frequency range, we used the high- pass
20/ 72
filter to improve the S/ N ratio . The old board that used the two- stage amplifier, the filter frequency is
not high enough that the sunlight noise effect the pulse signal. If the frequency is higher, the signal of
the second stage amplifier is very weak. The signal acquisition circuitry were redesigned and adjusted
to reduce the optical noise due to sunlight and oscillating noise.
Figure 1- 15 Pre- board sunlight signal spectrum
Figure 1- 16 Outside testing signal on time domain
21/ 72
Figure 1- 17 Roof testing signal on time domain
1.3.4.7 Arrangement of Printed Circuit Boards in Signal Box
The new system uses 24 detection channels on one side, i. e. 24 circuits same as the circuit shown in
Figure 1- 13 should be installed in a signal box. To make the signal box have a smaller size which
should fits the optical system, the arrangement of all the components and the PCB boards should be
carefully designed. At present, we use large scale integrated circuits to reduce the required space of
PCB board, so that 8 circuits for 8 channels can be designed in one small sized board. 3 such boards
make 24 channels for detection in one side. These 3 boards are installed in the signal box horizontally
as 3 layers with one end connecting the APD board and the other end connecting a back board which
has some connectors to connect the main control board by cables.
1.3.5 Power Supply Circuit and DC/ DC Converter
The power supply circuit provides various DC powers needed by the electronics system of LBDS. Its
input is 24VDC which comes from a powerful 24VDC power supply and its outputs are + 5V, - 5V,
and - 12V, as shown in Figure 1- 18. The VKP60xP 60 Watt triple output half brick DC/ DC converter,
The 12VDC input is converted to + 5V, + 12V and - 12V DC power output by a powerful DC/ DC
converter. The + 5V is used by the main control system, and the + 24V is used to drive the laser diode
module. Two voltage regulators U2 and U3 are used to obtain - 5V from - 12V supply. The + 5V and -
5V are used to power the signal pre- processing circuitries on both sides. The - 12V is also used to
power the negative bias voltage circuitry for APD. In the previous design of the LBDS, the power
supply is a DC/ DC converter which converts the 24DC to + 5V, - 5V, + 12V, and - 12VDC. The reason
of using the 24VDC as the input of the power supply is that the LBDS can be made mobile with the
trolley on the truss, and the DC power supply can be easily obtained from the rails of the truss.
22/ 72
Figure 1- 18. Power supply circuitry using DC/ DC converter.
1.3.6 Temperature Compensation for Responsivity of APD
The responsivity of APD will change as the temperature of operating environment changes due to its
temperature performance [ 15], which results in the fact that the signal amplitude of APD output and
accordingly the amplitude of pre- amplified signal will be different in different temperature even
though the other detection conditions remain the same. The higher the environment temperature is,
the smaller the signal amplitude of APD output will be. Figure 1- 19 shows the relationship between
the amplitude of pre- amplified signal and the environment temperature. This temperature
performance of APD will lead to misdetection when LBDS works in a wide range of temperature,
since the LBDS detects vehicles just by comparing the amplitude of pre- amplified signal with a
threshold. If the high temperature makes the signal amplitude become so small that it is always lower
than the threshold even if no vehicle blocks the laser, the signal processing circuitry of LBDS will not
be able to distinguish whether there is a vehicle in the detection zone or not. Considering that LBDS
is designed to be used outdoors and the highest outdoor temperature in hot summer day may exceed
40° C , the temperature compensation is necessary for correct operation of LBDS.
23/ 72
Signal- Temperature
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
20 25 30 35 40 45
Temperture( C)
Signal( V)
Figure 1- 19. Relationship between signal and temperature.
Temperature compensation can be realized by changing the bias voltage of APD. The responsivity of
APD will change as the bias voltage changes according to its performance. More negative bias
voltage will bring to bigger signal amplitude. Figure 1- 20 shows the relationship between the
amplitude of the pre- amplified signal and the bias voltage, which was measured at temperature of
. To reduce the effect of temperature variation on the amplitude of the signal, the bias voltage
should be adjusted to more negative when the temperature rises higher. Although it seems possible
that big amplitude of signal in a wide temperature could be obtained only by setting the bias voltage
at a high negative value, it is not realizable. The bias voltage of APD has a highest negative voltage
limit which is different in different temperature. If the bias voltage exceeds this limitation, the signal
will vibrate and can not be used for detection.
25° C
The circuitry for temperature compensation is shown in Figure 1- 12. It is a temperature controlled
variable bias voltage supply. R1 and R2 are thermistors with negative temperature coefficient. They
are mounted on the covers of the APD arrays and work as temperature sensors. When temperature
rises, the resistances of R1 and R2 reduce, which lead to voltage increments on the outputs of U1 and
U2. The bias voltages on APD arrays are therefore raised. In this case, the signal decreasing tendency
caused by the higher temperature will be counteracted by the signal increasing tendency caused by the
more negative bias voltage. The result of a temperature compensation test of this circuitry showed
that the signal amplitude can be kept at 3V in a wide range of temperature from 14° C to 42° C .
24/ 72
signal- bias
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
- 350 - 340 - 330 - 320 - 310 - 300 - 290 - 280
bias voltage( V)
signal( V)
Figure 1- 20. Relationship between signal and bias voltage.
1.3.7 Data Acquisition and Output
The data acquisition is the operation that the Rabbit microprocessor obtains the detection data from
the signal pre- processing circuitries on both sides of LBDS, and the data output is the operation that
the detection data are transmitted from LBDS to the remote PC for calculating the traffic parameters.
To make the LBDS more flexible for the requirements of different applications, two kinds of data
outputs are designed in the LBDS. One of them is parallel data output through a parallel port, and the
other is serial data output by Ethernet communication through an onboard Ethernet port. The circuitry
is shown in Figure 1- 21.
25/ 72
1 G
2 A1 Y1 18
4 A2 Y2 16
6 A3 Y3 14
8 A4 Y4 12
U2A
74HCT244
19 G
11 A1 Y1 9
13 A2 Y2 7
15 A3 Y3 5
17 A4 Y4 3
U2B
74HCT244
1 G
2 A1 Y1 18
4 A2 Y2 16
6 A3 Y3 14
8 A4 Y4 12
U7A
74HCT244
19 G
11 A1 Y1 9
13 A2 Y2 7
15 A3 Y3 5
17 A4 Y4 3
U7B
74HCT244
D0
D1
D2
D3
D4
D5
D6
D7
D0
D1
D2
D3
D4
D5
D6
D7
ch1 of side A
ch2 of side A
ch3 of side A
ch4 of side A
ch5 of side A
ch6 of side A
ch7 of side A
ch8 of side A
ch17 of side B
ch18 of side B
ch19 of side B
ch20 of side B
ch21 of side B
ch22 of side B
ch23 of side B
ch24 of side B
D0
D1
D2
D3
D4
D5
D6
D7
binary digital signals
from signal pre- processing
circuitry of side A
binary digital signals
from signal pre- processing
circuitry of side B
VCC
1 A
2 B
3 CLR Q 4
Q 13
14 Cext
15 RCext
U8A
74HCT123
R1
C1
VCC
9 A
10 B
11 CLR Q 12
Q 5
6 Cext
7 RCext
U8B
74HCT123
R2
C2
VCC
VCC
1 OE
3 D0 Q0 2
4 D1 Q1 5
7 D2 Q2 6
8 D3 Q3 9
11 LE
13 D4 Q4 12
14 D5 Q5 15
17 D6 Q6 16
18 D7 Q7 19
U9
74HC373
1 OE
3 D0 Q0 2
4 D1 Q1 5
7 D2 Q2 6
8 D3 Q3 9
11 LE
13 D4 Q4 12
14 D5 Q5 15
17 D6 Q6 16
18 D7 Q7 19
U14
74HC373
POD1
POD2
POD3
POD4
POD5
POD6
POD7
POD41
POD42
POD43
POD44
POD45
POD46
POD47
POD48
Port A
PB2
PB7
RCM3200
RabbitCore Module
PF7 clock
Latch signal
ch1 of side A
ch2 of side A
ch3 of side A
ch4 of side A
ch5 of side A
ch6 of side A
ch7 of side A
ch8 of side A
Interrupt signal
PE0
Ethernet
Port
TCP/ IP
Parallel
Output
Communication
U1
PB3
PB4
PB5
PB6
ch21 of side B
ch22 of side B
ch23 of side B
ch24 of side B
ch17 of side B
ch18 of side B
ch19 of side B
ch20 of side B
POD8
Figure 1- 21. Data acquisition and output circuit.
1.3.7.1 Rabbit Data Acquisition
The output signal from one channel of the signal preprocessing circuitry is a binary logic signal. 24
binary signals for 24 channels in one side combine 3 bytes 8- bit data. The total 6 bytes of 8- bit data
from both sides are fed to 6 data buffer chips U2, U3, U4, U5, U6, and U7. U8A and U8B make the
interrupt generator and latch generator. When the microprocessor responds the interrupt application in
every clock cycle, it reads out all 6 bytes data from both sides through the data buffers. The advantage
of acquiring data in this way is that only one 8- bit input port in Rabbit microprocessor is needed for
reading out 6 bytes data, which reduces the numbers of the I/ O ports needed for data acquisition and
allows those I/ O ports resources to be used for other purpose.
1.3.7.2 Output Data Through Ethernet Communication
Since the microcontroller RabbitCore RCM3200 module has a 10/ 100Base- T Ethernet port available
on- board, the detection data can be transmitted via this port using TCP/ IP communication. The
network architecture for Ethernet communication in our application is client/ server. The remote PC
works as a server and the RabbitCore microcontroller in LBDS works as a client. Both of them are
assigned an IP address respectively. When LBDS starts operation, the RabbitCore microcontroller
will configure the Ethernet port and initialize the TCP/ IP interface at first, and then it tries to set up
the connection with the remote PC by sending a TCP segment to the PC. If the remote PC responds,
the microcontroller sends back the acknowledgement. Then, the connection between the PC and the
LBDS has been established, and the TCP socket is therefore opened. In the remote computer side,
after it runs our application program “ Xvehicle”, the server will create a TCP socket and wait for a
connection from LBDS by listening for the connection from the Rabbit module on the TCP port
“ 2002”. If the connection is established successfully through a server- client handshake, the server will
read the data coming from the TCP session and send the data to the user program through FIFOs. The
flow chart of the server is shown in Figure 1- 22. On the other hand, in the vehicle detection routine,
26/ 72
the microcontroller stores the detection data that obtained from the signal processing circuitries on
both sides into one of two data arrays in memory in every clock cycle. When the data array being
used is full, all data in this array will be put into a packet and the microcontroller will start to send the
packet to the remote PC via TCP/ IP communication, while the storage of the detection data will be
switched to the other data array. Both data arrays are used alternately, so the detection data can be
acquired every clock cycle and they can be sent to the remote PC in packets continually. Meanwhile,
when the Rabbit module sends out a package, a tcp_ tick (& sock) should be called. Otherwise, the
TCP package sending buffer will be full and the data package will not be sent out. With the Ethernet
communication, the operator’s commands for some actions of LBDS can also be sent from the remote
PC to LBDS to remote control the LBDS.
Figure 1- 22. The flow chart of TCP/ IP communication.
1.3.7.3 Output Data Through Wireless Ethernet
Besides the cable Ethernet communication with the remote computer, LBDS has the ability to do the
communication by wireless Ethernet communication. A wireless router is used in this case. The
Ethernet port is connected to the router which is installed near the LBDS by a cable. The remote
computer can receive the detection data through the wireless router by wireless Ethernet
communication if it has a wireless networking. Utilizing wireless network technology is more
favorable since the LBDS may be mounted above highway and the desired mobility of LBDS requires
the communication with the remote computer be wireless.
1.3.7.4 Parallel Port Output
Unlike the Ethernet communication, the parallel port outputs the detection data every clock cycle. The
six 8- bit detection data from both sides are connected directly to the inputs of the latches of parallel
ports U9, U10, U11, U12, U13 and U14. After the six 8- bit data from both sides are ready to be
readout in every clock cycle, a latch signal is generated to let the six latches U9 – U14 to refresh their
outputs. The parallel output is completely controlled by the peripheral circuitry, it will not occupy any
CPU time of the microcontroller. These 48 digital signals can be transmitted to a remote computer
through a DIO- 96 card which is installed in the computer to readout the data for calculating the traffic
parameters.
27/ 72
1.3.8 Optimal Self- Calibration
The amplitude of the signal output from a two- stage preamplifier will be affected by the operating
environment, e. g., sunlight, weather and road surface conditions, and the noise magnitude will also
vary as the environment changes. A threshold that is well configured for a particular environment
may not be suitable for another. If the threshold is fixed, incorrect detection may happen when the
environment changes. The reliability and the correctness of the detection result cannot be guaranteed.
As a result, automatically and periodically setting the threshold of the comparator to a proper value,
i. e., “ calibration”, is important to the proper operation of the LBDS.
Based on this consideration, a calibration method is used to make the detection more reliable and
flexible. The idea of this calibration is to set the threshold of the comparator U2 at the midpoint
between the lowest maximum amplitude of the pulsed signals and the highest amplitude of noise
disturbances, which will lead to stronger ability for laser signal detection and noise resistance. The
calibration can be done by adjusting the wiper position of the DCP, which is controlled by a
microcontroller. If a single threshold is used, the calibration is relatively simple. But after hysteresis is
employed in the circuitry, two thresholds will be in action. In this case, calibration is more
complicated.
Figure 1- 23 depicts equivalent circuitry of components U2 and U3A in Figure 1- 13. This is a
comparator with hysteresis. The threshold a can be calculated as follows: V
a ref out V
R R
V R
R R
V R
1 2
1
1 2
2
+
+
+
=
Assuming out can only be either 0V or 5V, we then obtain: V
( 0 )
1 2
2 V V V
R R
V R a ref out =
+
− =
or, 5 ( 5 )
1 2
1
1 2
2 V V V
R R
V R
R R
V R a ref out ⋅ =
+
+
+
+ =
Because of the hysteresis, the level will change not only due to the level of but also due to
the previous level of even when remains unchanged, i. e., if , only when
will jump to 0V, however, if
out V in V
out V ref V V V out = 5
= + > + in in a V V V out V V V out = 0 , only when , will
jump to 5V. As a result, the comparator with hysteresis can effectively avoid toggles on
when is close to one of the two thresholds and .
= − < − in in a V V V
out V out V
in V +
a V −
a V
Figure 1- 23. Comparator with hysteresis.
28/ 72
The idea of “ calibration” here is to set to a value where the amplitude margins of the signal and
the noise are the same. The principle of the calibration is described in Figure 1- 24. Lines ① and ②
are the relations between and , while line ③ is
ref V
V ref V a a ref V = V without feedback. Suppose
is the peak value of the pulsed signal and is the peak value of noise. Their relative values of
are and respectively. We want to find the optimal value that makes the length
between points and the same as the length between and . These two lengths are
amplitude margins of the signal and the noise. Here, is the level that will remain at ( 0V) if
the noise level is higher than this level, and is the level that will remain at ( 5V) if the
signal level is lower than this level. The value of can be derived by letting the length between
and equal to the length between b and . The calculation is as follows:
signal V
noise V
ref V H V L V ref V M V
a amH V b amL V
amL V out V
amH V out V
M V a
amH V amL V
amL M V
R R
V R ⋅
+
=
1 2
2 , 5
1 2
1
1 2
2 ⋅
+
⋅ +
+
=
R R
V R
R R
V R amH M ,
noise L V
R R
V R ⋅
+
=
1 2
2 , 5
1 2
1
1 2
2 ⋅
+
⋅ +
+
=
R R
V R
R R
V R signal H ,
Let: amL noise signal amH . V − V = V − V
Then we obtain:
5 5
1 2
1
1 2
2
1 2
1
1 2
2
1 2
2
1 2
2 ⋅
+
⋅ −
+
⋅ −
+
⋅ +
+
⋅ =
+
⋅ −
+ R R
V R
R R
R
R R
V R
R R
V R
R R
V R
R R
R
M L H M
That is ( ) ( )
1 2
2
1 2
2
H M M L V V
R R
V V R
R R
R ⋅ −
+
⋅ − =
+
, so we have: ( )
2
1
M H L V = ⋅ V + V .
If and are determined, can be obtained. But unfortunately and are unknowns
and they need to be measured by the LBDS itself. Since there are no A/ D converters available in the
LBDS, the and values have to be determined by moving the wiper of the DCP ( U4) between
its two terminals, which will adjust between 0V and 5V to search for the wiper position where a
transition on the output level of U3A occurs. The value of can be determined from the DCP
wiper position. The flowchart of the calibration procedure is shown in Figure 1- 25.
H V L V M V H V L V
H V L V
ref V
ref V
29/ 72
Figure 1- 24. Demonstration for calibration.
Since the DCP wiper position is related to , the value of can be set by moving the wiper to
certain positions. At the beginning of the flowchart, the laser diodes are triggered once every 2 clock
cycles, and the wiper position of DCP is set to the highest point, where . At this time,
and
ref V ref V
V V ref = 5
V V out = 5 V V out = 0 since the signal pulse amplitude is lower than 5V. The microcontroller then
controls the DCP to move its wiper downward step by step. The microcontroller tests out V in the
clock cycles when laser diodes are triggered. 10 values of out V will be tested over 20 clock cycles
every step. Testing 10 out V values over 20 clock cycles is due to the requirement to eliminate the
effect of amplitude fluctuation of the pulsed signal and to initialize to 5V for every cycle in
which
out V
out V value is tested. When the 10 values of out V tested over 20 clock cycles are all at high
level, the current step of the wiper corresponds to , and the microcontroller will record this step
for later calculations. Once has been obtained, the microcontroller will change to find the
value. At this point, the wiper of DCP will be set to the lowest position and then move upwards step
by step while 10 values of
H V
H V L V
out V will be monitored over 12 clock cycles at every step. If the 10 values
of out V tested over 12 clock cycles are all at low level, the current step of the wiper corresponds to
L . The microcontroller will also record this step for later calculations. V
30/ 72
Since we want to ascertain the maximum level of noise, , the strategy for determining it is as
follows. Set the DCP wiper to a certain step, initialize to 0V by triggering laser diodes only one
clock cycle at the beginning of the test in this step, stop the triggering of the laser diodes over the
following 11 clock cycles so that there is only noise on the input of the comparator, and then sample
10 values of
noise V
out V
out V over last 10 clock cycles in this step. At the first wiper steps of the DCP, is
small, so
ref V
a ref V
R R
V R
1 2
2
+
− = will be lower than and input noise will cause to stay at 0V.
But as the wiper of the DCP moves upwards step by step and increases, will jump to 5V as
soon as
noise V out V
ref V out V
a ref V
R R
V R
1 2
2
+
− = is greater than noise . By sampling V out V , the step corresponding to
can be determined and recorded. To initialize in this case, 12 clock cycles are used at every step.
The laser diode will be triggered once in the first cycle to force to 0V, but won’t be triggered in
the remaining ( 11) cycles. The values of
L V
out V
out V
out V sampled in the first two cycles will be given up because
the effect of the laser signal on the monostable multivibrator ( U3A) may still exist, while the values
sampled in later 10 cycles will be monitored for determining . In this way, and can be
obtained reliably.
L V H V L V
After L and have been obtained, the microcontroller can calculate according to V H V M V
( )
2
1
M H L V = ⋅ V + V and get the corresponding step setting for the DCP. Once the optimal step
setting is worked out, the microcontroller will first set the wiper of DCP to the bottom terminal, and
then move the wiper to the expected position.
The calibration procedure has the ability to provide an error message if the operating environment is
too difficult for the LBDS to operate correctly.
31/ 72
Figure 1- 25. Flowchart of calibration procedures.
1.3.9 Remote Control for Trolley through Wireless Ethernet
The LBDS may sometime be mounted on a trolley moving on a truss as a mobile LBDS. The motion
of the trolley should be remotely controlled by the operator who stands on the grand to let the trolley
move to the correct position. Radio remote control can be used for the remote control of the trolley,
but since the wireless Ethernet communication is used in LBDS, it is favorable to utilize the wireless
Ethernet communication for remote control of the trolley. To realize this idea, The Rabbit
microprocessor is used for the control, the new circuit board has been designed and fabricated as
shown in Figure 1- 26, and they communicate through the router. When the Rabbit microprocessor
receives a command for the movement of the trolley through wireless Ethernet communication, it will
transmit the corresponding control signals to the trolley, and also control the power supply of LBDS.
32/ 72
In this way, the remote computer can be used not only to receive the detection data, but also to control
the trolley.
Figure 1- 26 The trolley prototype
1.3.10 Software Design for Rabbit Microprocessor
Rabbit microprocessor is used in LBDS, so a control program for the operation of LBDS has been
designed for the microprocessor. The program is written mainly in C language with some parts
written in assembly language where the time constraints are strict. It consists mainly of a main
program, a calibration subroutine and an interrupt service subroutine. The flowcharts are shown in
Figure 1- 27.
The main program is used to initialize all programmable components, send real- time data to the server
and test whether re- calibration is necessary. After the system is powered on, the main program
initializes the system and then goes into an infinite loop.
The interrupt service subroutine is used to read sampling data from signal pre- processing circuits.
This subroutine will be carried out once every 100μs. All 6 bytes, i. e. 48- bit, of sampling data will be
readout at the beginning of the subroutine, and then the data will be put into a buffer for transmitting
to a computer through an Ethernet cable. To ensure continuous data transfer, two buffers are used in
turn to store and send data.
The calibration subroutine is used to set the wipers of all 48 DCPs to their optimal positions. The
procedure of the subroutine has been described above.
33/ 72
Figure 1- 27. The flowcharts of control program on Rabbit microprocessor.
readData()
This is an interrupt service subroutine. It reads in the 48- bit sampling data from port A and stores the
data into two array buffers in turn.
setXDCP ()
This function is used to initialize all 48 DCPs when the system powered on.
setPwm()
This function sets the PWM module with required parameters to generate the timing clock signal.
setPortInit()
This function initializes all Rabbit ports that will be used in this system.
calibration()
This function is used to set all 48 DCPs to their optimal positions.
1.4 New Design of the Mechanicle System
1.4.1 Breadboard for opticle system
An optical breadboard was purchased and used as the base of the LBDS. The breadboard is a 0.5in
aluminum optical breadboard that provides more stability as compared to the previous base made
from 0.125in aluminum with steel L beams attached to the bottom for rigidity. Using aluminum
reduces the weight of the overall system. Portions of the breadboard can be pocketed out to further
reduce the overall weight.
34/ 72
1.4.2 Acrylic Box for Whole LBDS
To better protect the LBDS, a housing was fabricated for the LBDS from acrylic as shown in Figure
1- 28. Further step are required to fully weather proof the LBDS.
Figure 1- 28. LBDS with housing.
1.5 Software on remote computer
With our application software “ Xvehicle” running on it, the remote computer can receive the
detection data from LBDS through wireless Ethernet communication or through parallel
communication. The received raw data will be then used to calculate the traffic parameters such as
velocity, acceleration and length of vehicles. These calculated parameters will be displayed on the
remote computer. Figure 1- 29 shows the snapshot of the screen of the remote computer.
Figure 1- 29. Snapshot of the screen of the remote computer.
35/ 72
2 Comparison Study with the Previous System
2.1 Longer length of the Detection Area
The length of the laser line on the road surface for detection used to be only 2.5m even if it has been a
combination of two laser lines, and the length of the view range on road surface of the optical system
is even smaller in the previous design, i. e., less than 2m. Since the width of a highway lane normally
measures 3.8m, the laser line of 2.5m and the view range of 2m make the previous design of the
LBDS can not be practically used on highway. With the new laser module and the newly designed
optical system, the laser line on road surface has been enlarged to 5m, and the length of the view
range of the new optical system can reach 4m. These improvements enable the LBDS to do practical
detection on highway.
2.2 24 channels
In the previous design, only 8 elements of an APD array on one side are used. 8 pair elements on both
sides make 8 channels for detection. Even though the view range measures 2m, the resolution of the
detection is only 2/ 8= 0.25m/ channel. In the new design, the numbers of the used elements of an APD
are increased to 24, making 24 channels on both sides. Considering the length of the view range has
been 4m, the detection resolution of the new system becomes 4/ 24= 0.166m/ channel, which is still
better than the previous system.
2.3 Stronger Anti- noise Ability
The new design uses several anti- disturbance technologies such as hysteresis on the comparator,
timing window, and optimal self- calibration when processing the detection signal. These technologies
make the LBDS obtain stronger anti- noise ability. So that LBDS can implement correct and accurate
detection reliably and stably in a detection site with heavy- noise.
2.4 Flexibility for Different Environment
The temperature compensation and the optimal self- calibration are firstly used in the new system
design. The temperature compensation let the variation on the responsivity of APD when the
environment temperature changes can be compensated, so that the amplitude of the analog signal can
be stabilized in a wide temperature range. The optimal self- calibration ensures all the DCPs are in
optimal positions for correct detection. The effect of the variation on the amplitude of the analog
signal, whatever it is caused by, can be eliminated by this algorithm. With these advanced
technologies, the new LBDS has obtained high flexibility to apply different environments, and it has
also got stronger ability to operate correct detection reliably for a long term.
2.5 Communication
The previous design only has two ways for transmitting the detection data to the remote computer,
i. e., cable Ethernet communication and parallel communication. In the new designed system, the
wireless Ethernet communication is added. With this new communication way, the LBDS obtains the
mobile ability and will be more flexible to different applications.
2.6 laser Alignment
The new system has a pair of laser pointers, which are not used in the previous design, to help the
alignment of the infrared laser light. This green laser pointer is very bright. It can be clearly seen even
in the sunlight. With these laser pointers, the infrared laser lines can be aligned to the correct
positions easily.
36/ 72
3 Testing
Several different testing methods are utilized with the Laser- based Detection System. Indoor lab
testing provides immediate feedback for system calibration, minor modifications, and software
changes. Roof testing at Bainer Hall provides a more realistic testing method that includes an actual
vehicle while allowing for minor system tweaks and electrical characterization. Mobile truss testing
provides a on- campus faster vehicle testing scenario. Highway tests provide results from field usage
and are completed at the highway I- 5 and highway I- 80 junction in Sacramento.
3.1 Indoor Lab Testing
Indoor lab testing allows for immediate feedback from adjustments and minor hardware and software
modifications, and provides a stable location for calibration of the LBDS. In order to help with system
calibration, a simulation device and multiple surface pallets were created as shown in Figure 3- 1. The
simulation device consists of a form board cutout that slides on an aluminum rod. The form board was
cut to have a curved front and straight back in order to test for accurate speed determination of all
sensor elements given variational object width and horizontal symmetry. Different types of surfaces
were used to test the LBDS for different laser absorptivity. Indoor tests were generally completed
using two asphalt roofing strips to simulate a road surface since it had the maximum laser absorptivity
of all the testing surfaces.
Figure 3- 1. Indoor lab testing setup of the LBDS.
Results from indoor tests varied depending on the test completed and the equipment that was utilized.
General electrical tests were completed using an oscilloscope in which the signal waveform was
monitored. Optical tests were completed using a CCTV camera with high sensitivity in the infrared
region attached to a standard TV via a coaxial cable. Optical testing focused on the orientation and
37/ 72
sharpness of the projected laser line. Indoor vehicle simulation using the simulation device was done
using a computer running the XVehicle program which displays the calculated velocity for all 24
sensor element pairs of the LBDS.
The indoor lab testing procedures are as follows:
1) Check whether the laser beam is in a proper position on the wall. Using the CCD camera
additional the filter, the laser beam can be display on the TV. If the laser line cannot be seen,
inspect the laser module opening first. If the laser can be seen in the opening, check whether
the laser line is being blocked by an obstacle. If the laser beam cannot be seen in the module
opening, check the clock circuit and the 24V DC power supply.
2) Check whether the two laser beams are aligned in one line ( both sides). Using CCD camera
connect with the optical system, you can see the laser beam on the TV, then adjust the laser
module to let the beam align.
3) Increase the aperture of the image lens to half opening.
4) Attach the oscilloscope probe to the output of the pre- amplifier ( pin 8) and move the laser
line by adjusting the screw on the rear of the laser mount until the maximum signal is
obtained.
5) Reduce the aperture until the width of the signal is about 100ns and not saturated. .
6) Push the reset button to communicate with a computer.
7) Push the calibration button to set DCP. The output of the digital signal displayed in the
computer can be checked in states of blocked or unblocked.
3.2 Roof Testing
The LBDS can be easily and conveniently setup on the roof of Bainer hall located on the UC Davis
campus as shown in Figure 3- 2. Roof testing provides for realistic controlled tests with system
accessibility for electrical signal measurements. The driver of the vehicle is told to accelerate to and
maintain a speed of approximately 15MPH. The driver is then asked to repeatedly drive back and
forth while data are recorded. The XVehicle program, running on a laptop connected to the LBDS via
wireless 802.11b, is used to display the measured velocities of the sensor element pairs. Laser
blockage of the LBDS was also tested by having the vehicle slowly pass under the detection zone.
Confirmation of the laser blockage was done using the XVehicle program along with on- system
visual confirmation via red LEDs that correlate to the sensor element state.
38/ 72
Figure 3- 2 Roof testing setup of the LBDS at Bainer Hall.
3.3 Mobile Truss Testing
Highway or field testing requires the assistance of Caltrans workers and scheduling a suitable date. In
order to test the LBDS under highway velocities, a mobile truss is setup on a simulation site on old
Hutchinson Road in Davis. The LBDS is mounted to the mobile truss system as shown in Figure 3- 3.
Since it is impossible for a person to stand on the top of the truss to adjust the LBDS for different
heights, the LBDS is pre- adjusted to work at a distance of 5 meters in the lab and the truss is set at a
height of 5 meters for testing. Testing using the mobile truss provides sensor element pair velocity
results in XVehicle. The drive of the testing vehicle is once again asked to maintain a cruising
velocity. Multiple testing vehicles are utilized. The XVehicle program is run on a laptop that connects
to the LBDS over wireless 802.11b. A standard video camera may be mounted to the front of the
LBDS to capture video of the vehicle as it passes underneath the detector. The video and detector data
can be synced using the Video Sync program provided by Caltrans.
39/ 72
Figure 3- 3. A prototype of LBDS on trolley and truss.
Figure 3- 4. Field test on a roadway near UC Davis.
40/ 72
3.4 Outdoor Highway Testing
Field tests have been conducted near the junction of highway I- 5 and I- 80 in Sacramento. Figure 3- 5.
depicts the test site and the detection system mounted on the highway overpass.
Figure 3- 5. The prototype of LBDS being tested on highway near Sacramento.
Figure 3- 6. Result of the test on highway.
Figure 3- 6. shows a snapshot of the XVehicle program running on a laptop. The data given are the
velocities measured by each consecutive sensor element pair obtained with the passing of a single
vehicle. Ideally, all of the velocities would be the same since the sensor element measurements are
obtained from the same vehicle. The changes in the length are due to the curvature of the passing
vehicles. The results show a consistency in vehicle velocity measure along all the channels of the
detector with slight variation. The availability of 24 channels is a vast improvement over the previous
8. Most modern cars have a slight curve to the front and rear edges. A digital video camera is
41/ 72
generally attached to the front of the LBDS and pointed towards the ground. The video is downloaded
from the camera via an IEEE1394 fire wire port. The XVehicle user interface window contains a
menu bar, a status window, and a strip chart. The menu bar is used to control the system, i. e., to start
or stop data acquisition and to set parameters of the system. The signals from the electronic sensors
are displayed dynamically in the strip chart, which scrolls from right to left. The signal is at a high
level when a vehicle is not present, and changes to a low level when a vehicle blocks the laser. This
transition occurs over a time that is much shorter than the sampling interval. Since all the signals are
displayed at the same position, only one signal line can be seen on the chart when there is a vehicle
blocking the laser lines ( or no vehicle). Signals from different sensor elements can be distinguished
when they change from a low to high level ( or inversely) at different times. The front speed ( v1), rear
speed ( v2), acceleration ( a), and length ( l) of the vehicle are displayed on the lower part of the
window. Each line corresponds to one pair of sensor elements. Currently it can display twenty eight
pairs of sensor elements.
Figure 3- 7. Snapshots of the video sync program
It is also noted that we have incorporated a video analysis system ( Video Sync, provided by Caltrans)
for analysis of the performance of the LBDS system. Figure 3- 7. shows a snapshot of the Video Sync
program. The Video Sync program synchronizes the video recorded during highway testing with the
collected data. Since the Video Sync program takes in 60 Hz data encapsulated in the LOG_ 170
format, a quasi- standard format used in data communication in Caltrans traffic management system,
the 10kHz data output from the LBDS had to be reformatted. The left part of the figure shows the
recorded video from highway testing. The right side shows a graph of the collected data synchronized
with the video. When the video is played, the data graph will move from right to left. The green lines
“ A” and “ B” are indicators for the laser lines projected onto the road surface. The blue 0 line shown
in Figure 3- 7. corresponds to the front edge of the vehicle as it passes through the Laser A beam. The
blue 1 line corresponds to the front edge of the car as it passes through the Laser B beam. The blue 2
and 3 lines are the positions corresponding to the rear edge of the car as it passes through the Laser A
and Laser B beams. 24 of the available 48 channels from the LBDS are visible in Figure 3- 7. since the
program only supports a maximum of 8 channels. The upper 4 channels in the graph are the front 0- 3
channels while the bottom 4 channels are the back 0- 3 channels. The Video Sync software gives a
visual representation of what the system is actually capturing, giving an easier method to debug the
system. It also helps in checking the real- time accuracy of the system’s data output. In the results
42/ 72
shown, the Video Sync analysis verified that the vehicle bumpers crossed the laser sheets at about the
same time.
4 Conclusions
This report documented the work that was done to enhance the real- time laser- based non- intrusive
detection system for the true vehicular travel time measurement of a link of highway. The focus of
the ongoing research has been on enhancing reliability, accuracy and anti- disturbance capabilities of
the LBDS. New laser diode modules were purchased that are more powerful and project a longer laser
line than the previous modules. The projected laser lines have been enlarged to 4 m and almost cover
the full width of a lane on the highway. The laser lines now have a uniform intensity distribution
along the line, increasing accuracy at the edges of the detection zone. A new optical system has been
designed and manufactured that allows the avalanche photodiode ( APD) array in the LBDS to “ view”
the full width of a lane on the highway. Accordingly, the electronic systems have been redesigned.
The number of detection channels has been increased from the previous 8 channels to the present 24
channels, increasing the detection resolution. Use of a hysteresis comparator and a timing window has
reduced the effect of backdrop noise, enhancing the reliability and stability of the LBDS. Temperature
compensation for the responsivity of the APD has been implemented, which stabilizes the amplitude
of the analog signal from the APD. Temperature feedback improves system performance over a wide
range of temperatures, which afflicted previous designs. The optimal self- calibration capability of the
LBDS makes the system more flexible to different operating environments. The preprocessing circuit
board has been redesigned based on the analysis of the effect of sunlight and oscillating noise on the
system. A high- filter circuit was used to increase the S/ N ratio, which allows for stable long- term use.
Outputting the detection data through wireless Ethernet communication makes the LBDS mobile and
remotely accessible. With the above improvements on its performance, the LBDS has is now more
suitable for the practical application of vehicle detection on the highway. Field test results show
consistent results over all available sensor element pairs.
Future mechanical endeavors are aimed at weight reduction and profile compression in order to better
suite the mechanical constraints of available Caltrans structures and for overall ease in transportation
and maintenance. In order to further increase system performance, the use of linear CCDs as a
replacement for the APD arrays could be used to parallelize the detection algorithm and drive the
linear CCDs. The optical system could be enhanced to potentially view two lanes of highway instead
of one, which would reduce the number of detectors required for a given detection locus along the
highway. Further refinement of the circuitry would be done to reduce the power required and the heat
generated.
References
[ 1] J. Palen, “ The Need for Surveillance in Intelligent Transportation Systems,” Intellimotion,
vol. 6, no. 1, pp. 1- 10, Mar. 1997.
[ 2] J. Palen, “ The Need for Surveillance in Intelligent Transportation Systems --- Part Two,”
Intellimotion, vol. 6, no. 2, pp. 1- 17, Jun. 1997.
[ 3] “ Traffic Detector – Technical Appendix,” U. S. Dept. of Transportation, Federal Highway
Administration, April 1985.
[ 4] “ Traffic Detector Handbook” U. S. Dept. of Transportation, Federal Highway Administrations,
April 1985.
[ 5] T. M. Hussain, T. N. Saadawi, and S. A, Ahmed, “ Overhead Infrared Sensor for Monitoring
Vehicular Traffic,” IEEE Trans. Vehicular Tech., vol. 42, pp. 477- 483, Nov. 1993.
[ 6] H. H. Cheng, B. D. Shaw, J. Palen, J. E. Larson, X. Hu, and K. V. Katwyk, “ Non- Intrusive
Laser- Based System for Detecting Objects Moving Across a Planar Surface,” US Patent, No. 6,
404,506, June 11, 2002.
[ 7] B. Lin, H. H. Cheng, B. D. Shaw, and J. Palen, “ Optical and electronic design for a field
prototype of a laser- based vehicle delineation detection system,” Optics and Lasers in
Engineering, vol. 36, pp. 11- 27, Jul. 2001.
43/ 72
[ 8] H. H. Cheng, B. D. Shaw, J. Palen, J. E. Larson, X. Hu, and K. V. Katwyk, “ A Real- Time
Laser- Based Detection System for Measurement of Delineations of Moving Vehicles,”
IEEE/ ASME Trans. on Mechatronics, vol. 6, pp. 170- 187, Jun. 2001.
[ 9] Z. Wang, B. Chen, H. H. Cheng, and B. D. Shaw, “ Performance Analysis for Design of a High-
Precision Electronic Opto- Mechanical System for Vehicle Delineation Detection on Highway,”
Journal of Mechanical Design, vol. 125, pp. 802- 808, Dec. 2003.
[ 10] H. H. Cheng, B. Shaw, J. Palen, B. Lin, B. Chen, and Z. Wang, “ Development and Field Test
of Laser- Based Non- Intrusive Detection System for Identification of Vehicles on the
Highway,” IEEE Trans. Intelligent Transportation Systems, Vol. 6, No. 2, pp. 147- 155, June
2005.
[ 11] Zhaoqing Wang, Stephen S. Nestinger, Harry H. Cheng, Benjamin D. Shaw, Joe Palen.
“ REAL- TIME ARCHITECTURE FOR AN ELECTRO- MECH- OPTICAL SYSTEM FOR
DETECTION OF VEHICLES ON HIGHWAY,” Proceedings of DETC' 04: ASME 2004
International Design Engineering Technical Conferences And The Computers And Information
In Engineering Conference, Sept. 28– Oct. 2, 2004 Salt Lake City, Utah USA
[ 12] Zhaoqing Wang, Harry H. Cheng, Ben D. Shaw, Jeo Palen. “ Noise Rejection Using
Microprocessor Feedback Control for Variable S/ N Ratio Pulse Signals,”. Submitted to IEEE
Trans. on Instrumentation and Measurement
[ 13] Harry H. Cheng, Benjamin D. Shaw, Joe Palen, Bin Lin, Bo Chen and Zhaoqing Wang,
“ Development and Field Test of a Laser- Based Nonintrusive Detection System for
Identification of Vehicles on the Highway,” IEEE Trans. Intelligent Transportation Systems,
6( 2), pp. 147- 155, Jun. 2005
[ 14] Ping Feng, Zhaoqing Wang, Harry H. Cheng, Benjamin D. Shaw, and Joe Palen, “ Digitally
Controlled Optimal Self- Calibration for A Laser- Photodiode Array Based Vehicle Detection
System,” Proceedings of DETC/ CIE' 05: ASME 2005 International Design Engineering
Technical Conferences and Computers and Information In Engineering Conference, Sept. 24–
28, 2005 Long Beach, California, USA
[ 15] A. Prokes, and V. Zeman, “ Temperature Compensation of the Responsivity of Avalanche
Photodiodes in Free- Space Optical Communication Systems,” The IEEE- Siberian Conference
on Control and Communications, 2003. SIBCON 2003.
[ 16] Halvorson, G. A., “ Automated Real- Time Dimension Measurement of Moving Vehicles Using
Infrared Laser Rangefinders”, MS Thesis, University of Victoria, 1995.
[ 17] Larson, J. E., Van Katwyk, K., Cheng, H. H., Shaw, B., and Palen, J., “ A Real- Time Laser-
Based Prototype Detection System for Measurement of Delineations of Moving Vehicles”,
California PATH Working Paper, UCB- ITS- PWP- 98- 20. California PATH Program, Institute
of Transportation Studies, University of California, Berkeley, September 1998.
[ 18] U. Moon and B. Song, “ Background Digital Calibration Techniques for Pipelined ADC’s,”
IEEE Trans. Circuits Syst. II, vol. 44, pp. 102- 109, Feb. 1997.
[ 19] V. Liberali, F. Cherchi, L. Disingrini, M. Gottardi, S. Gregori, and G. Torelli, “ A Digital Self-
Calibration Circuit for Absolute Optical Rotary Encoder Microsystems,” IEEE Trans. Instrum.
Meas., vol. IM- 52, pp. 149- 157, Feb. 2003.
[ 20] J. Guo, W. Law, W. J. Helms, and D. J. Allstot, “ Digital Calibration for Monotonic Pipelined
A/ D Converters,” IEEE Trans. Instrum. Meas., vol. IM- 53, pp. 1485- 1492, Dec. 2004.
44/ 72
Appendix A: Circuit and Board Diagrams
Main control board top layer
Main control board bottom layer
45/ 72
Main signal board top layer
Main signal board bottom layer
46/ 72
Signal preamplifier and shaping board top layer
47/ 72
Signal preamplifier and shaping board bottom layer
48/ 72
Appendix B: Source Code
Source Code in rabbit 3200 for control and communication
# class auto
/*****************************************************************
* Pick the predefined TCP/ IP configuration for this sample. See
* LIB\ TCPIP\ TCP_ CONFIG. LIB for instructions on how to set the
* configuration.
* Config # 100 = lbds_ config. lib
*****************************************************************/
/* Defines the size of the buffer that is used to send TCP datagrams */
# define ETH_ MTU 1000
# define BUFF_ SIZE ( ETH_ MTU- 40)* 6 // must be smaller than ( ETH_ MTU - ( IP Datagram + TCP Segment))
# define TCP_ BUF_ SIZE (( ETH_ MTU- 40)* 10) // sets up ( ETH_ MTU- 40)* 6 bytes for Tx & Rx buffers
# define TCPCONFIG 100
# define TCP_ OPENTIMEOUT 100
//# define RETRAN_ STRAT_ TIME 12
/* Defines the interval( in seconds) to send out TCP packet */
# define HEARTBEAT_ RATE 5
/* Defines the local TCP port to send packets */
# define LOCAL_ PORT 2801
/* Where to send the heartbeats. If it is set to all ' 255' s, it will be a broadcast packet */
# define REMOTE_ IP " 10.0.0.11"
/* the destination port to send to */
# define REMOTE_ PORT 2002
# memmap xmem
# define D_ SIZE 100
# define D_ CHANNELS 6
# define PWM_ FREQUENCY 10000
/* define the same as the control. h in xvehicle */
# define START_ TASK 1
# define STOP_ TASK 2
# define PLAY_ TASK 3
# define RESET_ TASK 4
# define CALIBRATE_ TASK 5
# define CLOSE_ TASK 6
# define TROLLEY_ STEPLEFT 7
# define TROLLEY_ STEPRIGHT 8
# define TROLLEY_ MOVELEFT 9
# define TROLLEY_ MOVERIGHT 10
# define TROLLEY_ STOP 11
# define TROLLEY_ LOCK 12
# define TROLLEY_ UNLOCK 13
# define TROLLEY_ ACCELERATE 14
# define TROLLEY_ DEACCELERATE 15
# use " dcrtcp. lib"
/* define storage for the sampling data */
char data0[ D_ SIZE* D_ CHANNELS+ 10];
char data1[ D_ SIZE* D_ CHANNELS+ 10];
char * data_ array_ ptr;
49/ 72
/* Define storage for calibration */
char XDCP_ max[ 50], XDCP_ min[ 50]; // 0- 23 for left side, 24- 47 for right side
char testdata_ left[ 3], testdata_ right[ 3]; // storage for read data when calibration
char * testdata_ left_ ptr, * testdata_ right_ ptr; // pointers of the variables
tcp_ Socket tsock;
int count;
char flg, intflag; // intflag to indicate Rabbit has been interrupted
char tcpDataSendFlg; // tcpDataSendFlg to indicate the data stored in the memory is ready to send
char U12Shadow;
char send_ num;
int sequence1;
/***********************************************************
* Function Declaration
*
***********************************************************/
void readData();
void outData();
int sendPacket( char * buf);
void setPortInit();
void setDataInit();
void setInterrupt();
void setPwm( int requency);
void setHDCP();
void printResults();
void dispData( char * buf);
int openTcpSock();
void setXDCP();
void calibration();
void errorIndicate();
/***********************************************************
* main() function
*
***********************************************************/
int main() {
int cal_ key;
char oldflg, temp;
unsigned char command[ 1];
setPortInit();
setDataInit();
setPwm( PWM_ FREQUENCY);
setInterrupt();
testdata_ left_ ptr = testdata_ left; // save the address in the pointer
testdata_ right_ ptr = testdata_ right; // save the address in the pointer
setXDCP();
tcpDataSendFlg = 0;
oldflg= 0;
setHDCP();
if( ifstatus( IF_ ETH0))
printf(" ETH0 is up now.\ n");
while ( 1 ) {
tcp_ tick(& tsock);
// open a TCP socket for communication
if(! sock_ established(& tsock)){
if( openTcpSock()> 0)
50/ 72
sock_ mode(& tsock, TCP_ MODE_ ASCII);
}
if( BitRdPortI( PCDR, 1)== 0) {
costate {
waitfor( DelayMs( 10L ) );
if( BitRdPortI( PCDR, 1)== 0) {
while( 1) {
if( BitRdPortI( PCDR, 1)== 1) {
calibration();
break;
}
}
}
}
}
if( sock_ established(& tsock)&& sock_ bytesready(& tsock)){
if( sock_ aread(& tsock, command, 1)){
switch ( command[ 0]){
case CALIBRATE_ TASK:
printf(" calibration\ n");
U12Shadow &= 0xe0;
WrPortI( PGDR, & PGDRShadow, U12Shadow); // set PG4- PG0= 0, let Q5- Q1 of U12 = 0, store the output in
// U12Shadow
BitWrPortI( PCDR, & PCDRShadow, 1, 2);
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // stop trolley at first
calibration();
break;
case RESET_ TASK:
printf(" reset\ n");
forceSoftReset();
break;
case TROLLEY_ STEPLEFT:
printf(" trolley step moving left\ n");
U12Shadow &= 0xe0;
WrPortI( PGDR, & PGDRShadow, U12Shadow); // set PG4- PG0= 0, let Q5- Q1 of U12 = 0, store the output in
// U12Shadow
BitWrPortI( PCDR, & PCDRShadow, 1, 2);
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // stop trolley at first
U12Shadow |= 0x01;
WrPortI( PGDR, & PGDRShadow, U12Shadow); // set PG4= 0, PG3= 0, PG2= 0, PG1= 0, PG0= 1
BitWrPortI( PCDR, & PCDRShadow, 1, 2);
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // set Q1= 1 for trolley step move left
break;
case TROLLEY_ STEPRIGHT:
printf(" trolley step moving right\ n");
U12Shadow &= 0xe0;
WrPortI( PGDR, & PGDRShadow, U12Shadow); // set PG4- PG0= 0, let Q5- Q1 of U12 = 0, store the output in
// U12Shadow
BitWrPortI( PCDR, & PCDRShadow, 1, 2);
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // stop trolley at first
U12Shadow |= 0x04;
WrPortI( PGDR, & PGDRShadow, U12Shadow); // set PG4= 0, PG3= 0, PG2= 1, PG1= 0, PG0= 0
BitWrPortI( PCDR, & PCDRShadow, 1, 2);
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // set Q3= 1 for trolley step move right
break;
case TROLLEY_ MOVELEFT:
printf(" trolley moving left\ n");
U12Shadow &= 0xe0;
WrPortI( PGDR, & PGDRShadow, U12Shadow); // set PG4- PG0= 0, let Q5- Q1 of U12 = 0, store the output in
// U12Shadow
51/ 72
BitWrPortI( PCDR, & PCDRShadow, 1, 2);
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // stop trolley at first
U12Shadow |= 0x11;
WrPortI( PGDR, & PGDRShadow, U12Shadow); // set PG4= 1, PG3= 0, PG2= 0, PG1= 0, PG0= 1
BitWrPortI( PCDR, & PCDRShadow, 1, 2);
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // set Q5= 1 & Q1= 1 for trolley continous move left
break;
case TROLLEY_ MOVERIGHT:
printf(" trolley moving right\ n");
U12Shadow &= 0xe0;
WrPortI( PGDR, & PGDRShadow, U12Shadow); // set PG4- PG0= 0, let Q5- Q1 of U12 = 0, store the output in
// U12Shadow
BitWrPortI( PCDR, & PCDRShadow, 1, 2);
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // stop trolley at first
U12Shadow |= 0x14;
WrPortI( PGDR, & PGDRShadow, U12Shadow); // set PG4= 1, PG3= 0, PG2= 1, PG1= 0, PG0= 0
BitWrPortI( PCDR, & PCDRShadow, 1, 2);
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // set Q5= 1 & Q3= 1 for trolley continous move right
break;
case TROLLEY_ LOCK:
printf(" trolley brake\ n");
U12Shadow &= 0xe0;
WrPortI( PGDR, & PGDRShadow, U12Shadow); // set PG4- PG0= 0, let Q5- Q1 of U12 = 0, store the output in
// U12Shadow
BitWrPortI( PCDR, & PCDRShadow, 1, 2);
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // stop trolley at first
U12Shadow |= 0x08;
WrPortI( PGDR, & PGDRShadow, U12Shadow); // set PG4= 0, PG3= 1, PG2= 0, PG1= 0, PG0= 0
BitWrPortI( PCDR, & PCDRShadow, 1, 2);
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // set Q4= 1 for trolley locking, brake goes up
break;
case TROLLEY_ UNLOCK:
printf(" trolley unbrake\ n");
U12Shadow &= 0xe0;
WrPortI( PGDR, & PGDRShadow, U12Shadow); // set PG4- PG0= 0, let Q5- Q1 of U12 = 0, store the output in
// U12Shadow
BitWrPortI( PCDR, & PCDRShadow, 1, 2);
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // stop trolley at first
U12Shadow |= 0x02;
WrPortI( PGDR, & PGDRShadow, U12Shadow); // set PG4= 0, PG3= 0, PG2= 0, PG1= 1, PG0= 0
BitWrPortI( PCDR, & PCDRShadow, 1, 2);
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // set Q2= 1 for trolley unlocking, brake goes down
break;
case TROLLEY_ ACCELERATE:
printf(" trolley speed up\ n");
U12Shadow &= 0xe0;
WrPortI( PGDR, & PGDRShadow, U12Shadow); // set PG4- PG0= 0, let Q5- Q1 of U12 = 0, store the output in
// U12Shadow
BitWrPortI( PCDR, & PCDRShadow, 1, 2);
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // stop trolley at first
U12Shadow |= 0x18;
WrPortI( PGDR, & PGDRShadow, U12Shadow); // set PG4= 1, PG3= 1, PG2= 0, PG1= 0, PG0= 0
BitWrPortI( PCDR, & PCDRShadow, 1, 2);
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // set Q5= 1 & Q4= 1 for trolley speed accelerated
break;
case TROLLEY_ DEACCELERATE:
printf(" trolley speed down\ n");
U12Shadow &= 0xe0;
WrPortI( PGDR, & PGDRShadow, U12Shadow); // set PG4- PG0= 0, let Q5- Q1 of U12 = 0, store the output in
// U12Shadow
BitWrPortI( PCDR, & PCDRShadow, 1, 2);
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // stop trolley at first
52/ 72
U12Shadow |= 0x12;
WrPortI( PGDR, & PGDRShadow, U12Shadow); // set PG4= 1, PG3= 0, PG2= 0, PG1= 1, PG0= 0
BitWrPortI( PCDR, & PCDRShadow, 1, 2);
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // set Q5= 1 & Q2= 1 for trolley speed deaccelerated
break;
case TROLLEY_ STOP:
printf(" trolley stop!!\ n");
U12Shadow &= 0xe0;
WrPortI( PGDR, & PGDRShadow, U12Shadow); // set PG4- PG0= 0, let Q5- Q1 of U12 = 0, store the output in
// U12Shadow
BitWrPortI( PCDR, & PCDRShadow, 1, 2);
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // stop trolley
break;
default:
break;
}
}
}
printf(" testdata_ left= % x % x % x\ t\ t", testdata_ left[ 0], testdata_ left[ 1], testdata_ left[ 2]);
printf(" testdata_ right= % x % x % x\ n", testdata_ right[ 0], testdata_ right[ 1], testdata_ right[ 2]);
if ( tcpDataSendFlg) {
tcpDataSendFlg = 0;
// printf(" seq:% d\ n", sequence1);
switch ( flg) {
case 0: // data1 is ready to be sent
if( sock_ established(& tsock))
sendPacket( data1);
break;
case 1: // data0 is ready to be sent
if( sock_ established(& tsock))
sendPacket( data0);
break;
case 2:
// adjust potentiometer
// setXDCP();
break;
default:
break;
} // end of switch( flg )
} // end of if( tcpDataSendFlg== 1)
} // end of while()
}
/****************************************************************************
* setHDCP() For DHCP setup
*
* A very basic TCP/ IP program, that will initilize the TCP/ IP interface,
* if set TCPCONFIG as 3, DHCP ( Dynamic Host Configuration Protocol) or
* BOOTP ( Bootstrap Protocol) will be used to obtain IP addresses and other
* network configuration items.
* Otherwise, using the static IP address for this rabbit.
****************************************************************************/
void setHDCP() {
int status;
# if TCPCONFIG== 3
// Set runtime control for sock_ init()...
ifconfig( IF_ DEFAULT, // ( DHCP only works on default interface)
IFS_ DHCP_ TIMEOUT, 6, // Specify timeout in seconds
IFS_ DHCP_ FALLBACK, 1, // Allow use of fallbacks to static configuration
IFS_ ICMP_ CONFIG, 1, // Also allow use of directed ping to configure ( only if DHCP times out).
IFS_ UP,
53/ 72
IFS_ END);
printf(" Starting network ( max wait % d seconds)...\ n", _ bootptimeout);
# endif
status = sock_ init();
switch ( status) {
case 1:
printf(" Could not initialize packet driver.\ n");
exit( 1);
case 2:
printf(" Could not configure using DHCP.\ n");
break; // continue with fallbacks
case 3:
printf(" Could not configure using DHCP; fallbacks disallowed.\ n");
exit( 3);
default:
break;
}
# if TCPCONFIG== 3
if (_ dhcphost != ~ 0UL)
printf(" Lease obtained\ n");
else {
printf(" Lease not obtained. DHCP server may be down.\ n");
printf(" Using fallback parameters...\ n");
}
# endif
printResults();
}
void outData( char flg) {
BitWrPortI( PCDR, & PCDRShadow, flg, 4);
}
/**************************************************
* dispData( char * buf)
* the buf is a data read by readData in interrupt
* this function typically used as debug
*************************************************/
void dispData( char * buf){
static int i;
printf(" flg=% d\ n", flg);
for( i = 1; i <= D_ CHANNELS* D_ SIZE; i++){
printf(" % d", ( unsigned) buf[ i- 1]);
if( i% D_ CHANNELS == 0 )
printf("\ n");
}
}
/*************************************************
* setPortInit() set ports used as input or output
*
*************************************************/
void setPortInit() {
WrPortI( SPCR, & SPCRShadow, 0x80); // set port A as all inputs, p126
WrPortI( PBDDR, & PBDDRShadow, 0xff); // set port B as all outputs, p127
WrPortI( PBDR, & PBDRShadow, 0xff); // set PB7- PB0= 1
WrPortI( PCDR, & PCDRShadow, 0xea); // set PC0= 0, PC2= 0, PC4= 0, others=" 1", for TXD use 7407
54/ 72
BitWrPortI( PDDDR, & PDDDRShadow, 1, 4); // set PD4 as output
BitWrPortI( PDDDR, & PDDDRShadow, 1, 5); // set PD5 as output
WrPortI( PDDCR, & PDDCRShadow, 0x00); // set port D as standard output, not open drain
// BitWrPortI( PDDCR, & PDDCRShadow, 0, 4); // set PD4 as standard output, not open drain
// BitWrPortI( PDDCR, & PDDCRShadow, 0, 5); // set PD5 as standard output, not open drain
BitWrPortI( PDDR, & PDDRShadow, 1, 4); // set PD4= 1, enable U3B work
BitWrPortI( PDDR, & PDDRShadow, 1, 5); // set PD5= 1, enable U3A work
WrPortI( PEDR, & PEDRShadow, 0x00); // set port E outputs PE7- PE1 = 0
WrPortI( PEDDR, & PEDDRShadow, 0xfe); // set PE7- PE1 as outputs, PE0 as inputs for the interrupt
WrPortI( PECR, & PECRShadow, 0x00); // set PECR all bits as 0, for all outputs are clocked
WrPortI( PFDDR, & PFDDRShadow, 0xff); // set port F as all outputs
WrPortI( PFDCR, & PFDCRShadow, 0x00); // set port F as all standard output, not open drain
WrPortI( PFDR, & PFDRShadow, 0x01); // set PF7- PF1= 0, PF0= 1, but PF7 is used as PWM
WrPortI( PFFR, & PFFRShadow, 0x80); // set PF7= PWM
WrPortI( PGDDR, & PGDDRShadow, 0xff); // set port G as all outputs
WrPortI( PGDR, & PGDRShadow, 0xff); // set all output of port G as " 1", PG0- PG7= 1
BitWrPortI( PFDR, & PFDRShadow, 1, 0); // set INC( CLKD) of XDCP to 1
BitWrPortI( PCDR, & PCDRShadow, 1, 0); // set PC0= 1, U/ D( TXD) of XDCP = " 1", count up
WrPortI( PGDR, & PGDRShadow, 0xff); // set all output of port G as " 1", PG0- PG7= 1
WrPortI( PFDR, & PFDRShadow, RdPortI( PFDR) | 0x7e); // set PF6- PF1= 1, select all 6 hct373
WrPortI( PFDR, & PFDRShadow, RdPortI( PFDR) & 0x81); // set PF6- PF1= 0, set all 48 CS of DCPs to 1
WrPortI( PGDR, & PGDRShadow, 0x00); // set all output of port G as " 0", PG0- PG7= 0
BitWrPortI( PCDR, & PCDRShadow, 1, 2); // set PC2= 1
BitWrPortI( PCDR, & PCDRShadow, 0, 2); // set PC2= 0, set all outputs of U12 " 0", ( trolley control)
U12Shadow = 0; // initialize the shadow of U12
}
/**********************************************
* initializing the interrupter
* set the external INT0
* and interrupter handler as readData()
**********************************************/
void setInterrupt() {
// enable external INT0 on PE0, set the interrupt routin to readData
SetVectExtern3000( 0, readData);
// refer to p_ 95 00001001
WrPortI( I0CR, & I0CRShadow, 0x05); // enable external INT0 on PE0,
// falling edge, priority 1
WrPortI( I1CR, & I1CRShadow, 0x00); // disable external INT1 on PE1,
//
data_ array_ ptr = data0; // initialize the pointer pointting to data0 array
flg = 0; // initialize the flg
tcpDataSendFlg = 0; // initialize the flag for indicating sending the data
}
/**********************************************
* initializing PWM generator
* set the PWM frequency and the duty cycle
* define the output pin
**********************************************/
void setPwm( int frequency){
unsigned long freq;
int pwm_ options;
// request frequency PWM cycle ( will select closest possible value)
55/ 72
freq = pwm_ init( frequency);
printf(" Actual PWM frequency = % lu Hz\ n", freq);
pwm_ options = 0;
pwm_ set( 3, 0.1 * 1024, pwm_ options); // set PF7 as the PWM output port with 10% duration
}
/**********************************************
* initializing Data storage area
* define the memory area to store the data
**********************************************/
void setDataInit() {
int i;
/*
for( i= 0; i< D_ SIZE* D_ CHANNELS; i+= 3){
data0[ i]= 0xff;
data0[ i+ 1]= 0xff;
data0[ i+ 2]= 0xff;
data1[ i]= 0x00;
data1[ i+ 1]= 0x00;
data1[ i+ 2]= 0x00;
}
*/
memset( data0, 0x55, D_ SIZE* D_ CHANNELS+ 1);
memset( data1, 0xff, D_ SIZE* D_ CHANNELS+ 1);
send_ num = 0x31;
count = 0;
flg = 0;
sequence1 = 0;
}
/****** interrupt routines *********************************************
*
* readData(),
*
* use PE0 as INT. the data will be read from the input port
*
* Data read from PORT A, restore in the data0[ D_ CHANNELS* D_ SIZE] when flg= 0
* data1[ D_ CHANNELS* D_ SIZE] when flg= 1
* Each interrupt reads the PORT A for D_ CHANNELS times to get 48 channels from
* input data buffer. the PORT B is used to select which 8 channels
* will be read:
*
* for 0- 7 channels, set PB2 as 0 ( left 1- 8)
* 8- 15 channels, set PB3 as 0 ( left 9- 16)
* 16- 23 channels, set PB4 as 0 ( left 17- 24)
*
* 24- 31 channels, set PB5 as 0 ( right 1- 8)
* 32- 39 channels, set PB6 as 0 ( right 9- 16)
* 40- 47 channels, set PB7 as 0 ( right 17- 24)
*
************************************************************************/
nodebug root interrupt void readData() {
intflag= 1;
// read data from PORT A and store them into data array ( data0[] or data1[] )
// outData( 0);
# asm
56/ 72
; select left side 1- 8 channels
ioi ld a, ( PBDR)
or 0xfc ; set PB7- PB2 = " 1", not select all 6 hc244
and 0xfb ; set PB2 to 0, select hct244 for left side channel 1- 8
ioi ld ( PBDR), a
ld hl, ( data_ array_ ptr) ; get the address of the data array working at moment
; dataa = A7 A6 A5 A4 A3 A2 A1 A0
; channel: 8 7 6 5 4 3 2 1
ioi ld a, ( PADR) ; get the data from the port A
ld ( hl), a ; store the data to a unit in data array
ld de, ( testdata_ left_ ptr) ; get the address of testdata_ left[ 0]
ld ( de), a ; save the data to testdata_ left[ 0]
; select left side 9- 16 channels
ioi ld a, ( PBDR)
or 0xfc ; set PB7- PB2 = " 1", not select all 6 hc244
and 0xf7 ; set PB3 to 0, select hct244 for left side channel 9- 16
ioi ld ( PBDR), a
inc hl ; get the address of next unit in data array
; dataa = A7 A6 A5 A4 A3 A2 A1 A0
; channel: 16 15 14 13 12 11 10 9
ioi ld a, ( PADR) ; get the data from the port A
ld ( hl), a ; store the data to the next unit in data0[]
; ld de, ( testdata_ left_ ptr+ 1) ; get the address of testdata_ left[ 1]
inc de
ld ( de), a ; save the data to testdata_ left[ 1]
; select left side 17- 24 channels
ioi ld a, ( PBDR)
or 0xfc ; set PB7- PB2 = " 1", not select all 6 hc244
and 0xef ; set PB4 to 0, select hct244 for left side channel 17- 24
ioi ld ( PBDR), a
inc hl ; get the address of next unit in data array
; dataa = A7 A6 A5 A4 A3 A2 A1 A0
; channel: 24 23 22 21 20 19 18 17
ioi ld a, ( PADR) ; get the data from the port A
ld ( hl), a ; store the data to the next unit in data0[]
; ld de, ( testdata_ left_ ptr+ 2) ; get the address of testdata_ left[ 2]
inc de
ld ( de), a ; save the data to testdata_ left[ 2]
; select right side 1- 8 channels
ioi ld a, ( PBDR)
or 0xfc ; set PB7- PB2 = " 1", not select all 6 hc244
and 0xdf ; set PB5 to 0, select hct244 for right side channel 1- 8
ioi ld ( PBDR), a
inc hl ; get the address of next unit data array
; dataa = A7 A6 A5 A4 A3 A2 A1 A0
; channel: 8 7 6 5 4 3 2 1
57/ 72
ioi ld a, ( PADR) ; get the data from the port A
ld ( hl), a ; store the data to the next unit in data0[]
ld de, ( testdata_ right_ ptr) ; get the address of testdata_ right[ 0]
ld ( de), a ; save the data to testdata_ right[ 0]
; select right side 9- 16 channels
ioi ld a, ( PBDR)
or 0xfc ; set PB7- PB2 = " 1", not select all 6 hc244
and 0xbf ; set PB6 to 0, select hct244 for right side channel 9- 16
ioi ld ( PBDR), a
inc hl ; get the address of next unit in data array
; dataa = A7 A6 A5 A4 A3 A2 A1 A0
; channel: 16 15 14 13 12 11 10 9
ioi ld a, ( PADR) ; get the data from the port A
ld ( hl), a ; store the data to the next unit in data0[]
; ld de, ( testdata_ right_ ptr+ 1) ; get the address of testdata_ right[ 1]
inc de
ld ( de), a ; save the data to testdata_ right[ 1]
; select right side 17- 24 channels
ioi ld a, ( PBDR)
or 0xfc ; set PB7- PB2 = " 1", not select all 6 hc244
and 0x7f ; set PB7 to 0, select hct244 for right side channel 17- 24
ioi ld ( PBDR), a
inc hl ; get the address of next unit in data array
; dataa = A7 A6 A5 A4 A3 A2 A1 A0
; channel: 24 23 22 21 20 19 18 17
ioi ld a, ( PADR) ; get the data from the port A
ld ( hl), a ; store the data to the next unit in data0[]
; ld de, ( testdata_ right_ ptr+ 2) ; get the address of testdata_ right[ 2]
inc de
ld ( de), a ; save the data to testdata_ right[ 2]
inc hl
ld ( data_ array_ ptr), hl ; store the position pointer
ioi ld a, ( PBDR)
or 0xfc ; set PB7- PB2 = " 1", not select all 6 hc244
ioi ld ( PBDR), a ;
# endasm
count += D_ CHANNELS;
if( count >= D_ CHANNELS* D_ SIZE){
sequence1++;
count = 0;
tcpDataSendFlg = 1; // set flag to indicate the data is ready to be sent
if( flg){
flg = 0;
data_ array_ ptr = data0;
}
else{
flg = 1;
data_ array_ ptr = data1;
}
}
58/ 72
}
/***************************************************
* Print some of the DHCP parameters received.
***************************************************/
static void printResults( void) {
auto long tz;
auto word i;
printf(" Network Parameters:\ n");
printf(" My IP Address = % 08lX\ n", my_ ip_ addr);
printf(" Netmask = % 08lX\ n", sin_ mask);
# if TCPCONFIG== 3
if (_ dhcphost != ~ 0UL) {
if (_ dhcpstate == DHCP_ ST_ PERMANENT) {
printf(" Permanent lease\ n");
} else {
printf(" Remaining lease = % ld ( sec)\ n", _ dhcplife - SEC_ TIMER);
printf(" Renew lease in % ld ( sec)\ n", _ dhcpt1 - SEC_ TIMER);
}
printf(" DHCP server = % 08lX\ n", _ dhcphost);
}
# endif
if ( gethostname( NULL, 0))
printf(" Host name = % s\ n", gethostname( NULL, 0));
if ( getdomainname( NULL, 0))
printf(" Domain name = % s\ n", getdomainname( NULL, 0));
# if TCPCONFIG== 3
if ( dhcp_ get_ timezone(& tz))
printf(" Timezone ( fallback only) = % ldh\ n", tz / 3600);
else
printf(" Timezone ( DHCP server) = % ldh\ n", tz / 3600);
# endif
for ( i = 0; i < *_ last_ nameserver; i++)
printf(" DNS server #% u = % 08lX\ n", i+ 1, def_ nameservers[ i]);
ip_ print_ ifs();
router_ printall();
}
/*******************************************************************
* int sendPacket( char * buf)
*
* Send data to the remote Data. the zise of package is 601.
* It should consider the situation:
* that the data can't send out because the send- buffer is full.
*
*
*******************************************************************/
int sendPacket( char * buf) {
static long sequence;
auto int length, retval;
auto int ltlen;
static char buf_ last[ D_ SIZE* D_ CHANNELS];
char zero_ buf[ D_ SIZE* D_ CHANNELS];
# GLOBAL_ INIT
{
sequence = 0;
}
memset( zero_ buf, 0xFF, D_ SIZE* D_ CHANNELS);
sequence++;
length = D_ SIZE* D_ CHANNELS;
59/ 72
/*
if( strncmp( buf, zero_ buf, length)== 0)
if( strncmp( buf, buf_ last, length)== 0)
return 1;
*/
sock_ flushnext(& tsock);
tcp_ tick(& tsock);
/* there is enough room in the buffer
* before adding data with a blocking function.*/
if( sock_ tbleft(& tsock) < length) {
sock_ abort(& tsock);
return 0;
}
retval = sock_ awrite(& tsock, buf, length);
// printf(" sendout:% d\ n", retval);
switch ( retval) {
case 0:
printf(" Insufficient free space in transmit buffer!\ n");
break;
case 1:
printf(" Len is greater than the total socket receive buffer size!\ n");
break;
case 2:
printf(" The socket has been closed for further transmission!\ n");
break;
case 3:
printf(" len< 0 or parameter wa invalid!\ n");
break;
}
sock_ flush(& tsock);
// strncpy( buf_ last, buf, length);
return 1;
}
/*************************************
* int openTcpSock()
*
* open a tcp socket
* return 1 successfully
* return - 1 unable to open TCP session
*************************************/
int openTcpSock(){
// printf(" seq:% d\ n", sequence1);
sock_ abort(& tsock);
// printf(" try to open socket ... \ n");
if ( ! tcp_ open( & tsock, 0, resolve( REMOTE_ IP), REMOTE_ PORT , NULL )) {
puts(" Unable to open TCP session!\ n");
return - 1;
}
else if ( sock_ established(& tsock)) {
printf(" Established OK!\ n");
return 1;
}
else{
// printf(" Failed to establish a socket!\ n");
return - 1;
}
}
/*************************************
* void setXDCP()
60/ 72
*
* initialize all the XDCP
*************************************/
void setXDCP() {
int i;
char value;
WrPortI( PGDR, & PGDRShadow, 0x00); // set PG7- PG0 = 00000000 ( select odd 24 DCPs, 1,3,5... 47 )
value = RdPortI( PFDR) | 0x7e;
WrPortI( PFDR, & PFDRShadow, value); // set PF6- PF1= 1, select all 6 hct373
value = RdPortI( PFDR) & 0x81;
WrPortI( PFDR, & PFDRShadow, value); // set PF6- PF1= 0, set odd 24 CS of DCPs of total 48 DCPs to 0
// ( using AD5222 )
BitWrPortI( PCDR, & PCDRShadow, 1, 0); // set PC0= 1, U/ D( TXD) of XDCP to count up
for( i= 0; i< 130; i++) {
BitWrPortI( PFDR, & PFDRShadow, 0, 0); // set INC( CLKD) of XDCP to 0
BitWrPortI( PFDR, & PFDRShadow, 1, 0); // set INC( CLKD) of XDCP to 1
} // set XDCP to top
BitWrPortI( PCDR, & PCDRShadow, 0, 0); // set PC0= 0, U/ D( CLKD) of XDCP to count down
for( i= 0; i< 77; i++) {
BitWrPortI( PFDR, & PFDRShadow, 0, 0); // set INC( CLKD) of XDCP to 0
BitWrPortI( PFDR, & PFDRShadow, 1, 0); // set INC( CLKD) of XDCP to 1
} // set XDCP to midway, Vw= 2.0V
WrPortI( PGDR, & PGDRShadow, 0xaa); // set PG7- PG0 = 10101010 ( select even 24 DCPs, 2,4,6... 48 )
value = RdPortI( PFDR) | 0x7e;
WrPortI( PFDR, & PFDRShadow, value); // set PF6- PF1= 1, select all 6 hct373
value = RdPortI( PFDR) & 0x81;
WrPortI( PFDR, & PFDRShadow, value); // set PF6- PF1= 0, set even 24 CS of DCPs of total 48 DCPs to 0
// ( using AD5222 )
BitWrPortI( PCDR, & PCDRShadow, 1, 0); // set PC0= 1, U/ D( TXD) of XDCP to count up
for( i= 0; i< 130; i++) {
BitWrPortI( PFDR, & PFDRShadow, 0, 0); // set INC( CLKD) of XDCP to 0
BitWrPortI( PFDR, & PFDRShadow, 1, 0); // set INC( CLKD) of XDCP to 1
} // set XDCP to top
BitWrPortI( PCDR, & PCDRShadow, 0, 0); // set PC0= 0, U/ D( CLKD) of XDCP to count down
for( i= 0; i< 77; i++) {
BitWrPortI( PFDR, & PFDRShadow, 0, 0); // set INC( CLKD) of XDCP to 0
BitWrPortI( PFDR, & PFDRShadow, 1, 0); // set INC( CLKD) of XDCP to 1
} // set XDCP to midway, Vw= 2.0V
BitWrPortI( PCDR, & PCDRShadow, 1, 0); // set PC0= 1, U/ D( TXD) of XDCP to count up
WrPortI( PGDR, & PGDRShadow, 0xff); // set PG7- PG0 = 11111111
value = RdPortI( PFDR) | 0x7e;
WrPortI( PFDR, & PFDRShadow, value); // set PF6- PF1= 1, select all 6 hct373
value = RdPortI( PFDR) & 0x81;
WrPortI( PFDR, & PFDRShadow, value); // set PF6- PF1= 0, set all 24 CS of 24 DCP chips to 1
}
/**************************************************
* void calibration()
*
* calibrate all the XDCP to optimal position
*
**************************************************/
void calibration() {
char Step_ counter; // step counter for XDCP
char XDCP_ max[ 48], XDCP_ min[ 48], XDCP_ set[ 48]; // 0- 23 for left side, 24- 47 for right side, (_ max & _ min
// for maximum and minimum step count, _ set for setting of XDCP )
char XDCPSetDone_ max[ 48], XDCPSetDone_ min[ 48]; // store the symbol of the channel which has been set
61/ 72
char samplingData[ 12][ 6]; // store the sample data of 10- 12 times, [][ 0]-[][ 2] for left side, [][ 3]-[][ 5] for
right side
// char leftXDCP_ max[ 8], leftXDCP_ min[ 8]; // store maximum and minimum step count for left XDCP
// char rightXDCP_ max[ 8], rightXDCP_ min[ 8]; // store maximum and minimum step count for right XDCP
// char leftXDCPset[ 8], rightXDCPset[ 8]; // save the setting of XDCP
char temp1, temp2, temp3;
int i, j, symbol;
char value;
memset( samplingData, 0x00, 12* 6);
BitWrPortI( PBDR, & PBDRShadow, 0, 0); // set PB0= 0, turn on the LED indicator
WrPortI( PGDR, & PGDRShadow, 0x00); // set PG7- PG0 = 00000000 ( select odd 24 DCPs, using AD5222)
value = RdPortI( PFDR) | 0x7e;
WrPortI( PFDR, & PFDRShadow, value); // set PF6- PF1= 1, select all 6 hct373
value = RdPortI( PFDR) & 0x81;
WrPortI( PFDR, & PFDRShadow, value); // set PF6- PF1= 0, set 24 odd CS of all 48 DCPs to 0
BitWrPortI( PFDR, & PFDRShadow, 1, 0); // set INC( CLKD) of XDCP to 1
BitWrPortI( PCDR, & PCDRShadow, 1, 0); // set PC0= 1, U/ D( TXD) of XDCP to " 1", count up
for( i= 0; i< 130; i++) {
BitWrPortI( PFDR, & PFDRShadow, 0, 0); // set INC( CLKD) of XDCP to 0
BitWrPortI( PFDR, & PFDRShadow, 1, 0); // set INC of XDCP to " 1", genarate a negtive pulse on INC
line
} // set XDCP to top
BitWrPortI( PCDR, & PCDRShadow, 0, 0); // set PC0= 0, U/ D( TXD) of XDCP to " 0", count down
for( i= 0; i< 8; i++) {
BitWrPortI( PFDR, & PFDRShadow, 0, 0); // set INC of XDCP to 0
BitWrPortI( PFDR, & PFDRShadow, 1, 0); // set INC of XDCP to 1, genarate a negtive pulse on INC line
} // set XDCP to step 120
WrPortI( PGDR, & PGDRShadow, 0xaa); // set PG7- PG0 = 10101010 ( select even 24 DCPs, using AD5222)
value = RdPortI( PFDR) | 0x7e;
WrPortI( PFDR, & PFDRShadow, value); // set PF6- PF1= 1, select all 6 hct373
value = RdPortI( PFDR) & 0x81;
WrPortI( PFDR, & PFDRShadow, value); // set PF6- PF1= 0, set 24 even CS of all 48 DCPs to 0
BitWrPortI( PFDR, & PFDRShadow, 1, 0); // set INC( CLKD) of XDCP to 1
BitWrPortI( PCDR, & PCDRShadow, 1, 0); // set PC0= 1, U/ D( TXD) of XDCP to " 1", count up
for( i= 0; i< 130; i++) {
BitWrPortI( PFDR, & PFDRShadow, 0, 0); // set INC( CLKD) of XDCP to 0
BitWrPortI( PFDR, & PFDRShadow, 1, 0); // set INC of XDCP to " 1", genarate a negtive pulse on INC
line
} // set XDCP to top
BitWrPortI( PCDR, & PCDRShadow, 0, 0); // set PC0= 0, U/ D( TXD) of XDCP to " 0", count down
for( i= 0; i< 8; i++) {
BitWrPortI( PFDR, & PFDRShadow, 0, 0); // set INC of XDCP to 0
BitWrPortI( PFDR, & PFDRShadow, 1, 0); // set INC of XDCP to 1, genarate a negtive pulse on INC line
} // set XDCP to step 120
BitWrPortI( PDDR, & PDDRShadow, 0, 4); // set PD4= 0, turn off right laser
BitWrPortI( PDDR, & PDDRShadow, 0, 5); // set PD5= 0, turn off left laser
while( 1) {
costate {
waitfor( DelayMs( 5L) ); // wait for 5ms
break;
}
}
BitWrPortI( PDDR, & PDDRShadow, 1, 4); // set PD4= 1, turn on right laser
62/ 72
BitWrPortI( PDDR, & PDDRShadow, 1, 5); // set PD5= 1, turn on left laser
Step_ counter= 120;
for( i= 0; i< 48; i++) {
XDCP_ max[ i] = 0;
XDCP_ min[ i] = 0;
XDCP_ set[ i] = 0;
XDCPSetDone_ max[ i] = 0;
XDCPSetDone_ min[ i] = 0;
}
for( i= 0; i< 12; i++)
for( j= 0; j< 6; j++)
samplingData[ i][ j] = 0; // set 2D sampling Data array to 0
intflag= 0;
while( 1) {
for( i= 0; i< 10; i++) {
for( ; ; ) {
if( intflag== 1) {
intflag= 0;
break;
}
}
samplingData[ i][ 0]= testdata_ left[ 0]; // store the sampling data of left side
samplingData[ i][ 1]= testdata_ left[ 1];
samplingData[ i][ 2]= testdata_ left[ 2];
samplingData[ i][ 3]= testdata_ right[ 0]; // store the sampling data of right s