Coil Gun with Targeting System
Group 8
Brian Hoehn
Kwok Ng
Ricardo Reid
Josef Von Niederhausern
EEL 4915
Fall 2010
Table of Contents
Coil Gun Executive Summary Overview ........................................................... 1
Similar Projects ........................................................................................................ 2
Field Generation Subsystem ............................................................................. 3
Projectile ................................................................................................................... 4
Barrel ......................................................................................................................... 6
Rifling ........................................................................................................................ 7
Force ......................................................................................................................... 8
Wire size selection...................................................................................................19
Magnetic Wire........................................................................................................................ 20
Heat Study ............................................................................................................................. 21
Current Pulse Generation .......................................................................................23
External Iron.............................................................................................................24
Velocity Detection ...................................................................................................25
Testing......................................................................................................................29
Heat Dissipation .................................................................................................................. 33
Magnetic Circuit ..................................................................................................................... 34
Switching System ....................................................................................................36
Design ................................................................................................................................... 39
Bleed Resistor ....................................................................................................................... 40
Damping Resistor .................................................................................................................. 42
Digital Voltmeter ......................................................................................................43
Triggering System ...................................................................................................43
Motion Control and SensorSubsystem .......................................................... 46
Motion Control .........................................................................................................47
DC Motor Vs AC Motor .......................................................................................................... 48
Stepper Parameters .............................................................................................................. 53
DC step motor controller ....................................................................................................... 56
Real Time Control ................................................................................................................. 63
DC Motor Controllers ............................................................................................................. 64
Sensor ......................................................................................................................67
FOV request for this camera ................................................................................................. 67
Power supply for camera and motors ................................................................................... 69
Camera Position Configuration ............................................................................................. 69
Image collection and store .................................................................................................... 70
Movement Algorithm ............................................................................................................. 71
Motion Vector Determination ................................................................................................. 72
VIDEO DECODER ................................................................................................................ 74
Power system ................................................................................................... 76
Power System Overview .........................................................................................76
Power Supply ...........................................................................................................77
Energy Sources .......................................................................................................78
Solar Cells and Panels ............................................................................................80
AC to DC Power .......................................................................................................81
Stepping up AC vs. DC Power ................................................................................83
Regulators Comparison ..........................................................................................84
Linear Regulator .................................................................................................................... 84
Boost Regulators ................................................................................................................... 85
Buck Regulators .................................................................................................................... 86
Capacitor Charging Source ....................................................................................87
Digital Voltmeter ......................................................................................................90
Digital Thermometer ................................................................................................90
Testing Current ........................................................................................................91
Testing Voltage ........................................................................................................92
Circuitry Protection .................................................................................................94
Controls and SoftwareSubsystem .................................................................. 95
Position Determination ...........................................................................................96
Software Architecture .............................................................................................98
Microcontroller ......................................................................................................102
Programming the Microcontroller ........................................................................................ 105
Choosing a Microcontroller .................................................................................................. 105
PIC Microcontrollers ............................................................................................................ 108
TI MSP430........................................................................................................................... 110
FPGA ......................................................................................................................112
FPGA Design Process ........................................................................................................ 113
Editing, Compiling, and Synthesizing: .................................................................113
Choosing an FPGA ............................................................................................................. 114
User Interface.................................................................................................. 115
Controls and Software Budget .............................................................................118
Executive Summery ....................................................................................... 123
Partial Circuit Schematic Overview ......................................................................124
Circuit Schematics ................................................................................................125
Milestones ..............................................................................................................127
Coil Gun Executive Summary Overview
A pulsed linear induction motor commonly known as a coil gun is not considered
an efficient system. Some enthusiast has achieved efficiencies of 10% but most
range from one to two percent energy transfer. This project was initially
motivated by a similar project found on YouTube. This was a high voltage self
targeting coil gun designed by students at a University in California. Their project
was able to track a laser pointer deliver a projectile using 400 volts. Additionally,
their project was able to take inputs for the color designation of a target and
automatically move to and fire.
Additional motivation for building a coil gun was based on power systems. One
of this group members is fascinated by power systems and felt this would be a
challenging project. Another motivating factor was a fascination of group
members with electromagnetic field. Additionally, several members of the group
are fascinated with projectile weapons and the development and research of
such. Group members who were not as motivated in the electromagnetic or
power system aspect of this project were able to find subsystems which allowed
them to research and design aspects which were similar to other projects
considered by this group.
From a firearm standpoint current technology does not allow for a military use of
coil guns. One of the immediate benefits is the ability of a coil gun to fire a
projectile relatively silent compared to conventional firearm. One of the major
negative aspects of a coil guns when used military or law enforcement is the
heavy cumbersome load needed to generate the current source. The weight
necessary to accommodate the current source, a capacitor bank or battery,
reduces mobility driving a mounted implementation. Even if the current source
could be mitigated the strength of a coil is proportional to the number of turns
which tends to create a heavier unbalanced barrel assembly. This project will
use a mounted coil gun to achieve the maxim velocity possible. At the time of
this paper the goal was to achieve 100 feet per seconds or approximately 31
meters per second. The design will allow for either manual or automatic firing
modes. In the automatic firing mode, the coil gun will be able to optically track a
target and fire.
For design purposes the coil gun system was broken down into four functional
subsystems; Field Generation Subsystem, Power Subsystems, Sensors and
Motion Subsystem, and Controls and Software Subsystem. Each subsystem was
managed by one of the group members as shown in the block diagram Figure 1.
In this paper the subsystem will contain the research, design, and test specific to
1
that subsystem. At the end of the paper a total system test will be given. This
approach allowed each member to work on their respective subsystem and to
compile the paper in a modular manner. An administrator does not mean that
one person researched and designed the complete subsystem.
Controls and
Software
Subsystem:
Microcontroller,
FPGA, Firing
Circuit, Software
Architecture,
User Interface
Administrator:
Brian Hoehn
Sensor and Motion
Subsystem:
Camera, Video
Convertor, Image
Analysis Motor, Motor
Controller , Tracking
Algorithm
Administrator:
Kwok Ng
Field Generation
Subsystem:
Barrel, Solenoid,
Projectile, Capacitor
Bank, Magnetic wire,
SCR
Administrator:
Josef Von
Niederhausern
Power Subsystem:
Power Sources, Energy Sources, Protection Circuitry, Capacitor Bank
Charging, Heat Dissipation, Current & Voltage Testing, Switching Systems
Administrator:
Ricardo Reid
Figure 1 Automated Coil Gun subsystems
Similar Projects
The a search of the University of Central Florida School of Electrical Engineering
and Computer Science web site revealed two previous coil gun projects. The
first group was Group 8 during the Summer/Fall semesters of 2007. Group 8’s
project was focused on a high arcing projectile, a mortar application of a coil gun,
to achieve a given accuracy.
Duringthe Fall2008/Spring2009 semesters Group 13 tackled the coil gun project.
The primary propose of this groups project was to create a “high powered mass
accelerator” which would accept user inputs. Group 13’s project is similar to this
in that both projects are attempting to deliver a high power punch with
2
theprojectile.A major difference between this project and Group 13’s is an
automated optical targeting system.
Field Generation Subsystem
The Field Generation subsystem primary purpose is to take a large current pulse
of short duration and convert it to an electromagnetic field. The primary principle
behind the field generation subsystem is to transfer as much energy from the
energy storage, capacitor bank or battery, to the projectile. Unfortunately coil
guns are inherently inefficient which complicates this task. The field generation
subsystem was looked at from an analytical view. Matlab Mfiles were generated
to calculate the field strength and can be found in Appendix C.
Early on in the project decision-making process, consideration was given to the
number of solenoids used. Multiple solenoids trigged at precision timing has
advantages when trying to achieve high velocities. The timing of each solenoid
could be triggered by an optical sensor place inside the barrel with a variable
coefficient which could be manipulated for optimal acceleration. However, it was
determined shortly after the project choice than one solenoid should be used to
reduce the efforts of timing a series of solenoids. Since only one solenoid will be
used higher efficiency is a priority in this project. This may prove to be an added
difficulty in the project as coil guns are not known for their efficiency. A two
percent energy transfer is considered a good efficiency for coil guns by the
hobbyist community.
In order to maximize force applied in the forward direction, the current pulse
needs to be finished when the projectile has traversed one half of the coil. If the
projectile passes the half way point and the current is still flowing the field exerts
a force in the opposite direction resulting in a reduce muzzle velocity known as
‘Suck Back.’ The field generation subsystem internal and inter subsystem
interface is show in block diagram Figure 2.
3
Controls and
Software
Subsystem
Power
Subsystem
Sensor and
Motion Control
Subsystem
Trigger/
SCR
Capacitor
Bank
Barrel
assembly/inductor
Projectile
acceleration
Projectile velocity
detection
Figure 2 Field generation Block Diagram
Projectile
The first step in building a coil gun is to select a projectile. Several shapes and
packages were considered for the projectile. Since velocity is a primary design
choice in this project, with the benefits of low friction and light mass a spherical or
BB shape would have been the intuitive choice for a projectile. The lighter mass
would be impacted much greater by the same force on a heavier mass. Low
friction would reduce the amount of energy wasted to heat and noise.
However, further inspection of the magnetic system used to propel the projectile
the force applied to the projectile is directly proportional to the amount of ferrous
material. In selecting a projectile it was examined how the projectile affects the
magnetic field. Several coil gun enthusiasts have done this before.
Saturation is the point where all magnetic regions are aligned and a further
increase in the strength of the magnetic field will not be beneficial to the system.
The information for the saturation table 1 was taken from coilgun.info.com.
Supermalloy would be an optimal projectile material but the price and availability
is beyond the scope of this project.
4
Material
Permeability
Maximum
At B = 20 Permeability
Gauss
rolled 180
2000
Cold
Steel
Iron
Purified iron
Supermalloy
200
5000
100000
5000
180000
800000
Saturation Flux
density
B gauss
21000
21500
21500
8000
Table 1 Saturation Density
Iron or steel rods are also available which will fit better in the barrel with much
less air gaps. These rods would provide the opportunity to vary the length of the
projectile for optimization. However, currently a quarter inch number 2 Philips
head bit one inch long is selected as the projectile. This projectile fits within the
barrel but leaves some air gaps which translate into efficiency loss. Currently,
the projectile is made from cold rolled steel due to the availability, consistent
mass, length, and diameter. The Phillips head bits are readily available at any
local hardware store. Further research of the projectile found that Barry Hansen
had simulated the variation of the projectile to coil ratio shown as figure 3. It was
found that the maximum work done was when the length of the projectile was 75%
of the coil length. During the fabrication of this project the iron dowels from
Lowes will be tested from lengths of 20mm to 5 mm to pick an optimum projectile.
However, it is readily apparent that a spherical projectile would not be an efficient
vehicle for energy transfer. In addition to the Phillips head bits a length of 16
gauge cold rolled steel was purchased. This will allow the variance of projectile
length during the prototype phase.
5
Figure 3 Variation of Projectile
（Reprinted with Permission of Barry Hansen)
It was considered to build a Mfile which would simulate the variance of the
projectile while holding other things constant. However, this turned out to be a
more difficult analysis than expected and seems to be beyond the scope of this
project. As a result, it was determined this would be better done during the
prototype phase.
Barrel
The barrel choices were limited to what is available at local home improvements
stores. When selecting a barrel several considerations should be taken into
account. Permeability (μr) will allow the field to penetrate the barrel and engage
the projectile. For non ferrous materials μr is approximately equal to free space
(μ0). Low coefficient of friction (COF) will allow the projectile to be guided by the
barrel with minimal energy loss. A brass barrel would have a low COF and since
brass is not ferrous its permeability is not an issue.
6
A barrel made of brass would provide a durable material which would increase
total shots possible per barrel. The projectile path will be much more controlled
in a rugged metallic barrel. However, the metallic conduction properties would
increase a risk factor for safety issues. This could be mitigated with an insulator
between the solenoid and the barrel. An insulator would increase the inner
diameter of the solenoid and decrease the force exerted on the projectile.
Another aspect of the barrel is eddy currents. Eddy currents are circulating
currents in a conductor which will induce a magnetic field opposite to the original
field. Obviously, eddy currents will reduce the velocity of the projectile, which is a
major design point of this project. Slotting, or notching, the barrel will reduce the
amount of eddy currents.
Since this project is uninsured, maximum safety is a major design parameter.
Safety and accessibility has driven the selection of a non-conducting barrel.
Currently a quarter inch, or 8 mm, inner diameter polyvinyl chloride (pvc)
commonly known as PVC pipe which is commonly used for refrigerator water
lines, for ice makers and cold water, has been selected available at Home Depot
or Lowes. The outer diameter of this barrel is 13 mm and will be the inner
diameter of the driving solenoid. In an effort to increase consistency a rigged
support structure could be implemented to reduce barrel vibrations induced by
the projectile. This can be a dowel or forward support structure.
Rifling
To give the coil gun the most accuracy as possible we have decided to look in to
the possibility of rifling the barrel for which the projectile will travel through.
Rifling is the idea creating grooves throughout the inside of the barrel. The
grooves in the barrel force the projectile to spin. This spin stabilizes the projectile
improving its aerodynamics and accuracy.
Rifling is measured by the twists per inch, i.e. one turn in ten inches (1:10). To
increase the speed of the rifling you would shorten the distance per twist, while
increasing the distance per twist slow the speed of the rifling. The ‘speed’ of the
rifling refers to the rotation of the projectile through the rifling not the actual
velocity of the projectile. To determine the speed of the rifling, or twist rate, need
for a specific projectile the shape, length and weight will all be taken into account.
In most cases slower twist rates are used for projectiles that are short and have a
larger diameter, and faster twist rates are used for projectiles that are long and
have a small diameter. Since the projectile we will most likely being using is a
drill bit, which is long and has a small diameter, we will use a twist rate of 1:10 or
7
faster. To calculate a more precise twist rate we can use the Greenhill Formula
that was developed by George Greenhill, in 1879, to do just that. The formula
follows:
CD 2 SG
Twist 
L 10.9
Where:




C = 150 (use 180 for muzzle velocities higher than 2,800 f/s)
D = bullet's diameter in inches
L = bullet's length in inches
SG = bullet's specific gravity (10.9 for lead-core bullets, which cancels out
the second half of the equation)
There is also a formula to calculate the projectile’s revolutions per minute due to
the rifling. The formula to calculate the rotational speed of the projectile is MV(in
fps) x (12/Twist rate in inches) x 60.
Other than creating the rotation of the projectile, the rifling has other needs that
need to be addressed. We want the rifling to be the correct size so that projectile
will ‘swage’ when fired. We don’t want the diameter of the rifling to flux,
consistency is a must. The spacing of the twist and groove width should also be
uniform throughout the barrel.
There any many methods to manufacture rifling in a barrel. The two techniques
we will most likely use either the cut rifling method or the broached rifling method.
The cut rifling, also known as single point cut rifling, is done by using a machine
tool to cut one groove at a time. Instead of cutting on groove at time, all of the
grooves can be cut in one motion using a progressive broach bit; this would be
known as the broached rifling method.
Due to the cost and unavailability of rifling a barrel or purchasing a pre-rifled
barrel the idea of choosing to rifle the barrel of the coil gun seems to be a little
unrealistic, and we will only be done if circumstances allow for it.
Force
The first step is to calculate the amount of force needed to move the 5 grams 31
m/s. Using an acceleration distance of 2 inches or 0.0508 meters and the end
velocity of 31 meters per second, the acceleration period is calculated by:
8
ta 
d 0.058

 1.63ms
v
31
The acceleration is calculated by:
a
v
31
m

 18917.3 2
t a 0.00163
s
Using Newton’s Second Law:
F  ma  (0.005)(18917.3)  94.5newtons
Converting to Joules:
J  FD  (94.5)(0.0508)  4.805 joules
Evaluation of the above process using the kinetic energy equation yields:
Ke 
1
1
mv 2  (0.005)(312 )  4.805 joules
2
2
The coefficient for Polyvinyl chloride (PVC) was not readily available so the force
calculated above will be overestimated by 20% which gives 94.5 Newton’s of
force. This may seem like a small amount but considering the poor efficiency of
coil guns this will not be a menial task. The period of acceleration will also be the
current pulse period. Using an average two percent energy transfer about 250
joules will be needed.
Consideration was given on how to create a suitable current pulse. Since the
force exerted in a magnetic field inside a hollow solenoid is directly proportional
to the current. A large pulse, 1000 amperes, is desired for one to three
milliseconds. This can be accomplished several ways. The most common is to
use a bank of capacitors. One of the major benefits of using a capacitor bank is
relatively short charge times. Additionally the capacitors will deplete quickly
which broadens the triggering switch selection.
B   0 NI  (450)(1000)  0.56teslas
B = is the magnetic field in Teslas
 0 = permeability of free space
N = the number of turns per meter
I * = is the current in amperes
9
With the minimum design found above, an analytical optimization was performed.
Analysis of a finite solenoid can be done by using equations found on
www.netdenizen.com. A diagram showing the dimensional analysis is shown in
figure 4.
Figure 4 Dimensional Coil Analyses
The general equation for magnetic field strength of a finite air filled solenoid is
given by:

 r 2 x 2 r 
 r 2  x 2  r 
2
2
2 
1
2 


B
x2 ln
 x1 ln  2




2
2
2
2
2(r2  r1 ) 
 r1  x2  r1 
 r1  x1  r1 

 0 in
This equation was used to show the field as a function of power:
B  0
P
G
r1 
Where G is the unit less geometry factor:
G

1
 G
8  2  1
1
 G2 
10
    2     2
G1     ln 
 1  1     2





    2     2
G2     ln 
 1  1     2


r2
r1

1
2r1





x1  x 2
2r1
However, while attempting to simulate the above equation in it in Matlab
erroneous inductances were exhibited resulting in the results being discarded.
This was a major setback in for optimization of the coil gun. Being able to plot
the field strength with respect to starting position would have been a beneficial
design asset. However the ratio of Alpha and Beta were said to be optimized at
3 and 2 respectively. The alpha ratio was chose first and resulted in the outer
diameter of 39 mm being selected. Early attempts at setting the Beta ratio to 2
severely reduced the amount of current running through the coil, 300 amperes.
Therefore, this design was not able to optimize the Beta ratio as will be shown
later on, and it was later calculated at 1.
Continuing with the Matlab simulation, the magnetic field found at the midpoint of
a coil is given by:
2



 r 2   l   r 

2
 2 2

 0 in 
B
x
ln


2
2
2r2  r1  


l
2
r


r




1
1

2



Or:
2



l


2


r2     r2 


 jl 
2
B  0  x2 ln 

2
2 


l
2
 r1     r1 

2



Where j = the current density.
11
All the above equations were simulated in Matlab and the following plots were
generated. The equations were written to M files and all variables were held
constant except one. This process was performed several times. When the final
coil dimensional values were selected the final iteration was captured for this
initial project documentation. Firs the length of the coil was varied as shown in
Figure 5 this was done by holding all the values constant and setting up a loop
which iteratively added one millimeter increments up to one meter. Figure 6
shows the magnetic field strength as calculated from the on axis mid point. The
unbroken line is showing the partial wire input into the field strength. This is not
particle. The program was corrected to take into account the wire gauge. A
second iteration of the M file accounted for the wire diameter and would not allow
for partial wires. This resulted in the jagged line shown below. Figure 7 shows a
complete view of the coil length variations. Even if the calculated numerical
values are incorrect the graph has merit in showing the diminishing return on field
strength. This plot was used in conjunction with Barry Hansen’s inductor
simulator and resistor, inductor, capacitor simulator to pick the nominal length of
26 millimeters.
l
x2
B
r2 r1
x1
Figure 5 Varying Coil Length
12
Figure 6 Varying Coil Length
Figure 7 Varying Coil Length
13
As the length of the coil increases the midpoint magnetic field increased. This
variable was used to optimize current given an optimal inner and outer diameter.
Holding the length constant the diameter of the coil was varied as shown in figure
8and produced the plot shown in figure 9. This was done using an iterative loop
which varied the outer diameter of the coil from the inner diameter to one meter
in 1 millimeter steps. Again the line is jagged to show the partial wire gauge
steps did not increase the number of turns. The field strength shown on the left
is measured at the center point on axis of the coil. Even if the calculated
numerical values are incorrect the graph has merit in showing the diminishing
return on field strength. Using this graph and the recommendation to set alpha to
3 the outer diameter was selected at 39 mm.
l
x2
B
r2
r1
x1
Figure 8 Varying Coils Outer Diameter
Figure 9 Varying Coil Outer Diameter
14
Then the inner diameter of the coil was varied as shown in figure 10 to produce
the plots shown in figure 11.
l
x2
B
r2
r1
x1
Figure 10 Varying Coil Inner Diameter
Figure 11 Varying Coil Inner Diameter
The final dimensional analysis was to vary the American Wire Gauge while
holding all other things constant as shown in figure 12. As can been seen there
is a significant gain as the American Wire Gauges increases, which is an actual
decrease in the diameter and an increase in the turn density. Unfortunately, the
15
current used in generating this plot, 848 amperes, would exceed the coils
capacity to be able to dissipate heat for anything larger than 20 AWG.
Figure 12 Varying American Wire Gauge (AWG)
The first equation given in this section was used to vary the starting position
shown as figure 13 of the projectile to produce the plots shown in figure 14.
l
x2
B
r2 r1
x1
Figure 13 Varying Projectile Starting Position
16
Figure 14 Varying the Projectile Start Point
After all the physical dimensions of the coil were varied, the current was varied to
produce figure 15. This plot show the calculated strength of the on axis midpoint
located within the coil. As anticipated the coil has function which is shown to be:
y  3 *10 21  4.5 *10 3  1.5 *10  16
Figure 15 Varying the Current
17
Barry Hansen’s simulators located at www.coilgun.info were used to optimize the
length of the coil. Holding the inner and outer diameters fixed and systematically
adjusting the length produced Figure 16. It was noted that the length calculator
in this simulator differed significantly from the unsuccessful Matlab simulation. A
comparison of the two showed that it appears the Inductor simulator is not using
Pi x 2r to calculate the circumference of the coil and hence the length of wire per
turn. The values calculated using the simulator may be shorter than actual wire
length. The code written for Matlab can be found in Appendix C.
Figure 16 Inductor Simulator
(Reprinted with permission of Barry Hansen)
Using the inductances shown above the RLC simulator on the same website was
used to show currents. The length of 26 mm was selected to give the most
18
current by reducing the resistance and inductance. The RLC response is shown
with the coil dimensional parameters as figure 17.
Figure 17 RLC simulator
(Reprinted with permission of Barry Hansen)
Wire size selection
Current Density is defined as the amount of current flowing through a wire. As
the current flowing through a wire increases some of the energy is dissipated as
heat. If the wire in not able to dissipate fast enough the wire will melt. So
consideration was given to the size of the wire used in the driving solenoid, since
there will be a large pulse traversing it. Since the current used to produce the
field will be of short duration the following equation given by the Handbook for
Electrical Engineers for wire size used in fuses was used to select a nominal wire
size.
33
 Tm  Ta 
I2

S  Log 
2
A
 234  Ta  1 
Where:
19
I = the current in Amperes
A = the area of wire in circular mils
S = the time current flows in second
Tm = is the melting point of conductor (for copper this is 1084.62° C)
Ta = is the ambient temperature (in Celsius)
Using the average indoor temperature of 27° C the above equation was used to
find the Current Time product. With the estimated 1000 amperes desired for the
field generation this yielded three nominal choices 18, 16, and 14 American Wire
Gauge (AWG) shown in table 2. These theoretical values were de-rated a further
20% to allow for a more realistic view. The smaller gauge wire will allow for more
turns per meter and yield a stronger field. A larger gauge will allow for more
current to flow with less wire resistivity. Additionally, the larger gauge will be able
to dissipate heat faster than a thin wire and allow for a shorter firing interval. A
MatLab Mfile was created and an iteratively ran though the following American
Wire Gauges to show the time to melt at 1000 amperes. The turns per meter are
useful to show the gain in field strength since field strength is proportional to the
number of turns as shown below.
AWG
14
15
16
17
18
Current
(amperes)
1000
1000
1000
1000
1000
Duration
melt (ms)
19.0
15.0
11.9
9.5
7.5
to De-rated
20 %
15.2
12.0
9.5
7.6
6.0
Turns/Meter
(bare)
614.4
689.8
774.6
869.9
976.9
Table 2 Time to melt
Currently 16 AWG is selected for building the driving solenoid. Further
investigation is needed to evaluate the thermal resistance of the insulator of the
wire; this is to be done in the Magnetic Wire section.
H = Magnetic Field (tends to magnetize space)
 = Magnetic Flux Density (total magnetic effective results)
  0  r H
Magnetic Wire
20
Since the magnetic field strength is proportional to the number of turns per unit of
linear measurement, magnetic wire should be used. Magnetic wire is classified
as an inner conductor, usually copper or aluminum, thinly coated with a thin
polymer for insulation to maximize turns per inch. It is available in circular,
square, rectangular, rounded square, or flat shapes. It should be noted that even
though it’s called magnetic wire, the wire itself is not magnetic.
Magnetic Wire is rated by the thermal capacity of the polymer. The National
Electrical Manufacturers Association (NEMA) has set forth standards for
categorizing Magnetic Wire (MW) with the nomenclature MW-NA where NA are
alphanumeric values. Some common types of magnetic wire insulators are
polyurethane, polyamide, polyester, polyester-polyimide, polyamide-polyimide (or
amide-imide), and polyimide. Table 3 shows the Thermal Classes of insulators.
Thermal Class
105° C
Insulation Type
Polyamide, Polyvinyl Acetal, Polyurethane,
130° C
155° C
Polyurethane
Polyester, Polyester (amide), Polyester (amide)
(imide), Polyurethane
Polyurethane 180, Polyurethane Nylon 180,
Polyester Nylon, Polyester-imide, Polyesteramide-imide
Glass
Fibers,
Dacron
Glass,
Poylester
200,Polytetra fluoloethylene (Teflon)
Polyimide,
180° C
200° C
240° C
Table 3 Thermal Classes of insulators
Heat Study
The power consumed by the coil was calculated as:
P  I 2 R  (6932 (0.184)  88,366watts
This power delivered in this calculation is much higher than would be expected.
The total energy converted to heat was then calculated by assuming a sin
function for the current pulse:
t1
4
t0
0
E  R  I 2 dt  0.184 Sin ( I ) 2 dt  0.00076544 joules
Since the firing period is very short, one to two milliseconds, no cooling will be
considered to take place from the environment. Or in other words, all heat will be
21
considered to dissipate in the wire and insulator per shot. This is important when
considering the grade of Magnetic wire which will be used. Again the equation
does not yield values which seem practical.
Since the above calculations returned numbers which this group had very low
confidence in an educated guess was made for the temperature rating of the coil.
A spool of 64 feet of Magnetic Wire with an AWG of 16 was chosen for this
project, BulkWire part number Wire-MW-16-1/2. The temperature rating is 200°C
and is made from a modified polyester resin. There is an overcoat applied which
is amidi imide resin which the website is calling a “High tech enameled insulation.”
64 feet or 19.5 meters of MW-16 is priced at $12.68.
The coil will be wound with a coil guide made from a bolt the same diameter as
the barrel. The ends the center part of the bolt from point A to point B is where
the coil length will be controlled see figure 18. On opposing sides of the washers
will be two nuts per side. Turning theses nuts in counter directions will make a
solid point which can me manually adjusted as needed. When the first layer of
coils is complete a thin drop of super glue will be applied to the top and bottom of
the coil. This process will be performed for each new layer. When the adhesive
has had sufficient time to dry it the coil can be safely removed and mounted on
the barrel assembly.
Figure 18 Hand Made Coil Guide
22
Most of the fabrication for the field generation subsystem can be done in the
garage. The winding of the coil can be done by hand. Using a bolt with the
same diameter as the barrel a set of locking nuts can be used to create a coil
guide used for winding the coil. Another method would be to attach the coil guide
apparatus to drill and using a low setting have one person guide the coil while the
other controls the drill. If done properly this would create a uniform and tight coil
can be constructed. The preferred method is to use a lath. If a lath can be found
which has a low enough rotational setting, one person could use the coil guide
apparatus to wind the coil. Depending on the lath if it is a machine lath or a
woodworking lath will determine the coil guide apparatus. A Machine lath will
allow the use of the bolt coil guide apparatus. If a woodworking lath is to be used
a wooden dowel the same diameter as the coil will be needed wooded washers
would be made. The wooden washers would be made a larger dowel and drilled
to the inner diameter of the barrel dowel. This would then be fastened to the
wooden barrel dowel.
Current Pulse Generation
An industrial automotive battery was proposed and considered. The Duralast 12
volt 950 cold cranking amps (CCA) heavy industrial battery, Autozone part
number (31-950), could provide a sufficient charge necessary for demonstration
of the coil gun. However, concerns arose as to the practicality of using a battery.
It was noted that the charge time for an industrial automotive battery is
significantly longer than a capacitor bank. At approximately 60 pounds this
would be a large and cumbersome current source.
In addition to a battery and capacitor bank, it may be possible to use a
transformer to provide a high current capacity on demand. This would be highly
desirable option and allow for a rapid firing of the coil gun. One of the immediate
problems with this is the need for a fast switch which could handle high currents.
Finally, a capacitor bank could be used for the current source. According to
wiki.4hv.org capacitors used in coil guns should have a low equivalent series
resistance. Intuitively this makes sense since the goal is to deliver maximum
current to the coil. Readers were warned that super capacitors, appealing as
they may be with high capacitance, were not to be used in coil gun applications.
A pulse discharge will wreck the super capacitor. In addition to low equivalent
series resistance a low equivalent series inductance is also good. However, if
the inductance of the capacitor is much less than the inductor the equivalent
series inductance can be ignored.
23
For this project 5 millifarads are needed to generate the current pulse.
Electrolytic capacitors were selected because they are inexpensive and have a
high capacitance for their size. When designing the coil firing circuit it will be
important to prevent any current from entering the electrolytic capacitors from the
wrong polarity. This could cause a catastrophic failure of the capacitor. This will
be done with diodes known as fly wheel diodes.
The Type 450C 105°C Ultra-Ripple, Long-life, Inverter Grade, Radial Leaded
capacitor was the capacitor of choice for the capacitor bank due to the low
equivalent series resistance. The equivalent series resistance is 6.9 milli ohms
and Five 1000 μ Farad capacitors in parallel would provide the desired 5000 μ
Farads. However, due to unavailability of the part, Mouser part number 598450C102M350AJ8, and manufacturer part number 450C102M350AJ8 this part
was not selected.
Secondary and design choice is the Type CGS High-Cap Screw Terminal
Aluminum Electrolytic Capacitor. Manufacturer part number CGS102T350V4C,
and Mouser part number 539-CGS350V102V4C. This is a 350 Volt 1000 μ
Farad capacitor with an equivalent series resistance of 140 milli ohms. The price
is $28.80 each. Five of these capacitors will make the capacitor bank which will
provide access to 225 joules.
1
1
Pe  CV 2  (0.005)(300 2 )  225 joules
2
2
The five capacitors will be connected in parallel using a copper bus. The screw
termination of the capacitors will allow for holes to be drilled in a copper bar and
fastened with screws.
The calculated kinetic energy needed in the Force section was 4.8 joules. If a 2%
energy transfer is assumed this would be barley enough to accelerate the
projectile to 31 meters per second.
External Iron
External iron can be used to reduce the reluctance of the coil. The external iron
will guide the flux lines to the central axis of the solenoid. This should improve
the energy transferred to the projectile. In order for the external iron to be
effective, it should be used on all sides to include the ends, allowing just enough
room for the barrel and projectile to pass through. Iron pipes used for plumbing,
a common item found at the local hardware store, can be used for the external
iron. Caps can be used for the ends of the barrel assembly. One of the major
24
tradeoffs for this project is the additional weight that will be added. Since the
barrel will be moved by servos a heavy firing tube is undesirable. Additionally,
the coil will have a longer cool down time due to the enclosed system. Finally, it
is possible that eddy currents in the external iron could be counter productive to
the firing process. This could be reduced by notching the external iron. Figure
19 shows the external iron concentrating flux lines.
Figure 19 External Iron Guiding Flux Lines
(Reprinted with permission of Barry Hansen)
Velocity Detection
Since the major design requirement is to achieve a velocity of 31 meters per
second. A built in velocity sensor will be used to improve performance in the
prototyping phase of this project. There are several methods to determine the
velocity of a projectile. The methods covered in this project are electrical contact,
mechanical contact, field sensing, and optical sensing.
25
The electrical contact method would use the projectile to complete a circuit as it
passes through a point in the barrel. This could start a counter which would be
stopped as the projectile continues down the barrel. One of the major design
issues with this system is that the projectile needs to make good contact to
complete the circuit. This means that it must either fill the entirety of the barrel or
have the electrical leads protruding into the barrel. This would be a poor design
approach. Since maximum velocity is a primary design goal protrusions into the
barrel would be counter productive. Additionally, since the design of the coil will
need to be optimized in the prototype phase the projectile may be altered from its
original design which could severely cripple the electrical contact velocity
detection method. Even if the above concerns were mitigated, there is additional
potential for the electrical contacts to reduce velocity.
The second method, mechanical contact, is similar to the electrical contact
method. This would use thin springs protruding into the barrel that would
depress an electromechanical switch. Again two sensors would be need. The
first would trigger a counter and the second would stop the counter. As
discussed in the electrical contact sensor the mechanical contact would also
inhibit the velocity of the projectile.
The third method would be to use sensing coils at the end of the barrel. As the
projectile passes through the sensing coil a current is induced which could be
used to trigger a counter. When the projectile passes the final coil the counter
would be stopped and the velocity could be determined by the distance between
the two sensors and the time it took to traverse that distance. A major benefit to
this type of velocity sensing is the noninterference into the barrel or with the
projectile. It would also allow for the coils to be slide easily along the barrel for
adjustment. Since the project is a coil gun the use of sensing coils is appealing.
An article from Sensors Magazine was published about the United States Air
Force used a coil sensing system at a laboratory on August 1, 1998. Kaman
Instrumentation Corp. designed the Kamen KD-2300 Dual Sensor Velocity
System. This system is used by the Impact Physics Lab 0.50 caliber light gas
gun facility to measure and compare the projectile performances. This system
uses a one megahertz carrier frequency to produce eddy currents in the
projectile. The eddy currents allow the sensing coils to know exactly where the
projectile is at. This system allows for the detection of high velocities up to 3000
feet per second and with an excellent precision. The coil sensing method would
be an excellent project in itself. However, since this project is bound by the time
and money the coil sensing method will not be used.
The fourth method would be to use optical sensors to capture the velocity of the
projectile. Similar to the inductive sensing coils, an optical sensing system would
be unobtrusive. However, if the optical sensors are incorporated into the barrel
26
they will not be adjustable like the coils. Holes will need to be drilled in the barrel
and the sensors seated below the path of the projectile. The potential arises that
the holes where the sensors were drilled will allow the projectile to catch and
tumble or lose velocity. This is especially likely to happen when the projectile is
not rounded or if the projectile does not fit the barrel.
The theory for optical velocity sensing is simple. Using two sensors place in the
barrel at known distances, the time it takes the projectile to get from the first
sensor to the second the velocity can be calculated. The infrared light emitting
diode will be constantly emitting. When the projectile passes through the
between the light emitting diode and infrared sensor the sensor will trigger a
counter. When the projectile passes through the second set of emitter sensors
the counter will stop.
For the optical sensing the infrared spectrum was chosen. Infrared signaling is
commonly used in household applications with remote control devices being the
most obvious. The benefits to using an infrared sensing system are that the
visible lights will not trigger the sensor on accident. For the light emitting diode a
SE3455-003, manufactured by Honeywell, was selected. This light emitting
diode will output 935 nanometer wavelengths in a 90° beamwidth. For the optical
sensor, a SD3410-002, also manufactured by Honey well, will be used. This
sensor has a 90° or 12° nominal acceptance angel option and is mechanically
and spectrally pared with the SE3455-003 light emitting diode. Both the sensor,
Mouser part number 785-SD3410-002, and the light emitting diode, Mouser part
number 785-SE3455-003, are currently available.
When the sensor conducts the LM393 comparator, Mouser part number 863LM393NG, will trigger a counter. The counter will be housed in a Programmable
Integrated Circuit. PIC16 will be used as the counter. When the comparator
initiates the PIC the program will start counting. The output of the PIC will be
sent to an eight bit display as shown in figure 20.
27
Figure 20 Speed trap Calculation
(Reprinted with permission of Donnie James)
The sensors will be positioned three inches apart with the last sensor a half an
inch from the end of the barrel. Combined they create a speed trap. The speed
trap is optical sensors are shown in figure 21.
Figure 21 Optical Sensor
Data collected from the optical speed trap will be compared to a look up table to
arrive at the actual projectile velocity. Testing the optical speed trap can be
done with a chronograph. Firing the projectile through a chronograph and
28
comparing out puts will verify functionality of the optical speed trap.
Thecomplete circuit is shown as figure 22.
Figure 22 Total Speed trap
(Reprinted with permission of Donnie James)
Testing
The primary tests for the Field Generation Subsystem will be to show that the
magnetic wire will be able to handle the current without damage to the wire or the
insulator. After procuring the capacitor bank a length of wire will be tested
against a large current source. This needs to be done early next semester so
that if the enamel is insufficient a higher temperature magnetic wire can be
procured.
The final test will be to see if the field generated can achieve velocities of 31
meters per second or greater. This can be tested by several methods. The first
method would be the linear drop test. This method uses the linear distance
traveled from a given height to show the velocity. A simple equation, distance
traveled = time * velocity. Setting the coil gun on a level surface 4.8 meters, the
measuring the linear distance traveled by the projectile. Another method, if
available will be to use a gun chronograph. The Chronographs use two optical
sensors to determine the velocity. A built in Chronograph would be highly
desirable for this project and would be much cheaper than purchasing new test
equipment. Chronographs pricing starts at $50 dollars, if enough funds are left
29
after purchasing components this test equipment will be procured. A paint ball
chronograph will work just as well.
Barry Hanson has devised a test to find the inductance and strength of the coil.
Using a resistor decade box in series with the coil, adjust the decade box until
the voltage is the same, across the decade box and the inductor. When the
voltages are the same the resistor setting will be equal to the reactance of the
coil. Where, n is the number of turns and f is the frequency.
L
Xl
2nf
Measuring the coil strength can be done by setting the projectile on its back and
placing the coil above it see figure 23. Connect the coil to a variable voltage
source. Increasing the voltage until the projectile is barely touching the table.
Once this is met record the voltage in a table comparing the height and voltage
needed to lift the projectile. Mr. Hanson calls this voltage force a one G. This
process will be contused for variation of elevation points.
Figure 23 Coil Strength Test
(Reprinted with permission of Barry Hansen)
Use the following equations to complete the table.
depth  projectilelength  Height
Forcemax 
VMax
Vi
30
Since air filled coils have linear properties, up until saturation of course, the
Force at maximum voltage can be calculated. Plotting the data collected in this
table will produce a plot which will show the magnetic field strength in terms of G
see figure 24. It should be noted that this is not a measure of actual Standard
given by the International System of Units. It is however a method of measuring
the field strength with respect to the projectile. Even though it is not an
International System of Units standard the plot would be helpful in determining
the nominal position for the starting projectile. Once this accurate and specific
point has been found for the firing solenoid, a mechanical loading system will be
built to deliver the projectile to the same nominal position consistently.
Figure 24 Magnetic Field Strength Test
(Reprinted with permission of Barry Hansen)
Field Generation Design Summary
The Field Generation subsystem was designed with the following parameters
table 4 and cost estimate table 5. This subsystem will take one fourth of the
estimated project cost.
Parameters:
Coil Length
Coil Inductance
Coil Resistance
26 mm
0.352 m H
0.184 ohms
31
Inner Diameter
Outer Diameter
Wire Gauge
Wire Insulation
Projectile Mass
Projectile Length
Projectile Width
Turns in coil
MW linear length
MW weight
Winding density
Force needed to accelerate
Acceleration Length
Work needed
Field strength at Coil Midpoint
Power dissipated in coil
Capacitors
IR emitter
IR Sensor
Comparator
PIC
Current trigger
Fire control trigger
13 mm
29 mm
16 AWG
High Tech Enamel
5 grams
25.4 mm
6.35 mm
171
13.97 m
.16 kg
7.5 turns/cm
94.5
0.0508 m
4.805 joules
3.851 Tesla
5 x 1000 μ Farad
SE3455-003
SD3410-002
LM393
PIC16
SCR
555 Timer
Table 4 Field Generation Design Parameters
Item
Magnetic Wire
Capacitor
PVC Barrel
Projectile (bits)
External Iron
IR LED
IR Sensor
LM393
PIC16
8 bit display
SCR
555 timer
Subsystem Total
Cost Ea
$12.68
$28.80
$2
$8
$20
$4.80
$7.10
$0.50
$2.00
~$10
$120
$5
Qty
1
5
1
1 Pkg
1
2
2
1
1
1
1
1
Total
$12.68
$140.00
$2
$8
$20
$9.60
$14.20
$0.50
$2.00
$10
$120
$5
$343.98
Table 5 Field Generation Cost Estimate
32
Heat Dissipation
The coil gun is going to generate massive amounts of heat. There is a lot of heat
coming from a lot of power dissipation among electrical devices. Magnetic fields
generated by change in electrical fields create most of the heat. Keeping the heat
to a minimum is going to be a huge obstacle. There are several manners of
which the heat can be reduced. The key to keeping heat to a reasonable value is
by making sure to account for the power generated by the device used to
dissipate heat. Power generated by the heat dissipation devices should not
interfere with the 300V being discharged of the gun. The usual choice for cooling
an object is the use of fans. Fans are great coiling devices for the project
because they can transfer the heat away from the gun. Heat is an adversary of
every electrical component in the circuit. Taking care of the heat increases the
efficiency of the gun. The use of natural air to cool the electrical component is not
possible.
Heat sinks can also be used as a cooling device. A heat sink transfers heat from
one medium to another. In most cases the medium is air or liquid. Liquid cooling
works a lot better but is very expensive. Liquid cooling involves liquids such as
liquid nitrogen which is fairly expensive. Despite being expensive this is not the
only drawback. It can be very dangerous due to its extremely low temperature.
When this liquid comes in contact with human tissue it leads to frostbite. Another
issue arises when it comes in contact with a hotter substance. It begins to boil
immediately after contact and turns into a gas.
Many circuit elements of the coil gun will be exposed to large amounts of heat.
The charge resistor will experience a majority of the high amounts of heat, only
second to the actual solenoid. Since the charge resistor for the gun is going to be
light bulb power dissipation through heat is not an issue. The light bulb does a
good job of dissipating the heat by emitting light. The diodes in the rectifier
circuit are going to experience high switching current; therefore they are subject
to high heat as well. To prohibit these from overheating a fan will likely be used
to cool the devices. Capacitors might experience high heat because of the large
amounts of energy being discharged. Keeping the capacitor bank cooled will
reduce the risk of premature failure of an individual capacitor. This is a major
concern because high voltage rated capacitors are expensive.
Computer fans are great choices for use in a coil gun. There are several different
types of computer fans to choose from. There are fans designed for all purposes.
Computer game enthusiasts use fans with high CFM. CFM stands for cubic feet
per minute. A Cubic feet per minute refers to the measure of air flow that is
generated by the fan. Fans also are described by their revolutions per minute or
RPM. Fans with higher RPM have higher decibels. Some fans come with fans
that can be adjusted electronically or manually. Fans that can be manually
33
adjusted use a potentiometer to allow change of speed. Fans can be adjusted
thermally or by computer hardware or software too. In the coil gun there isn’t a
need to have an adjustable fan. Since computer fans are 12 V using a smaller
voltage would cause the fan to run at a slower speed and is less noisy.
For the coil gun multiple fans and heat sinks will be used to assist in dealing with
heat dissipation. For the capacitor banks cap coolers will be used to deal with
the heat that the AC to DC power convertor generates. The capacitor cap
coolers are very effective for keeping the capacitors at a safe operating
temperature. The weights of the capacitors coolers are fairly light. From first
glance they appear to be a permanent attachment to the capacitor. The solenoid
of the coil gun is going to need more than just a fan or heat sink to deal with heat
dissipation. The heat dissipation in the solenoid is going to need some type of
insulation to account for the high heat dissipation that is not just on the surface
but internal as well. Surface heat is easy to cool because fans and heat sinks
easily dissipate that heat, but internal heat dissipation can be troublesome.
Internal heat usually depends on the type of conducting material and its
resistance. The higher the resistance in the material usually means more power
is being dissipated. To cool the internal heat it may be necessary to use different
material, material size, or adding some type of thermal compound. Fans and heat
sinks do not efficiently help with dealing with internal heat, they are more for the
surface area of conducting material. Two small fans were chosen to cool
electrical components of the coil gun. The two fans chosen to be used will be
computer fans. The function of the fans will be to provide the gun with cool air
while expelling the hot air as well. The size of each of the two fans is 80mm. This
is the standard size that is currently used in desktop computers. Computer fans
are excellent cooling devices. They have proven to be the essentials of desktop
computers for years. The two fans may have built in potentiometers to increase
fan speed if for some reason it is needed.
Magnetic Circuit
To better understand the physics of a coil gun it is necessary to be
knowledgeable of magnetic field generation. A magnetic circuit is closely related
to that of a basic electrical circuit. In a magnetic circuit a magnetomotive force
(mmf) denoted by the letter F provides a flux through a ferromagnetic core. The
equation used to describe mmf is as follows:
ℱ = 𝑁𝑖
The letter N is the number of turns in the coil and i is the current through the wire.
The magnetomotive force should be looked at similar to the voltage source of an
34
electrical circuit. The purpose of mmf is to create an electromagnetic field by
passing current through a wire that contains multiple turns. The mmf has polarity
as well. To determine the polarity of mmf we use the right hand rule. We take the
fingers of the right hand and curl them the direction of the current through the
wire, whichever way the thumb is pointing is the positive direction of the mmf.
This concept is based on the basis of the Biot-Savart law.
The Biot-Savart Law explains how a magnetic field denoted by the letter B is
generated by an electric current. With Biot-Savart’s law we can calculate the
magnetic field that is being generated by the current that we provide the gun
with. B =µH. µ is simply the magnetic permeability of the material. H is the
magnetic field intensity. The magnetic field intensity is the effort exerted by the
current to create a magnetic field. This is explained through Ampere’s and
Maxwell laws. To understand how we can increase the magnetic field we have to
know the relation the magnetic field intensity has on the magnetic field. The
magnetic field intensity, H is equal to mmf divided by length of the core. In a coil
gun this is length the projectile travels before leaving the end of the solenoid. The
main concept that should be evident between magnetic field and magnetic field is
that they are proportionate to one another. In are design of the coil gun we will be
looking to maximize the magnetic field and intensity.The magnetomotive force
can be expressed in terms of flux and reluctance. The equation is as follows:
ℱ = øℛ
Phi in the equation is the flux. The flux can be thought of as the magnetic current
that is produced by the magnetomotive force. The r is the reluctance, better
known as the resistance of the magnetic circuit. The reluctance in the coil gun
should be kept to a minimum so that flux will flow freely. The reluctance is
dependent on two major factors the length and the cross sectional area of the
solenoid. The reluctance is given by the equation:
ℛ =
𝑙
µ𝐴
To obtain the least resistance in our circuit we must minimize the length of our
solenoid as well as increase our cross-sectional area. We do have control of our
length of solenoid but not over the cross-sectional area because of other factors
pertaining to the projectile being used. It is important to visualize the conceptual
idea of having minimal reluctance in the magnetic circuit and more flux.
Permeability is going to play a vital role in the coil gun magnetic circuit. Air and
the projectile are going to be the core of the magnetic circuit. The core is
composed of ferromagnetic material. In the coil gun barrel assembly the iron bit
is the ferromagnetic material which has a very high permeability. The high
35
permeability means that the degree of magnetization is high. The projectile will
be able to react to the magnetic field that is induced from the current traveling
through the coil.
Switching System
In order for the capacitor bank to release stored energy to the coil it must have a
switching system. A switching system needs to be implemented in order to get
current from the bank to the solenoid. Normally this can be done by any type of
switch. With a high voltage, high current coil gun the use of a normal switch is
not an option. A normal switch is not an option because they cannot handle high
currents. The high currents will cause such devices to fail. The coil gun must use
some type of switch which can handle a high current surge and block high
voltages. There are multiple ways of which this can be done with a bipolar
junction transistor, metal-oxide semiconductor, silicon-controlled rectifier, or an
insulated gate bipolar transistor. All devices can control high current surges. For
the coil gun it is ideal to want a device that will be able to switch on and off via
current or voltage. The best device to implement switching is a transistor. This is
not the only device that can be used for quickly switching the current from the
capacitor bank to the solenoid.
A SCR is a Silicon-controlled rectifier. Despite the name SCR it is nothing more
than a thyristor. The name SCR represents a brand of thyristors but is commonly
used to describe all types of them. It is composed of four layers including P and
N type semiconductor materials. In the coil gun as soon as current travels
through the silicon-controlled rectifier’s gate after exceeding its holding current, it
will turn on. The holding current is dependent on the specifications of the siliconcontrolled rectifier. When it is turned on it will then release the stored capacitor
bank charges to the coil via current. The device will remain on as long as the
current through the anode remains above the holding current. The siliconcontrolled rectifier in the coil gun will stay on until after every shot. The problem
with a silicon-controlled rectifier is its properties prohibit it from turning off. Once
getting a current to flow through the gate the device is in active bias and
conducting. At that point in time it cannot be turned off in the middle of the
current pulse. Using a SCR in a coil gun might be necessary because of its low
cost to voltage ratio. The figure 25 shows a silicon-controlled rectifier that is used
in a high powered system. From the picture it does not look similar to a diode as
seen in circuit schematics. It none the less functions very similar to a diode. The
picture shows the little wire that represents the gate that the triggering system
needs to be connected to.
36
Figure 25 Silicon Controlled Rectifier
A bipolar junction transistor can be used for switching. For a coil gun it is
beneficial to use an NPN type. A NPN transistor can be used as a switching
device when the device is operated in saturation mode. For the bipolar transistor
to operate in saturation mode there must be a large current supplied to the base.
The current inputted in the base terminal allows the transistor to pass larger
collector to emitter current. This transistor in a coil gun would not be ideal at all.
The current supplied to the base current has to be as much as 10% of the
collector to emitter current to have the device operate in saturation mode. Since
the coil gun uses a high current surge from the collector to the emitter it would be
inefficient to use this device.
An IGBT is an Insulated Gate Bipolar Transistor is shown in figure 26. It is known
for its fast switching while generating large pulses. An IGBT utilizes favorable
characteristics from metal-oxide field effect transistors and bipolar junction
transistors. The cross section of an IGBT shows its similarities to the two
common transistors. It contains three terminals, one for the gate which comes
from a MOSFET. This is helpful because it is easily gate driven as well. The
other two terminals are similar to that of a BJT they are the emitter and collector
terminals. IGBTs are commonly used in many appliances. In order to rapidly fire
a coil gun it might be suitable to use one. There are drawbacks in using IGBTs
shown in figure NN. They are a bit expensive at high current ratings. In order to
activate an IGBT in a coil gun it is necessary to use a gate driver. Another issue
that arises when using IGBTs is voltage spikes. If there is not an IGBT with a
high enough voltage rating and a short distance in wire between the capacitor
and IGBT voltage spikes become an issue when the IGBT is turned off.
37
Figure 26 IGBT
It is possible to use an n-channel power MOSFET for the coil gun switching
system. The problem with a power n-channel MOSFET with a coil or solenoid as
a load is the back electromagnetic field. The back electromagnetic field would
likely destroy the device over time. In a coil gun the back electromagnetic field
would occur too frequently because a magnetic field is being generated each
time the projectile is being shot. For a coil gun circuit there would need to be the
addition of a flywheel diode in parallel with the solenoid to reduce the back
electromagnetic field. The flywheel diode is used to suppress voltage spikes in
switching circuits when the switching device is off. For our coil gun we will use a
flywheel diode in our circuit as extra precaution against a back electromagnetic
field. A flywheel diode is not a specific type of diode; it is simply a description of a
diode in parallel with an inductive load of a switching circuit. A power MOSFET
is shown in figure 27. The figure does not appear to be a standard MOSFET.
That is because a power MOSFET has multiple MOSFETs combined. The
additions of MOSFETs add strength to the power MOSFET so that it can handle
more power.
38
Figure 27 Power MOSFET
Transistors are excellent switching devices in electronics. They are usually
inexpensive in simple applications that do not require huge voltages or currents.
It may be possible to use passive components to construct switching circuits but
will require a lot of time and many passive component combinations. Transistors
are active components that already come with beneficial characteristics to
switching on and off a circuit. The ability to amplify makes it very useful even in a
coil gun. Many of the transistors earlier mentioned allow large current and
voltage to flow through the capacitor bank to the coil or solenoid. Despite their
many advantages there were numerous drawbacks found from using the above
devices for switching in a coil gun because of the high electromagnetic field that
is produced. It is apparent that high electromagnetic fields in circuits are no good
to electrical components, especially those made of silicon. Silicon products are
great for high current and voltage circuits but inducing voltage to create current
through magnetic fields cause complications.
Design
In the coil gun design a silicon-controlled rectifier was chosen. The rectifier will
be able to switch on the large current needed to travel to the coil. The siliconcontrolled rectifier chosen is 100 amps with a 1400 amp holding current. The
duration rating needs to be short so a low current rating of below 6 milliseconds
39
will be used. In choosing the rectifier current rating it was pertinent to know that a
SCR can handle 10 times the current rating for brief periods of time. This
exceeds the current done in simulation but is just a precaution if the group
decides to increase voltage to obtain a higher velocity projectile with respect to
the 400V capacitor power rating. Some difficulty was found while trying to choose
the appropriate switching devices. Power n channel MOSFETs and NPN bipolar
junction transistors were quickly eliminated from our choices. The problem with
the power n-channel MOSFETs were the back electromagnetic field. The
transistor would need a flywheel diode which isn’t a big issue since one is going
to be used either way. The issue is there may still be enough back
electromagnetic field to hinder the function of the MOSFET. The MOSFET
transistor experiences voltage spikes when the device is off. The choice of using
the NPN bipolar junction transistor as a switching device for a coil gun was far
from consideration. This device presents too many problems in a coil gun circuit.
It is highly inefficient when it comes to amplifying current. The base current would
need to be almost 10 percent of the discharge current which is massive. An IGBT
was the biggest competition for the silicon- controlled rectifier. It has many of the
characteristics except for a major problem with voltage spikes. From the ones
looked at for the design of the coil gun, most were very expensive. If price
weren’t a concern this device would of likely been a good choice.
The silicon-controlled rectifier will be connected between two parallel circuits.
The anode terminal will be connected to the circuit with the flywheel diode and
coil in parallel. The cathode terminal will be connected to the parallel capacitor
bank. The gate current supplied to the device from a direct current
Bleed Resistor
When the coil gun is turned off there should be no type of activity going on. In an
ideal coil gun system the stored energy would fully discharge once the gun has
shot the projectile and is turned to the off position. This is not an Ideal coil gun,
so issues in the circuitry will arise. The stored energy in the capacitor bank is
going to need a way to filter out, when the gun is turned off. A bleed resistor will
be used so rectify the situation; it ensures that capacitor bank is fully drained.
The bleed resistor is a normal resistor that will be placed parallel to the highpower capacitor, which in this case is the capacitor bank. Choosing the value of
the resistor can be found utilizing a few simple equations. It is important to
choose the correct resistor because choosing a value too low can cause
problems with heat due to power dissipation. The first thing done is find the
charge time of the capacitor bank. The bleed time will be one hundred to a
thousand times the value calculated for charge time. The equation to calculate
the resistor value is give below:
40
𝑅 = 𝑡/𝐶
The letter t in the equation is the bleed time and C is the capacitance of the
capacitor bank. The value for the coil gun of the capacitor will likely be over
100kΩ. Anything lower would overheat. To calculate the power used by the
resistor take the voltage squared divided by the resistance that was calculated in
the previous equation. This is the power rating that will be used when obtaining
that resistor. The bleed resistor is an excellent safety addition. It ensures safety
when storing the device for long periods of time that there is no storage of charge
in the capacitor banks. This always prolongs the life of the capacitor bank too
because no work is being done while the device is off. Every high-power device
should have a bleed resistor. The bleed time is not too important for anything
other than long term storage. It will not affect the charge time either, but it is a
good thing to know because it illustrates how the capacitor bank is drained. If
more capacitors are added to the capacitor bank of the coil gun, then a larger
bleed resistor will be used. The above equation will help in calculating the actual
resistance of the bleed resistor.
Below are two tables with bleed resistances that will be used for the coil gun.
The first table,shown as table 6, is for bleed resistances that are one hundred
times the charge time. The second table,shown as table 7, is for bleed
resistances that are a thousand times the charge time of the capacitor bank. The
resistor values from the first table appear to be more reasonable values and
easier to find. In the coil gun those will be used. From the values in the first table
the average bleed resistance is 99.33 kΩ. As a precaution a 150 kΩ resistor will
be used in the circuit.
Voltage
300
350
400
Bleed Time
288
392
512
Capacitance
0.004
0.004
0.004
Bleed
Resistance
72000
98000
128000
In KΩ
72
98
128
Table 6 Bleed Resistance
Voltage
300
350
400
Bleed Time
2880
3920
5120
Capacitance
0.004
0.004
0.004
Bleed
Resistance
720000
980000
1280000
In KΩ
720
980
1280
Table 7 Bleed Resistance
41
Damping Resistor
The projectile should travel from the breech of the barrel forward. A possible
problem that may occur is suck back. Suck back is when the projectile travels
forward through the solenoid and then before completely leaving the solenoid is
sucked back and shot out through the breech. It usually occurs when the timing
of the current through the coil is incorrect. The current through the coil should not
be on when the projectile has traveled half way through the coil. The addition of
the flywheel diode or protection diode in the coil gun does prevent back
electromagnetic field and protects the solenoid. It however does have a
drawback that can cause suck back. Although the flywheel diode opposes back
electromagnetic field, it may circulate current to the coil. If this occurs a damping
resistor will need to be placed in the coil gun. The flywheel diode may possibly
be removed from the circuit if the damping resistor is added. A damping resistor
will make the RLC circuit critically damped. Since the circuit becomes critically
damped there would be no need for a protection diode. The electrolytic
capacitors only need protection when the RLC circuit is under damped. If the
damping resistor is needed, it will be placed in series with the coil, capacitor
bank, and silicon-controlled rectifier. The addition of the damping resistor will
also reduce the potential of ringing than could occur after each discharge of the
capacitor bank in the coil gun. The ringing in the capacitor bank after discharge
can be somewhat annoying at times. Eliminating this constant ring would
beneficial. Below is the equation used to calculate the value of the damping
resistor that will be used in the circuit:
𝑅 2 𝐶 2 − 4𝐿𝐶 = 0
𝑅=√
𝑅=
4𝐿
𝐶
4∗(352∗10−6 )
√ (4∗10−3 )
= .593 Ω
The equation above takes into account the capacitor banks equivalent series
resistance, coil resistance, and the wire resistance. The R that is obtained above
is the resistance to keep the series RLC circuit to be critically damped. It is
evident that the resistance depends largely on the value of the capacitance. The
inductance looks like it would be a factor since the equation has the inductance
times a multiple of 4, but the value is extremely low. In the coil gun a resistance
of .583 ohms will be used as the damping resistor.
42
Digital Voltmeter
In most coil gun projects the designers choose to connect a voltmeter or
multimeter to measure the voltage the gun is operating at. In this coil gun it was
decided that it would be better to have a more permanent solution to measure
voltage across the capacitor banks while they are being charged. A digital
voltmeter will be included in the gun’s circuit to easily display voltage. An
additional feature may be to have a current meter added as well. For the coil gun
it is a necessity to monitor the voltage being used at all times to make sure the
maximum voltage is not exceeded. The voltmeter ensures that the voltage across
the capacitors does not exceed the maximum voltage rating of 400V. As a
precaution the banks voltage will be well below the maximum voltage rating.
Keeping a significantly lower voltage than the maximum voltage rating maximizes
the life and quality of the capacitors. The digital voltmeter will be placed in
parallel with the capacitor bank. The amp meter if added will be positioned in
series with the parallel capacitor bank and the solenoid.
Triggering System
There are few ways to design a triggering system for a coil gun. The three
systems found that will be decided between are the open loop triggering system,
the optical triggering system, and then finally the induced voltage sensing.
The first method that will be discussed is the open loop triggering system shown
in Figure 28. This method is executed by sending a pulse of current through the
coil at an exact time period and then the current extinguishes itself. To best
determine the time period to use a trial and error method is used. The design of
the open loop triggering system will contain 555 timer circuits which are configure
to output pulses. The figure below will be used to help describe the operation of
the timer circuit. The output pin 3 and the discharge pin 7 are low in the nontriggered state. The capacitor C1 in the non-triggered state stays discharge.
When a negative edge is sent through the node at the trigger pin 2 the discharge
pin goes into a floating state and the output goes high. At that point C1 begins to
charge, once the threshold pin 6 becomes two thirds Vcc the output then
becomes low. Once this occurs pin 7 then becomes low and C1 discharges.
Now the device is in a reset state. For this circuit, Pin 4 and 5 are not a part of
the functionality. On this circuit Pin 4 acts as the active low and Pin 5 is coupled
to the ground.
43
Figure 28 Open Loop Triggering System
(Reprinted with permission of Barry Hansen)
The approximate output pulse is estimated using the following formula:
𝑡𝑝 = 1.1 ∗ 𝑅1 ∗ 𝐶1
This is achieved using the function of using two thirds of the voltage supply and
solving for the time constant.
Because of its pulse triggering signal is precision, despite the coil and power
supply configuration, optical triggering is another system that is under
consideration. Another plus to this system is the fact that this system is immune
to electromagnetic interference for the most part. The figure below shows the
configuration of the circuit discussed. The first circuit shown has two transistors,
one being connected to the other through its emitter, a switch and the actual coil.
The circuit below the first circuit is an improved version, and more likely choice
for this project.
For the first circuit, the phototransistor, or SFH309, conducts as the gate is not
obstructed by the projectile. At this time the majority of the voltage drops through
the resistor going into the collector, making the voltage low at the collector of the
phototransistor. The voltage across the collector resistor controls the mosfet, or
BUZ10, which in turn is the switching device. Once the projectile creates an
obstruction to gate of the mosfet the current through the phototransistor becomes
low and the voltage across the collector approaches the approximate value of the
voltage supplied. At this point the mosfet is now on and the current is now flowing
through the coil that drives the projectile. As the current starts flowing through
the coil the projectile starts accelerating and leaves the gate clearing the
obstruction of the gate. The collector voltage then lowers back down, the current
44
starts flowing through the phototransistor, and the mosfet is turned back off. To
help lower a voltage spike as the switch turns off a diode is placed across the coil
driving the projectile. The diode across the coil is called the commutating diode.
The next circuit, shown in Figure 29, is the improved version of the first circuit
discussed. A comparator is created using an op-amp. The improved circuit can
sink and source more current than the first circuit due to the op-amp. Since
turning the mosfet on and off is controlled by applying a voltage to the gate and
in turn removing the applied voltage to gate, the op-amp allows for a quicker
on/off switch due to the lower resistance of the current source/sink than the
phototransistor alone.
Figure 29 Optical Triggering Systems
45
The induced voltage triggering method is the final triggering method that will be
discussed. The idea of this method is to use the magnetic flux from the coil and
the projectile used. By investigating the induced voltage in the sensing coil,
shown in the Figure 30 below, the idea of this triggering circuit starts to make
sense. The waveform in the figure is created by the induced voltage in the coil.
The biggest obstacle of this method is due to the fact that determining the
relationship between the projectile and the induced voltage in coil requires a lot
of trial and error to best achieve the most accurate timing that would we would
like for the trigger.
Figure 30 Voltages in the Sensing Coil
Out of the three methods, the first two seem to be the better choices because of
simplicity. Those two methods again are the open loop triggering system and the
optical triggering system. To decide between these two methods we must delve
deeper into more specifics of each system. Which system would be more
reliable and which would be more inexpensive, is the main question that has to
be asked. Because of simplicity the 555 timer seems to be the best choice for the
triggering system.
Motion Control and Sensor Subsystem
46
Motion Control
The motion control system of the coil gun is constructed by the FPGA, motor control
connector, two servo motors and two motion planes. The FPGA provides the control
PWM signal to the servo motors, and the servo motors make the two motion plane rotate
right and left or up and down. Then, the barrel of the coil gun will move to same direction
of the plane move.
The most important part of the motion control which is the motor. In order to choose the
proper motor for this project, the RPM and Motor Torque has been determined. For the
RPM, if the motor move at the speed (v) of 3 m/s that mean the target can move 3
meters (dh) in 1 second in the horizontal direction. The distance (d) from the gun to the
target is 2 meters. ∠A is the angel velocity of the require for the gun to follow the target,
it can be calculate as ∠A= arctanc(dh/d)= 56.31o. And the RPM can convert by
∠A/360*60=9.385 R/M. The other Once the RPM has been determined, the next step is
determine the torque require for move the gun. There two difference rotating plane in
this project. One plane is directly attached to the gun and it makes the gun move up and
down. Another plane is at the bottom, it make the gun move left and right. The weight of
the barrel is 1 lb. However, it’s better to over estimating it, since it’s better to use the high
power motor to list the light item than use the low power motor to list the heavy item. So
total weight for the barrel is estimate to 1.5 lb. Tq = r*F which is 120 in-oz. For the
bottom plane, even though, total weight is a little higher than the upper plane. However,
the r is much small than the upper plane, so if use the same torque motor is powerful
enough to move the plane move right or less.
Figure 31 and Figure 32 show two different modes in the Fire control system.
Either it cans manual control by the computer or it can track and shot the target
by the corresponding color. In the manual mode, shown in figure 33, first the user
control tracking system and told the motors to control gun point to the desired
location then the user shoot the target by pushing the button on the fire control.
Different from the Manual mode, the auto mode, Figure 34,user doesn’t’ need to
tell the tracking system which target needs to point at. But user needs to tell the
tracking system which color of target needs to track. Then the tracking system
will tell the motors turn and make the gun follows the color target. Also at the
same time, the sensor will tell detect the target is the right color target the user
looking for. If the target color is wrong the sensor sent back the data to the
tracking system and tell it stop tracking. Otherwise, the sensor will tell the fire
control to shoot the target.
47
Figure 31 Manual Mode Control
Figure 32 Auto Mode Control
DC Motor Vs AC Motor
There are many different types and sizes of electric motors. Electric motors can
generally be divided into 3 types: alternating current (AC) motors, direct current
(DC) motors and Universal motors. A DC motor will only run in the DC current
source, on the other hand the AC motor only can run in the AC current source.
The universal motors can use both AC current and DC current, however, it cost
more money as well, so it has put aside from this project immediately. Compare
both AC motor and DC motor. Usually the AC motor has larger size than the DC
motor. As show on Figure 33 the AC motor is 4 times larger than then high
torque DC step motor, for the small torque DC motor is only 1/20 of the size of
AC motor. Also, AC motor and DC motor have different construction. Usually DC
motors utilize permanent magnets so none of their energy needs to be used in
the creation of an electromagnet as in AC motors, so the DC motor can obtain a
relatively high amount of mechanical power than the AC motor. As mention
above, this project require tow electronic motor in order to adjust the fire direction.
The precision is very important for the gun fire system. Compare the precision of
the DC motor and AC motor, the DC motor will be better than the AC motor. Of
course it is possible to use an AC servo instead, but it requires using the encoder
48
and decoder to send the message to the motor, that increase the difficulty of the
design. As the all the consideration, the DC motor have been choice instead of
the Ac motor.
Figure 33 Size of DC motor and AC motor
Brush Motor Vs Brushless Motor
DC electric motor also can be divided in two types which include brush motors
and brushless motors. For all these types, brush electric motors are most
common used. Since they are easy to build and very effective. However there is
major drawback for this type of motor. They use carbon brushes to transfer
electrical current to the rotating part, and these brushes wear over time and
finally will result in the failure of the electric motor for long run. For the DC
brushless motor eliminates the brushes, but it cost more money and requires
much more complicated drive electronics to operate. So it is very hard to make a
decision between chose between the DC brush and DC brushless in the very
beginning. After doing further research, one found out that the DC step motor
(one kind of the DC brushless motor). Since it is easier to control compare to the
other type of motor. For example, if the voltage pulse sequences were applied to
the DC motor, each of the plus will make the DC motor a given number of the
degree, move no more or less. If the DC step motor have been chose, it much
easier to prevent the overshoot than use the other motor. Another reason, DC
motor movement is predictable. For, example, if the minimum angle per step is
1.8 degree. After it move 200 step or 200 pluses were applied, then the step
49
motor will move 360 degree. If the brush motor was brush motor or the regular
brush less motor were chose, if the same sequence of pluses were applied. They
will never act the same, since they are very easy effect by the outside
environment. (Temperature, battery power and smoothness of the operation
surface) As a result, if two different brush motor were apply the same sequence
of pluses, they will not have the same result. Of course this problem can be solve
by adding an encoder and make the system become close loop system. Just like
the DC servo motor, it has the internal encoder and the price is not too expensive
for this project. So in the follow top will show what are the differences between
them how to operate them.
Different Internal Connection of DC Motors
There are two different kinds of step motor which are permanent magnet and
variable reluctance step motor. Both of them have different range of revolution,
the coarsest one can turn 90 degree each step, while the higher resolution one
can divide up to 1.8 even 0.72 degree each step. Do in the base research, 1.8
degree step motors cost around 10 dollars to 50 dollars, but the 0.72 degree step
motor cost 30 dollars to 100 dollars. With the help of the appropriate step motor
controller, the step motor can be run n half step, and some of the controllers can
handle even smaller fractional steps or microsteps.
The simple idea for the step motor run is the rotor all way trying to fixed angle
and hold the angel until the torque exceeds the holding torque of the motor after
the winding of the motor is energized, at the time the torque exceed the rotor will
turn, try to keep it the equilibrium. Following is two example showing how
different configuration works. The Figure 34 shows the typical schematic diagram
for variable reluctance stepping which has three winding. As show on the figure
the common wire connect to the positive side of the power supply and the
winding are sequentially power up. As showing on this figure, the minimum step
for this motor is 30 degree and there are 6 poles and 4 teeth for this motor, also
each winding has been wrapped around tow opposition pole. At the time the
winding number 1 is turn on, 2 and 3 is off. The rotor teeth in the vertical
direction are attracted to the number 1 winding’s poles. If the winding number 2 is
turn on and winding number 1 and winding number 3 turn off, then the rotor teeth
in the horizontal teeth will be attracted by number 2 winding’s poles. If the power
sequentially apply to the 3 windings then the power will motor can rotate
continuously. Just like the figure showing the power sequence of the 3 windings
can make the motor to rotate 2 revolutions. In here, 1 represents power on and 0
represent power off.
50
Figure 34 Variable Reluctance Motors schematic diagram
Control logic sequence of Variable reluctance Motors
(Permitted was given by Professor Doug Jones)
Two of wire will be connect to the center of the winds which call the center taps.
Usually the center taps will connect to the positive direction of the power supply,
and two ends of each winding are alternately grounded to the opposing direction
of the power supply. As show on the Figure 35, if the current flowing from the
center tap 1 to the winding end a, at this time, the top stator will attract by the
North Pole, and the bottom stator will be attract by the South Pole. So the rotor
will fix in to the corresponding position as well. If corresponding central tap and
winding end turn on in the sequence of 1a, 2a, 1b and 2b. Then this rotor will
continually rotate. The corresponding logic circuit will be as show below which
can make the motor rotate 24 steps.
51
Figure 35 unipolar stepping motors schematic diagram
Control logic sequence of unipolar stepping motor
(Permitted was given by Professor Doug Jones)
The third type is the bipolar motor, as show on Figure 36 which has the identical
mechanism as the unipolar motr, but it only have two windings without any center
taps. The motor itself become simpler than the unipolar one but it request to
reverse the polarity for each pair of the motor poles which make the drive
circuitry more complex. For the bipolar motor is very command to use an Hbridge to for the control circuit of the windings. In order to make the controlled
independently, H-bridge will allow the polarity of the power applied to the end of
each winding. So it has different control circuit as show below in order to make
the motor turn two revolutions, the plus and minus symbols for the different
polarity of the power applied to each winding.
Figure 36 Bipolar Moter stepping motors schematic diagram
Control logic sequence of bipolar motor stepping motors
52
All figures are showing the 30 degrees per step motor, for higher resolutions
motor, the rotor will have more proportion and as well as the poles. Also different
kind of the motor should apply different of the control sequence in order to make
it work correct. For example, in this project, the 1.8 degree step motor will be use,
the similar idea can apply to make it rotate, and the only different is the teeth
number of the rotor or the wings. And the relationship between windings number
w and teeth number will be invert. Also, some command configuration, such as
biporlar, bifilar motor and multiphase motor, have not been discussed here, since
they are not relate to this project.
Stepper Parameters
In order to chose the motor to meet the specified performance requirements.
Some motor parameters, such as dynamic torque, holding torque, phase
inductance, torque stiffness and rotor inertia, have been known and understand
first.
Holding Torque Vs Dynamic Torque
Holding torque is the torque for the motor to prevent a static load from pulling the
system out of step. Difference as holding torque, the dynamic torque is the
torque when the motor in motion, it relate to the current in the stator
electromagnet. Assume the full current is apply, the relationship between the
dynamic torques and holding torques can be express the equation , where Td is
the dynamic torques and the To is the holding torques. If the system has a
perfect switching time relative to the rotor position (closed-loop control need to
apply), then the system can reach the ideal level, at this time the Td will reach
the maximum value as show on equation, which can get the Tmax=0.90To. If the
motor was operate in the open loop control system, and then the minimum
dynamic torque can be calculated by the equation , which can get the
Tmin=0.63To. The typical dynamic torque is T=To(0.90+0.637)/2=0.768To. For
the 2 phase bipolar motor, the dynamic torque only 76.8% as the holding torque.
Step motor accuracy and holding Torque
The step motor accuracy not only depends on the motor construction but also
has the proportional relationship with the motor’s torque stiffness. On the other
hand that mean the relationship between the torque and the angle displacement
will be show as follow equation, which N is the number or the rotor teeth and To
is the maximum holding torque. If calculate the derivate for both side, then we
53
can get the equation where dT is change in the torque and the dq is the change
in angle. dT/dq is represent the torque stiffness, which is depend on the value of
the number of the rotor (N) and the maximum static holding torque(To). If the N
and To is improve, then the torque stiffness will improve.
Current Rise Time and Torque
The majority loss in the torque is deal to when motor running in the high speed,
the pulse switching times can become shorter than the motor phase current
rising time. So the pulse will end before the phase current rise to certain level
that can provide the request torque. In Figure 37 showing the simple example the
how the motor phase inductance and the resistance R limit the phase current rise
time, which also can express as the equation on the equation on the right side. In
here, t is current rise time also can define as L/R, V is the power supply voltage,
L is the motor phase inductance. In order to keep the rise time small, then V
need to be as high as possible, but the L and R need to be as large as possible.
As show on figure, if the voltage pulse switching time is t1, then at the end of the
voltage pulse the current can rise to the 0.63 of the maximum value. When the
voltage pulse switching time change to the t2 half of t1, at the end of the voltage
pulse, the current can only can rise to 0.39 of the maximum current. If the one
need to the optimized the performance of the system, the current has to reach 90%
of the maximum current. That mean the 2.3 times of the t1 pulse switching time
need to apply.
Figure 37 Current Rise Time and Torque
54
DC motor Selection
Conclude all the parameter and compare with the data sheet of DC motor usually,
usually there is at least two parameters will list clearly. They are the RPM rotation
per minute and Torque. And these tow parameters will be consider and cumulate
when picking up the DC motor. The first factor is the RPM can make sure the
motors to achieve the require velocity as desired for this project. The second
factor is the minimum torque requires making the plane rotate in order to change
the direction of the gun.
It is different if this project is use the small rotating frame and large small rotating
frame. If this project is used the small rotating frame it will require more RPM to
achieve the desire velocity, but it will take less torque to move it. On the other
hand, the large rotating frame, it doesn’t need higher RPM to achieve the desire
velocity, but it will need more torque. In order to determine these two factors.
Some of the requirements such as, the size of the rotating frame and the velocity
of the target move, as list in the requirement, If those requirements were known
then the minimum PRM can be determined. After the RPM is determined, the
power that need to move the gun can be determined. Furthermore, the minimum
torque needs to move the gun can be determined as well. As these two factors
were determined, then one can design how powerful of the motor need to buy for
this project. Of course, one can by a motor with very high speed and large torque,
that is sure can move the gun but this will cost the unnecessary power consume
as well as money. So the goal of this charter is determine the minimum power
consume DC step motor for this project in order to save money. And the following
will show how to choose the right size of the motor for this project.
Determine the Motor RPM
As state in the request, the target will move at the speed at 3 m/s and gun is 2
meters from the target. As showing below the target will move 3 meter in 1
second from A to B. AC is the distant of from the gun to the target which is 2
meters. One need to calculate the angle of ∠ACB, and then one can get the
angular velocity for the gun. ∠ABC=arctan(AB/AC)=56.31o. That means it needs
to turn 56.31 degree in 1 second in order to follow the target. And the RPM can
calculate by RPM=∠ABC/360*60=9.385 R/M. According to the research, most of
the DC step motors is much higher than this RPM value, so for this project, this
factor will be less consideration, but this factor can help to determine the stability
for this project.
55
Determining Motor Torque
Once the RPM has been determined, the next step is determine the torque
require for move the gun. There two difference rotating plane in this project. One
plane is directly attached to the gun and it makes the gun move up and down.
Another plane is at the bottom, it make the gun move left and right. The weights
of the gun plus the two planes have been estimate to be 20 lb; of course the
small weight will be easier to control. However, it’s better to over estimating it,
since it’s better to use the high power motor to list the light item than use the low
power motor to list the heavy item. In order to find out the force the follow
equation need to use, F=W*Crr, W is the weight, and Crr is the rolling friction.
And the power can be calculate by using equation P=F*v. Finally, the torque can
be get by equation τ=P/w. If the gun’s weight is 25 lbs include the plane 1 lbs, the
total weight is 26 lbs for the first motor need to move. According to the table
8below Crr for hard steel ball bearing on steel is 0.0002 to 0.0010. To be safe,
the maximum rolling friction 0.001 will be use in the calculation. And final result of
the torque is 8.185 ounce-force inch. This will be enough torque to move the
upper plane up and down. The second plane will have different weight which is
25lbs and the rolling friction is higher than the first plane which is 0.5. Use the
similar calculation and get the final result is 25.22 ouce-in.
Crr
0.0002
0.0010
0.0002
0.0010
0.0022
0.005
b
to 0.5 mm
Description
Railroad steel wheel on steel rail
to 0.1 mm
Hardened steel ball bearings on steel
to 0.1 mm
production bicycle tires at 120 psi (8.3 bar)
and 50 km/h, measured on rollers
Table 8 Rolling resistance coefficient
DC step motor controller
Base Step Motor controller
In this section, some elementary control circuit for difference kind of the motor
will be discussed. In order have better explanation, all the circuits assume the
motor’s rated voltage is small than the motor power supply voltage. Of course,
this assumption will limit the motor performance.
As show on the Figure 38, this is the typical controllers for a variable reluctance
stepping motors. In this figure, boxes are used represent switches; a control unit,
56
not shown, but is this project the control unite will be the FPGA board which can
providing the control signal to turn on and turn off the switch at the appropriate
time in order to spin the motor in certain way. The control unite will be control by
the computer or interface controller (FPGA board in this project), with software or
program language can directly generating the output signal (in this project a
binary adder signal will use the control 4 lead of the DC step motor as show on
Figure 39) to control the switches, However, since the current through the motor
winding cannot be turned on the off instantaneously without involving infinite
voltage, at this time the switch controlling a motor is opened, as the result, the
voltage spike will damage the switch. So the kick back voltage needs to be take
care off to prevent the damage of the system. There are tow ways to deal with
this problem, one is bridge the motor winding with a diode, that can be able to
conduct the full current through the motor winding, but it will only conduct briefly
each time the switch is turned off, as the current through the winding decays.
Another way is bridge the motor winding with a capacitor. At this time when the
switch is closed, the capacitor will discharge through the switch to ground. When
the switch is opened, the stored energy in the motor winding will charge the
capacitor to a voltage significant above the supply voltage.
Figure 38 Variable Reluctance Motors Controller
(Permitted was given by Professor Doug Jones)
57
Figure 39 Unipolar Motors Controller
(Permitted was given by Professor Doug Jones)
As mention before the full step the plot of torque versus angular position for the
rotor relative to some initial equilibrium position will be approximate a sinusoid.
So how would be a half step or micro stepping looks like. If two motor windings
have been powered simultaneously that will produce a torque versus position
curve that is the sum of the torque versus position curves for the two motor
windings taken in isolation. For a two winding permanent magnet motor, the two
curve will be S radians out of phase, and if the currents in the two winding are
equal, the peak and the peaks and valleys of the sum will be displaces S/2 radian
from the peaks of the original curve, as show in Figure 40 this is the basis of half
stepping, the relation shipping between the single winding holding torque h1 and
the two winding holding torque h2 will be express as follow h2 = 20.5h1. This
assumes that no part of the magnetic circuit is saturated and that the torque
versus position cures for each ideal sinusoid. The figure 41 is the torque versus
angular position for microsteping control which allows even smaller steps by
using different currents through the tow motor windings and the following
formulas showing the key characteristics of the composite torque. h = ( a2 +
b2 )0.5 and x = ( S / ( /2) ) arctan( b / a ) where a is torque applied by winding
with equilibrium at 0 radians; b is torque applied by winding with equilibrium at S
radians; x is equilibrium position, in radians.
58
Figure 40 Bipolar Motors Controller
(Permitted was given by Professor Doug Jones)
Figure 41 Half Stepping and Micro stepping
(Permitted was given by Professor Doug Jones)
Limiters of the Micro stepping
The micro stepping can reduce step angle to smaller. However, it also has its
limited. The first limit is the dead zone limited. For example, Figure 42 presence
of a dead zone has a significant impact on the utility of micro stepping! If the
dead zone is x° wide, then micro stepping with a step size smaller than x° may
not move the rotor at all. Thus, for systems intended to use high resolution micro
stepping, it is very important to minimize static friction. The second problem
involves the non-sinusoidal character of the torque versus shaft-angle curve on
the real motor which is call the detent effects. The motor is at its expected
position at every full step and at every half step, but there is a significant
positioning error in the intermediate positions which is showing in figure The third
problem arises because most applications of the micro stepping involve digital
control. As the result, the current through each motor winding is quantized,
59
controlled by a digital to analog converter. For instance, if PWM current limiting
circuit is used, the current through each motr winding is not held perfectly
constant, but rather, oscillate around the current control circuit’s set point. The
effect of this quantization is easily seen if the available current through one motor
winding is plotted on the X axis and available current through the other motor
winding is plotted on the Y axis which is showing in the Figure 43
Figure 42 Dead zone impacts
Figure 43 Micro step Accuracy
60
Step Motor Control Design and Difficulty:
As show on the Figure 44 is the Unipolar step motor control sequence. If the
circuit is turned on in the sequence of 1a , 2a, 1b, and 2b, that will make the
motor turn clockwise. For reverse direction of the motor one just need to make
the circuit turn on in the sequence of 2b, 1b, 2a, and 1a. In the step motor which
plan to use can be divide to two group, white, orange , and blue is one group
Black, yellow and red is another group. If connect the circuit from orange, red,
blue and yellow then the motor will turn clockwise. In the reverse direction (yellow,
blue, red and orange ), the motor turn will turn counter clockwise.
Figure 44 Unipolar Step Motor Control Sequence
(Permitted was given by Professor Doug Jones)
So if one connect the circuit like Figure 45 below, the common wires connect to
the power supply and winging wire connect to the one end of the switch. If one
turn on the switch as shown on the sequence above. Then the step motor can
turn the corresponding direction as state above. Digital switch is one of the
options for this design, but the digital switch is expensive, For example, 2A
LTC1477 is cost $6.95 for switch. So the total needs to cost $27.8 dollars to
accomplish this design.
61
Figure 45 Step Motor Controller wind Winging Location
(Permitted was given by Professor Doug Jones)
Another solution is use Power Mosfet to design the digital switch. The circuit will be as
Figure 46 shown below. The IRF 1324 Power Mosfet has up to 24 minimum breakdown
voltages. Also it has 195 A continuous Drain current. It can support enough voltage as
well as the current for the motor use in the project. The drawback of this design, the
FPGA output did not have enough output current to turn on the mosfet. To solve this
problem, the comparator or Amplifier can be use to solve this drawback. LM339
comparator has been considering for this design. Since it can step up 4 inputs just use one
IC, it has 1mA forward current and up to 20 mA reverse breakdown current. Most
importance is the cost only 1 dollars for online price.
62
Figure 46 Digital Switch Circuit
Real Time Control
Open loop Vs close loop motor control
The real time control means the control system need to know when the next step
should be taken and how should it operate. This question needs to depend on
the application. The similarities between different applications are sufficient to
justify the development of fairly complex general purpose stepping motor
controllers. Most of the motor control may be based on open loop or closed loop
control models. For example, figure is a control block diagram with shaft encoder
to provide the feed back to the control system. If the shaft encoder is rotated into
a position where the output of the shaft encoder translates to a control vector that
holds the motor shaft in its initial position, the motor shaft will not rotate of itself,
and if the motor shaft is rotated by force, it will stay wherever it is left. This
position of the shaft encoder is relative to the motor as the neutral position. This
insures this control system will always produce the maximum torque the motor is
able to deliver at any speed with the limit. Open loop control is most often use in
the step motor control system since the advantage of the step motor is can know
the step what have done and predict the next step. As show on figure 47 , the
basic principle is very similar to the close loop, but the different is the feedback
loop is broken, instead use the simulation model of the response of the motor
and load to the control vector. At any instant, the actual position of the rotor is
unknown. However, the position can be predicting base on an assumed rotor
position and velocity. So the simulation model has to been constructed in order to
63
generate the simulated shaft encoder. As a result, the figure will have the same
accurate as the motor control with a closed loop system.
DC Motor Controllers
The first motor controller has been considered is Phidget Unipolar USB motor
controller which as shown on the Figure 47. It has the usb port that makes it
easier to be control by a computer. External power supply apply insure the power
operate in its request power range. It has 2 motor connect in each side make it
possible to control up to 4 unipolar step motors. Also, Phidgets also offer its own
Application Programing Interface to be download that make it extremely easy to
operate in by the Window, Linux and Mac OSX operate system. It also support
VB6, VB.NET,C#.NET, C#, C++, LABVIEW, PYTHON, MAX/MSP and COCOA
and so on different kind of popular computer program language. It offers the user
the maximum amount of freedom to operate this board. It looks like it’s the
perfect solution for this project to control a DC motor. However, the drawback for
this controller is the price, it cost $73.05. It’s over the budget for this project.
Figure 47 Phidgets Unipolar USB 4 Motor Stepper Controller
Step Motor Vs Servo Motor
According to the research, step motor and the servo motor are very often use for
position control. Compare for both step motor and servo motor. For the step
motor, it is the open loop control, but the servo motor use the close loop control
that can determines accuracy and resolution.; for the step motor, it can continue
rotate but the servo motor usually has the rotation limit; step motor usual is a
brushless motor it has long life than the servo motor; servo motor has higher
64
output power relative compare to the step motor. So over all, the servo motor is
better choice for this project.
Nylon gear are very common in servos, they extremely smooth and is not easily
to wear. Also, they are very light weight. But deal to lack in durability and strength.
They haven’t been use for this project. Instead the nylon gear servo, metal gear
servo was use, although the weight is heavier than the nylon gear, but the side
load is much greater. Also, the drawback of this kind of gear is it will be slowly
wear or lost. As show on the Figure 48 below, SAVOX SC-0252 is used in this
project, this servo motor is only require 4.0 to 6.0V to operate and the weight of
the servo motor is only 49g, but the torque of the motor is up to 10.5 kg-cm, and
it can run at 0.19 s/60 degree. Compare to the step motor， sanyo Denki 85004，
has been consider，which weight is 600 g, 10 time of the servo motor that has
use now. and the this step motor only has 55.3 OZ-in of the torque.
Figure 48 Savox Servo Motor
Motor Control
The real time control means the control system need to know when the next step
should be taken and how should it operate. On the Figure 49 is a control block
diagram with encoder to provide the feed back to the control system. When the
encoder is rotated in a position where the output of the encoder translates to a
control signal to hold the motor in the initial position, then the motor will not move
by itself. If the motor is rotated by external force, it will stay wherever it is left
before so that the system can provide the maximum torque to the motor. For
open loop control is very often use in the step motor control system since the
advantage of the step motor is it can predict what next step should be. As show
on the Figure 50 , the feedback loop is broken. So the actual position of the rotor
is unknown, however, one can construct the simulation mode to generate the
encoder, so that it can predict the motor position as the close loop system did.
65
Figure 49 Close Loop Control
Figure 50 Open Loop Contol
Digilent PmodCON3 Servo Connector Module Board has been use for motor
control. This board is not exactly a motor controller. It is only a connector board
and makes the Digilent system board easily control the servo motors when the
PWM signal generate by the system board. This connector board has 4 sets of
the pins can connect up to 4 servo motors. Also, it has 1 terminal power supply
block, which can power up 6V source which can insure sufficient power to run the
servo motor at anywhere from 50 to 300 ounce/inch of the torque. There are 6
pins header for this connector board and it compatible for most of the Digilent
system board as well as the BASYS FPGA Board which is use in this project. In
this case, the BASYS FPGA Board will act as a motor controller to provide the
control PWM signal to the motors. The PWM signal can be use to control the
direction and degree of rotation. The range of the pulse signal is from 1ms to
2ms. 1ms plus signal will cause the motor to turn all the way in one direction.
2ms plus signal will cause the servo turn all the way in other direction.
Figure 51 PmodCON3 Servo Connector Module Board
66
Motion Control Concussion
As show on the Table 9 below, use the servo motor as who motion control has
the advantage of low cost, ( save up to 50 dollars in the control circuit ). Servo
motor also has higher torque but lower power consume advantage. The most
importance, it is servo motor has much lower weight compare to the step motor.
Step Motor
Servo Motor
Cost
$19.99
$23.99
Torque
120 oz-in
146 oz-in
Weight
34.3 oz
1.8 oz
Power
24V
Controller Cost
$70
2V to 6V
$10
Table 9 Step Motor Vs Servo Motor
Sensor
FOV request for this camera
After decided what kind of the camera that can be used for this project, the next
step is to determine what charities needs for the camera. For the charities one
need to determine mainly 3 parts. The lens focal length and camera revolution. In
order to determine these charities, some requirements need to be found out,
such as FOV (field of view), Resolution and the Working distance.
First one need to determine revolution, according the research in order to make
the accurate measurement. This project needs to use of two pixels to represent
the smallest feature we need to detect. The relationship between field of view,
revolutions and smallest feature size as equation show below
Sensor resolution (S) = (FOV / size of smallest feature) x 2
67
Also the relationship between the Focal length, File of view sensor size and
working distance has the following relationship
Focal length x FOV = sensor size x working distance
By research we know that the lenses are manufactured with a limited number of
standard focal lengths. Common lens focal lengths include 6 mm, 8 mm, 12.5
mm, 25 mm, and 50 mm. when the lenses with short focal lengths (less than 12
mm) produce image with a significant amount of distortion that will affect the
sensor’s accuracy. As request in this project the working distance is 3 meters
and the smallest feature size is 0.1 meter. The minimum FOV can be calculate
as follow. Assume the camera plane in the center of the target screen. And this
screen size is 2 meters by 3 meters. The minimum FOV for horizon is 2*tan1((3/2)/10)=17.06 degrees. The same way to calculate the minimum FOV for
horizontal is 2*tan-1((2/2)/10)=11.41 degrees.
Knowing what kind of the camera and what minimum requiems for this project.
Next goal is to choose the specific camera for this project. During the research
each camera has a field of view, resolution and frame rate that has to be taken
into account it did not restrict the choice because price and availability affected
the decision. In most case field of vies is immediately available and therefore has
not been the fact of affect the decision. A lot of the camera with this power and
within the affable rang of this project. However, most of the retailer don’t offers
enough technique information of the cameras but only with their detail project,
there is not enough documentation that limits the choice of this project, which
may have the risk for getting the information for this design. Due to the extensive
searching, Table 10 shown below, the cheapest digital camera also it has the
higher resolution. However, this camera does not have development style
connectors so it has been eliminated from this project. Another cameraC3038
has exactly the data output as for the video decoder board; however, the design
is complicated. So it also had been take out from this project. The last one is the
Swna sw-p dscex, which has NTSC, it can directly plug in the video decoder
board. Of course, the program needs to be designed in order to input the frame
from the camera.
Resolution
FOV
Lens Length
Frame Rate
Data out
Power
requirement
IMC CH-8028
800*600 Pixel
44.5 *57.4
4.0
30f/s
YUV=4:2:2,
RGB=5:6:5
C3038
356*292 Pixel
34.4*20.7
3.6mm
15f/s
YCrCb
4:2:2,
GRB 4:2:2, RGB
Swan sw-p dscex
380 TV line
45*100
6.0 mm
30f/s
YCrCb
4:2:2,
GRB 4:2:2, RGB
3.3VDC
3.3 VDC
8-12VDC
68
Interface
Cost
USB
9.99
IC2
31.98
NTSC
32.99
Table 10 Camera comparison
Power supply for camera and motors
When the power supply transform from the AC wall outlet, it out put the 120 V DC
volts, but for the camera (9v) and DC step motors (24V) which is much smaller
than the power supply than the camera and motors need. So the voltage dividers
may be need to this project. In order to ensure that the camera and DC step
motors have the require voltage and current need during operation one may use
the DC regulator for this project. Also the other reason this project need to use
the DC regulator because it can protect the power supply form damage if one
component had an internal fault.
For voltage one can simple use the similar circuit below to design the low
voltage power supply. IC linear voltage regulators come in a variety of sizes,
input and output ranges. One kind of regulator that fix this project is LM2937 3.3
voltage regulator made by National Semiconductor. This regulator is a small
three-pin IC. It has a maximum range of 26V DC input. This project will use one
for camera and two for motors. Linear regulator can easy build by Op-amps,
building a linear regulator from a discrete components allows for more design
freedom, if the time available
Camera Position Configuration
Figure 52 Camera Position Configurations
Difference camera position configuration has been developed and used, from the
Figure 52 above, VM1 is monocular eye-in-hand mode, VM2 is monocular standalone mode, VM3 is binocular eye-in-hand mode and VM4 is binocular standalone. Different configurations have difference effect on the visual system. For
example, the first one, VM1 is one of the most common use configurations. The
69
camera is attached to the robot’s hand. The task of this configuration is to try to
move the camera and get the image achieve the predefined image positions. The
current and the desired position image will be save when the visual servoing
starts. The second one is VM2 which also a one camera configuration, but the
camera is set aside of the servo, so the camera in here use as a global sensor
for the system. It usually required more accurate camera to perform the operation.
Every time after it tasks the target it needs to retrieve to the initial position, so it
can’t offer the smooth tracking procedure. But it can use for the suitable grasp
operation.
Different from one camera arrangement, VM3 and VM4 has two cameras in a
stereo arrangement. This configuration cans simply estimation the depth without
using the explicit models jus t like the one camera configuration. It can provide
complete 3D information about the scene. But the trade off is it require twice or
more computational time for each iteration. VM3, binocular eye-in-hand mode, is
not very often use for servoing tasks since its limit base inaccuracy and the
difficulty of reconstruction. The last configuration, VM4 binocular stand-alone is
very often using in the servoing tasks. Since it’s not only can easily make the
baseline long enough so that the depth estimates are become more accurate
compare to the eye in hand stereo configuration. Also, this configuration
increases the field of view which can make it easier to move robot and the target
the object simultaneously.
The proposal of this project is first detect the color object then calculate the
position of the object, then control the motor to point the object. So that can
haveenough time to process the calculation of the position use the monocular
eye-in-hand mode can simplifythe design.
Image collection and store
As mention, the camera captures image and storing it in the computer as using
for analyst. How do the pictures save in the computer? Every image that saves in
the computer will represent as pixel in the computer which is a tiny dot of color in
the computer screen and it containsthe color and locationinformation of the
object. The image also have a set number of pixels per size of the image known
as resolution, mean the number of pixel in square inch of the image. The same
as camera, the higher resolution means there are more pixels in a set area,
resulting in a higher quantity image, the bad thing about the higher quantity of
image require higher performing of processor as well as higher. That is also limit
of the project, so in this project the bad and white camera has been chose in
order to reduce processor perform time.
70
Image is stored in 2D matrices, which represent the location as well as the color
of the pixel. All image have X and Y component. If the image is only black and
white, then the data will store as 1 or 0 binary number. If the image is color image,
it will store as a set of number. So, the less color involved the faster the image
can be process. Below is showing the revolution of 5*6 black and white image
data store in the computer. And it require 5*6*1=30 bits to store memory
000110
101110
011110
010010
The other example will show a grayscale (8bit) image, and it requires 5*7*8=280
bits memory.
0 0 55 255 0 0
55 0 252 255 255 0 0
0 255 255 55 255 0 0
0 255 0 0 255 0 0
It is easy to see increase the revolution and information of the image it require
addition memory to for the process. Otherwise, it will slow down the speed.
Movement Algorithm
General Analyst
There are five steps in the movement algorithm, they are the frame input, color
detect, threshold, centroid and motion vector, as shown on Figure 53. The
frames will detect by the camera in the sequence order and the first fame will be
store in the memory as background fame. And the subsequent will compare to
the background frame to determine the next operation. If the differences of the
new frame with the background frame within the acceptable value, then the
system will input another new frame. Otherwise the image will be past to next
step to be threshold. After that, the centroid process will be operating to find the
center mass of the object. Then the comparison will be process in order to find
out the motion vectors between the center of the object and the center position of
the whole image. Final, the pan and tile vector was finned to determine the motor
movement.
71
Figure 53 Schematic diagram of the real-time
Closed-loop tracking algorithm
Treshold and Centroid
After using the background subtraction to eliminating the background and the
stationary object, the tresholded can be simplify by the binary image as show
below equation. Where “a” is range of the treashold value and it also determine
the accuracy of the tracking algorithm. The next step is to centroid of all the
pixels about the threshold is calculated by the equation, where W and H is
represented the width and the height of the frame. Then the central mass of the
moving object is determined.
1
I t ( x, y )  
0
I d ( x, y )  
I d ( x, y )  
W 1 H 1
X c    x  I t ( x, y )
x 0 y 0
W 1 H 1
Yc   y  I t ( x, y )
x 0 y 0
Motion Vector Determination
The final step is to determine the motion vector. To achieve this, the perspective
model for the camera will be constructing as below on figure. The original is the
location of the camera or the initial location of the gun mouth. If the gun initial
point is at the origin of the XYZ plane, so what angle it will change if the object
moves to the point Q. As show on the Figure 54, X is the distance that the object
moves in the horizontal direction. And Y is the distance the target move in the
72
vertical direction. And assume Z is the distance from the camera lens to the
target plane.
Figure 54 Motion Vector Determinations
Gravity effect on the vertical direction
The bullet more or less will affect by the gravity and this effect must be
consideration on this project system. As show on the photo above the initial
velocity is V, if we decompose it in to X direction Vx and Y direction Vy. t= X/Vx
(Vx =Cos(a)*V and Y without the gravity the bullet should shoot at the value of
Y=t*Vy (Vy=Sin(a). However, in the really situation the vertical distance the bullet
travel will effect by the gravity. So the actuate distant the bullet in the Y direction
will be Yac=Y-1/2*g*t^2. For this project, the distance the gun from the target is 3
meter and the initial velocity is 33 m/s. So the effect according to the angle
change will be as show below on Figure 55.
Figure 55 Gravity Effect
73
VIDEO DECODER
General information
The Video Decoder 1 Board (VDEC1), which shown on Figure 56, use the multi
format video decoder chip, ADV7183B, can convert most of the analog video
signal to the digital signal. This board offer the different video input format such
as S-video inputs, component and composite, with the simple converter video
socket use, then the camera can be attach to this board and operate. Also, the
VDEC1 offer the offer the perfect solution for the Diligent board which Hirose FX2
connector, simple plug in, then can be operate properly.
Figure 56 Video Decoder 1 Board Inputs and output connector
Limitation and Advantage
The video input format can be specified by the I 2C bus. In this project, the C3038
on board camera will be use. So the output format should use the YCrCb 4:2:2.
The video capture design function should program on the FPGA board. Also,
video stander information, such as color and brightness, will be detect by the
camera, then transfer by the VDEC1 board, and finally analyze by the FPGA
board. There is some limit of theVDC1 board, for example it limit in the input
signal voltage is 3.3 V analog signal and it is not support the 5 V input signal, but
this limitation can be solve by using the amplifier logic circuit to decrease the
signal voltage. The advantage of using this board since it has ADV7183B chip
which can oversamples the analog input by the factor f 4 also it has up to 54MHz
sampling frequency that reduce the need of the requirement for an input filter.
However, the antaliasing filter will be use in order to optimal the performance.
74
The simple emitter follower circuit will be use to implementing the buffer for the
antliasing filter, which reduce the cost for the filter and simplifier the design.
Video Selection
The VDEC1 is has 4 bits selector VID SEL[3:0] bits allows the user to chose the
digital core into the requested video standard. In default, the VID SEL[3:0] will
sent to 0000 which can auto detect the supporting format such as PAL, NTSC
and SECAM. When the system set in the autodetection mode, it will picks the
closest video standard and read back the through the status resister. The Table
11 shows the Global Status Registers information. And the following is show the
example of using SECAM525. If the AD_SEC525_EN was set to 0 (default), the
autodetection function will be enable. If the AD_SEC525_EN set to 1 the
autodetection function will be unable. If the AD_N443_EN was set to 0, then the
auto detection of NTSC will be disable, otherwise it will enable. If the
AD_P60_EN was set to 0, the autodetection mode of PAL system with 60Hz field
rate will disable. Otherwise, it will enable the detection.
VID_S
EL
0000
0001
0010
0011
0100
0101
0110
0111
1000
1001
1010
1011
1100
1101
1110
1111
Description
(default) Autodetect (PAL BGHID) <–> NTSC J
(no pedestal), SECAM
Autodetect (PAL BGHID) <–> NTSC M
(pedestal), SECAM
Autodetect (PAL N) (pedestal) <–> NTSC J
(no pedestal), SECAM
Autodetect (PAL N) (pedestal) <–> NTSC M
(pedestal), SECAM
NTSC-J (1)
NTSC-M (1)
PAL60
NTSC-.43 (1)
PAL-B/G/H/I/D
PAL-N (= PAL BGHID (with pedestal))
PAL-M (without pedestal)
PAL-M
PAL-Combination N
PAL COMBINATION N (with pedestal)
SECAM
SECAM (with pedestal)
Table 11 Global Status Registers for VDEC1
75
Power system
Power System Overview
The power system is the main component in the coil gun. In the coil gun it will
provide all electrical components with electricity throughout the whole circuit. It
will provide currents to power switches, charge a capacitor bank, and power DC
servomotors with controls used for automation and target acquisition. For the
high voltage and high velocity gun the power system will likely be the most
important part of the project. This subsystem will make sure that initial
specifications for the project are met. With a careful design of this system the coil
gun should be able to greatly exceed the current specified constraints. It will
encompass all that has been learned in the years as engineering students at the
University of Central Florida. The knowledge of electrical and magnetic fields will
be refreshed and tested with this project, specifically this subsystem.
The power system will have both types of sources AC and DC.
It will be
necessary to step up voltage through a transformer and convert AC to DC using
circuit elements to create a power convertor. This DC power generates a direct
current to then charge the capacitor bank. This of course is just the general
concept. It should be kept in mind that the design of a strong triggering and
switching system has to be established to account for the high currents and
voltage. The energy in the coil gun is being stored in high voltage rated
capacitors. Charge resistors will also be implemented in the circuit. The charge
resistor together with the capacitor bank will establish the time needed to charge
the coil gun before firing. An overview block diagram of the power subsystem is
shown in Figure 57 below.
76
Figure 57 Power Subsystem Block Diagram
Power Supply
The power supply for the gun can be any DC source. There are numerous
options for DC sources. One option is to use a power supply from a computer.
High power computer power supplies are now fairly cheap and easy to find.
Computer supplies use an 115V or 230V AC input. A nice characteristic of a
computer power supply unit is that it already converts AC to DC power. This
characteristic eliminates the need to construct a full-wave rectifier in the coil gun
circuit. In the coil gun the problem that one would run into using a computer
power supply unit deals with the fail-safe of the unit. A capacitor bank cannot be
hooked up to a computer PSU directly. The reason for this is that the voltage
must be kept above a certain value to remain on. If it goes below that value it is
designed to automatically shut off. In the coil gun, having the PSU and capacitor
bank hooked directly together would be an issue because the voltage will vary
with respect to time. The discharge of the gun or the PSU being turned on would
cause the PSU to turn off due to the voltage being dragged too low. The only way
77
to rectify the issue is to use a low resistance power resistor in series with the
capacitor bank. The disadvantages of that are slower charging capacitors. This
would add to the already slow charge time.
A car battery charger is another method that can be used to charge capacitor
banks. Most car batteries charge at slow rates because manufacturers are
worried about the safety of persons using the product. It might be possible to
manipulate the circuit into charging faster. They are cheaper than computer
supply units. Battery chargers are the worst way to charge a capacitor when
timing is a big concern. The time it takes to charge a capacitor is too long
especially considering its capacitance. For the coil gun to be able to continuously
be able to exert a projectile this could never be an option. It is too slow and very
time consuming. The choice of repeated fire would likely not be an option.
The best choice is likely AC power. It is inexpensive. With AC power the circuit
can be manipulated a lot more. It will be easier to add rectifiers and boost
converters if necessary. The time the capacitors banks take to charge will also be
a lot quicker than using a normal car battery or standard power supply unit. The
great thing about AC power is it is readily available. There is no need to worry
about current running out like a battery. If the coil gun needed to power any
additional controls or circuit devices it can easily be added with an AC source. If
the additional devices needed direct current they could just be connected to
another full wave bridge rectifier connected to the AC power source. The AC
power from the standard outlet is 120V. The frequency for the analog signal is 60
Hz here in the United States. The frequency is not any concern for the coil gun.
After the current goes through the convertor, only a small portion of the signal
passes through. This small portion is not enough to make a big difference.
Usually capacitors are put into circuits to eliminate the problem of a signal
passing through. The capacitor bank has more than enough capacitance to
eliminate the signal completely.
Energy Sources
It is possible to use an electrochemical cell battery for an energy source. The
best choice of a battery would be one that has real low internal resistance. The
lower the internal resistance the higher the power transfer from the battery. In a
coil gun maximum power transfer is a desired characteristic. Lead-acid and
nickel-cadmium batteries all have low internal resistance characteristics. The
disadvantages for a coil gun include longer charge time and lower current.
Battery voltages are DC power sources so varying voltages is not an option. This
could hinder the goal of reaching at least 100 ft/s for projectile speed. The
78
importance of being able to vary input voltages is the key to increase current
through the solenoid.
Capacitors are a favorable choice because they easily store large amounts of
energy. In a coil gun all the energy it needs can come from a capacitor. To give
the coil gun more energy and current more capacitors can be added to increase
capacitance. The weight of a capacitor is dependent on its capacitance. Normally
capacitors are not too heavy. For the capacitance in the coil gun capacitor bank
weight should not be an issue. Capacitors for an energy source are great but
expensive at high voltage ratings. Specification for the gun dictate 300V but
capacitor rating will likely be 400V if chosen to account for variations in voltages
from the power source. There are many types of capacitors to choose from that
will be discussed later in the report.
Using an inductor is also a possibility. Inductors possess the similar characteristic
of a capacitor in that it can store energy. To use an inductor in a coil gun a
number of factors need to be addressed. For the inductor to act as a storage
system an additional inductor is needed. The inductors are set up like a
transformer. The first inductor stores charges like a capacitor bank, and the
second inductor is the output. It outputs the stored energy. The first inductor or
primary windings induce a voltage on the secondary windings through a
magnetic field creating an output current. There are disadvantages and reason it
cannot be used in the coil gun that is being built. First off it is very bulky. The coil
gun that is being designed is supposed to be lightweight. The addition of
something of this magnitude added would add too much weight to the design.
Another issue that arises is keeping the charge stored on the primary windings.
Unlike a capacitor to store the energy in the primary windings a constant current
needs to be present. Despite the disadvantages if weight was not an issue this
would be perfect. The advantages of using this in a coil gun system are that the
desired output current and voltage can easily be reached. In a transformer the
desired voltage and current can be attained through adjusting the number of
windings and input voltages.
The best energy source to use in a coil gun is a capacitor. The capacitor was
chosen because of it is versatility. Energy from the capacitor can be increased or
decreased with the addition or subtraction of capacitors. There prices are
expensive for high voltage rating but there are many types and sizes to choose
from. A majority are also light weight, a benefit needed to keep the gun light. The
transformer or inductors as an energy source was hard to consider because of its
size. It had every advantage over the capacitors except for it being bulky. The
one thing that stood out about the inductors was their ability to not only store
energy like the capacitor but be able to step up or step down voltages. For a
higher powered coil gun where extremely high voltages were a concern this
would be a great energy source to use. The electrochemical battery as an energy
79
source as energy was quickly shot down after thorough research. A main
concern before the thought of even designing a coil gun was charge time.
Charging a battery takes fairly long and would not reach the goals of the project.
The chemical in the battery requires many safety precautions. Batteries can
explode if they are charged after the point of energy capacitance. If for some
reason the person charging the battery accidentally mismatched the positive and
negative terminal the battery may explode. Batteries are good sources of direct
current but not for long term use. Too many problems arise from repeated use of
a battery. It was quickly scraped from the design process. Although batteries as
energy source without capacitor banks was not feasible. It is possible to use a
battery to charge a capacitor bank instead of using alternating current. Charging
the capacitors through a battery does have some advantages. A circuit with a
battery and a capacitor bank do not need a full-wave bridge rectifier to charge.
This is because of the fact that batteries supply a direct current that does not
need to be converted to another form of power. The only thing that would maybe
be added to the circuit to deal with higher voltages is a boost regulator. This is of
course if the battery being used had a small amount of voltage. Multiple batteries
would have to be on standby incase the battery ran out. The coil gun would likely
eat up a lot of energy from the battery. Using a battery to charge the coil gun
might be good if the user wants the gun to be completely portable. Always having
to change batteries would become a nuisance. Not to mention when the battery
is running low it might be somewhat hard to calculate the charge time of the
capacitor banks. The variable amount of charge in the battery might also affect
the way the projectile is shot through the barrel. In the current coil gun
specifications the projectile must be shot at a specific speed all the time.
Anything deviated from that speed would present a problem in consistency This
is the reason that an AC outlet has to be used. With an AC outlet there is no
need to worry about deviations due to current. This is because the current
coming from the outlet is always 120 volts at a frequency of 60 Hz. When the AC
power is turned into DC power the current is still constant. The constant current
is all that should be worried about when supplying current to charge the
capacitors. From this constant current the charge time will always be the same.
Solar Cells and Panels
The current design of the coil gun prohibits it from being portable because of the
dependency of an electrical outlet. Since the gun is being designed to be efficient
and economically friendly the use of batteries will not be considered. If time
permits the gun will be made so it can be portable as well. Solar cells and panels
can slowly charge devices over a long period of time. They slowly charge
devices because a single panel can only produce a small limited amount of
power. In order to create a very large amount of power numerous panels will
80
need to be used. The solar cells use energy that comes from the sun in the form
of photons to produce electricity. Most solar cells are created using crystalline
silicon wafers. These cells are very sensitive and brittle. They must be protected
to ensure quality. Most solar panels are protected from the natural elements of
rain, wind, and moisture by glass. The glass protects the superstrate and the
substrate of the cells. Solar cells can be orientated in panels in a couple of ways.
One way is positioning the cells in series. If solar cells are placed in series there
is an additive voltage for each cell that is added. For more current transfer the
cells can be orientated in parallel to achieve maximum current. Earlier in the
report it was mentioned earlier about storing a coil gun with large amounts of
energy can be dangerous. For this reason if the coil gun were to have solar cells
to charge the gun when the gun is not in use there needs to be a switch to
control this method of charge. The gun should be charged only when it is going
to be fired sometime in the soon future. A switch would enable it to charge. When
the gun is not going to be charged without an electrical outlet the switch should
be turned to the off position. The gun will only be able to shoot one cycle
because the charge time with solar panels is more than a thousand times the
charge time of an electrical outlet. To charge the coil gun using solar cells
requires it to be exposed to sunlight for many hours. Even with ten solar panels
it would take a long time to charge the capacitor bank. This is okay because the
solar cells are only a secondary source or it can even be considered to be a
backup. Despite the drawbacks of solar cells there are many advantages. Solar
cells use natural elements to produce energy. They do not use non-renewable
resources therefore they are environmentally friendly.
They also have long
lifetimes as long as they are properly protected from moisture. Moisture corrodes
metal contacts and interconnects. The cells can overheat but the solar panels
usually consist of diodes that prevent overheating. Heat in the panels reduces
operating efficiency as well. Since the energy source is sun light there is going to
be heat either way.
AC to DC Power
The voltage in the coil gun will come from the standard 120V 60 Hz wall outlet.
The 120V from the wall is an alternating current. This will be used to charge the
capacitor bank. An alternating current is a bidirectional sinusoidal power source
with voltage and current being out of phase. In order to effectively charge the
capacitor bank the alternating current will be stepped up and need to be
converted to a direct current. Alternating current cannot be used because a
capacitor blocks directs currents but allows AC to pass through. The direct
current is blocked and charges the capacitor bank because a direct current
possesses infinite reactance. The conversion from an AC signal to a direct
current will need some form of a converter. For conversion a simple circuit is
81
needed. Almost all devices in a household use direct current to operate. The only
real use for AC is to transfer power from one place to another. Using direct
current for transferring power from long distances is very ineffective because the
currents and voltage are always in phase.
An easy way of converting AC to DC is to use a rectifier. The process by which
the AC is being converted into DC is known as rectification. A rectifier consists of
one or more diodes in a circuit connected in series with a resistor. Using only one
diode in series with a resistor is better known as a half-wave rectifier. A half-wave
rectifier filters either the top or bottom of the input and outputs either the upper or
lower portion of the wave. It is very ineffective for power transfer in a coil gun. In
a coil gun the power transfer should be maximized. It is of better advantage to
use a full-wave rectifier. A full-wave rectifier is similar to a half-wave rectifier
except it has more diodes in a different configuration. The configuration that the
diodes are formed in is known as a bridge. The full-wave rectifier as you can see
has four diodes connected to a resistor.
A full-wave rectifier does not filter out an AC signal completely. It only filters out a
majority of it. Rectification does not stop at a full or half-wave rectifier. A
capacitor is usually introduced into the circuit in parallel to create a more true
constant direct current. The larger the capacitor introduced the more constant
direct current there will be. Since all of the current is going into a capacitor bank
anyway, it may be possible to be able to get by with just a small value capacitor.
This addition of the capacitor does not filter out the sinusoidal signal completely
and may present a problem later with the triggering system. Hopefully the analog
signal is real minor. In a coil gun the issue is not of any major concern. The
problem can easily be addressed by additional circuit components if need be.
For the AC to DC power conversion in the coil gun a device is needed. There
was a choice of either buying an existing AC to DC converter or actually
constructing one from scratch. Buying one from the store is no fun. Constructing
one would be more of a challenge. Four diodes will be put in a bridge
configuration. Each diode will have a current rating of twenty amperes. In the
circuit all diodes will have the cathode on the right side of the diode facing
towards the outputted direct current. The left side of the bridge is the alternating
current and is connected directly to an outlet. The right side of the rectifier is
outputting direct current and is connected between the triggering system and
capacitor bank. No boost regulators of any type will be connected to the circuit.
There should be enough direct current from the conversion to charge the
capacitor banks in a reasonable amount of time. The current after the full-wave
rectifier will be measured to make sure enough current exists after the power
transfer. If the charge of the capacitors takes too long it may be necessary to
step up the voltage because the current is too low or the charge resistor has too
much resistance. Another way to get more current would be by possibly using
82
diodes with higher current rating, which isn’t an issue because they are
inexpensive. The temperature the diodes are operating could affect the power
transfer as well. The heat dissipated across the diodes in the full-wave rectifier
will also be monitored to make sure the current rated value of the diode chosen is
the correct one. A half-wave rectifier was not an option because of its poor power
transfer characteristics. It was very inefficient when compared to the full-wave
bridge rectifier. To use a half-wave rectifier would have been too much of a
hassle. The current and voltage would need to continually be stepped up from
transformers.
Stepping up AC vs. DC Power
Stepping up AC or DC voltages may be required to achieve the goal of having
high voltages throughout the coil gun circuit. This makes high voltage power
transfer an important issue. For a coil gun there should be minimal power loss
throughout the transferring of power. Most power loss can be controlled through
added devices or simply by choosing the correct circuit elements to transfer
power. The debate over whether stepping up AC or DC power has always been
a debate. Years ago Thomas Edison and Nikola Tesla fought a war on which
currents were better alternating current or direct current for electrical power. In
the paragraphs below the difference between the two will be discussed to find out
which is best for a coil gun system.
The process for stepping AC power is a simple concept to grasp. AC power uses
a transformer to either step up or step down the voltages or currents. In the coil
gun minimal power loss is a positive. In the coil gun circuit the voltage coming
from the outlet is low. The voltage from the outlet is only 120 V which is not very
high. In the coil gun circuit a desired characteristic is to have low voltage and
high current. A transformer usually has more power loss when there is low
voltage and high current rather than high voltage and low current. Transferring
high voltages for a coil gun is complicated. The safety would be an extra concern
as well. It does not seem too complicated except for the added weight.
Transformers are usually very heavy. One big thing that affects power loss in AC
power is the skin effect. The skin effect describes how alternating current has
higher resistances in the conductor or wire than direct current. Alternating
currents have a greater advantage of transferring power long distances.
DC voltages can be stepped up and down using small circuits for smaller circuit
configurations. Although they are not ideal for long distance power transfers they
work fine for a small circuit like a coil gun. DC step up circuits use active and
passive circuit elements to function. They are easy to use and very easy to
design. The only concern is making sure the circuit elements can deal with the
83
high flow of current. Power loss in the circuit can be an issue because of the way
the voltage is being regulated. There can be multiple step ups or step downs to
reach a specified voltage. The voltages in the coil gun will not be messed with
too much. The desired current and voltage will for the most part remain constant
before traveling into the capacitor banks. The addition of boost regulators for the
DC voltage step ups might be put into consideration for current and voltage
upgrades of the coil gun’s circuitry.
The transformer that was chosen for the coil gun was the HT97817 step up
transformer. The transformer was purchased at an electronics surplus store for
relatively cheap. The price for the transformer at the surplus store was $8.00. It
is used but in good working condition. This transformer steps up the 120V AC
voltage to 480V AC voltage. The primary of the transformer accepts 120 V
AC. The secondary of the transformer outputs 480V AC voltage. The capacitor
bank only needs 400V DC because of its voltage rating. Charging the capacitor
bank past the voltage rating can permanently damage the electrolytic
capacitors. As a precaution a voltage divider will be used to make sure the 400V
DC is the maximum output. Despite the weight of the transformer it was the
easiest solution for stepping up the AC voltage. It is a lot easier and cheaper to
step up AC voltages than it is DC voltages. The transformer weighs a little over
10 pounds. It is very heavy but will do the job.
Regulators Comparison
Linear Regulator
In designing the coil gun a linear regulator was taken into consideration. While
designing the coil gun circuit there was a possibility that the circuit might need
some type of voltage regulation. The voltage in the circuit and across electrical
components needs to be below their threshold voltages or ratings. Different
parts of the circuit are designed to withstand certain voltages. If at any time in the
operation of the coil gun it exceeds those voltages problems arise. A linear
voltage regulator can keep a constant voltage through the circuit. Linear voltage
regulators are not efficient at all. They do the opposite of what switching
regulators do. Linear regulators use either passive element in the breakdown
region or active devices such as bipolar junction transistors and metal-oxide field
effect transistors. These elements act as regulators. They act as regulators by
functioning as a variable resistor would. For voltage regulation they are
continually being adjusted to keep a constant specified voltage. The main
problem with the linear regulator is the way it dissipates the extra voltage. The
extra differential voltage is loss through an unconventional method. It loses it by
dissipating heat. Dissipating heat wastes large amount of energy. The output
84
voltage is the voltage that is used and controls the voltage throughout the circuit.
The linear voltage regulator uses feedback controls to adjust the input voltage to
keep the output voltage the same. The output is connected directly to the input
voltage.
The linear voltage regulator will not be used because it would hinder the ability of
maximum power transfer in the coil gun circuit. The circuit is already designed to
handle large amounts of currents. This implies that it can handle high voltage as
well. There is no step up transformers or DC to DC step up convertor so the
necessity of a voltage regulator is not an option. A linear voltage regulator might
be useful if the specifications in the coil gun design were changed. If for some
reason the speed of the coil gun shooting the projectile needed to be reduced
there are two methods. One would be by only partially charging the capacitors.
Another method would be by using a linear voltage regulator after the capacitor
bank. This would cause there to be a lower voltage across the capacitor bank
and their discharge would be significantly lower than normal. The significantly
lower voltage than normal would release smaller amounts of current through the
solenoid. The lower release of energy would cause the magnetic field of the coil
to slow the projectile down. Slowing down the projectile is not recommended for
the current design. If the current through the solenoid were reduced significantly
the projectile may not fire through the barrel at all. The suck back of the
projectile could occur. This would occur because the timing of the circuit would
be off due to a variation in the current pulse through the coil. The timing of the
circuit would need to be adjusted to properly shut the current off when the
projectile travels half the distance of the coil. The linear voltage regulator cannot
be used in the circuit. It presents too many problems and is not worth the
trouble. The drastic change in current would likely present a problem with the
holding current rating in the silicon-controlled rectifier as well. A switching
regulator would be a more ideal regulator to use if voltage regulation was needed
for the coil gun circuit.
Boost Regulators
Boost regulators were considered to help the coil gun deal with power transfer. A
boost regulator is similar to a linear voltage regulator. Boost regulators have
several advantages over the linear regulators. The switching efficiency of boost
regulators is much higher. The major advantage of using one in a coil gun is that
one can step up DC output voltages. It would be used directly after the AC to DC
full wave bridge rectifier to step up the output voltages. This would decrease the
charge time of the capacitor bank. The benefits of using boost regulators seem
too good to be true. The reason the boost regulator will not be used in the gun
comes from two factors. Boost regulators can be noisy and annoying just like the
85
ringing of capacitor banks after being discharged. Another drawback to using a
boost regulator is that it requires energy management to function properly. The
energy management issue can be addressed just by adding a control loop. There
is no real solution for the noise problem. The constant noise after trial and error
of constantly testing the gun would be irritating. In the future if one were to be
added it might be possible to sound proof the container containing the circuit. In a
coil gun soundproofing the circuit box seems to be a little extreme. With a
soundproof box the heat dissipation becomes a factor once again. The box would
likely get too hot and hinder airflow throughout the box. The boost regulator is
shown in Figure 57. The figure depicts the circuit elements that are need to
construct the boost regulator for the coil gun circuit. It is just a basic concept and
will need adjustments to possibly be used in the coil gun circuit.
Figure 58 Boost Regulator
Buck Regulators
A buck regulator is similar to a boost regulator. It contains all the components
that a boost regulator has except for its configured in the different manner. A
buck regulator consists of two switching devices; the switching devices are
usually a diode and a transistor. These switching devices control the inductor that
is connected to the load. When the switch in the circuit is on or connected the
voltage from the source stores energy in the inductor. When the switch in the
circuit is off the energy is discharged to the load. The function of the buck
regulator is the complete opposite from that of a boost regulator. Instead of
boosting DC voltage or stepping it up, it steps it down. A buck regulators main
purpose is to step down direct current. In the coil gun if for some reason the input
direct current were extremely high and the circuit could not handle it, it would be
necessary to use a buck regulator. If the input direct current were too high it
could damage certain components. Since every circuit component has a
maximum current rating exceeding these currents can cause serious problems.
The obvious problem is the overheating and malfunctioning of components. This
could lead to components igniting on fire. A buck regulator will be used if the
86
multimeter attached to the end of the full-wave bridge rectifier shows that the
current going into the capacitor bank is too high for the capacitors or the wire
chosen to handle. A high current does charge the capacitor bank faster which is
usually a plus, but too much current can have just as many disadvantages as
well. The buck regulator is shown in Figure 59. The figure depicts the circuit
elements that are need to construct the buck regulator for the coil gun circuit. It
is just a basic concept and will need adjustments to possibly be used in the coil
gun circuit.
Figure 59 Buck Regulator
Capacitor Charging Source
Capacitors store and release large amounts of energy. For a capacitor to
become charged it must be able to create an electric field across the insulator. In
order for an electric field to become present there has to be a potential difference
between the two conductors that lie between the dielectric or insulator. The
dielectric is non-conductive and does not permit charges to pass through but
stores the energy produced by the electric field. Therefore there must be work
done to move charges from one conductor to another. To charge a capacitor or
capacitor bank it is necessary to use a direct current source. An alternating
current source is not an option because at high frequencies current will flow
directly through the capacitor without any type of lag. At high frequencies a
capacitor has very little reactance. Since reactance is the opposition to
alternating current, a small reactance would exhibit little opposition to current
flow. AC circuits with capacitors are normally used for filters.
A charging circuit must have three elements of which include: the power source,
a resistor, and the capacitor. It is possible to connect multiple capacitors in
parallel to increase the capacitance of the circuit. The capacitance of capacitors
87
in parallel can be found by adding the capacitances up of all the capacitors
connected in parallel. The voltage across a capacitor is initially zero. When a
capacitor is connected in series with a power source and resistor current begins
to flow gradually charging a capacitor. The longer current passes through the
circuit the more charges are stored in the capacitor creating a higher voltage.
The voltage across the resistor and the voltage across the capacitor are inversely
proportional. It is for this reason, that there is less current going through the
circuit as time progresses. This is pertinent because it helps us to understand
why the rate of charging of a capacitor becomes slower with time. The time the
capacitor in a charging circuit takes to charge is related to its time constant. The
equation is below. In this equation R is the resistors resistance and C is the
capacitance of the capacitor. T is indicative of the time it takes the charging
current to fall 1/e of its initial current value. After 5RC the capacitor is nearly fully
charged.
In this equation R is the resistors resistance and C is the capacitance of the
capacitor. T is indicative of the time it takes the charging current to fall 1/e of its
initial current value. After 5RC the capacitor is nearly fully charged.
𝑇 = 𝑅𝐶
A light bulb is an ideal charge resistor for a coil gun. Light bulbs have great
characteristics for discharging capacitor banks, especially in our case where
safety is a major concern. They have low resistances at low power and high
resistances at high power, thus enabling a constant current sink. The problem
with using a standard resistor is overheating. Since a coil gun uses large
amounts of current and energy the resistor easily reaches its breakdown point.
This point occurs when the average power dissipation can no longer be safely
dissipated by the resistor. At this point the excessive power dissipation raises
the temperature of the resistor at which it can no longer function causing it to
burn out. When burn out occurs there is a possibility of a fire starting. Light
bulbs are perfect for power dissipation, where high heat is a concern.
The charge resistor for the coil gun is a light bulb. The type of light bulb chosen is
a 125W flood light bulb. In Table 12 shown below is the charge time it takes to
charge the capacitor bank at various voltages the coil gun will be operating at.
As the operation voltage increases so does the charge time and resistance.
Voltage
300
350
400
Capacitance
0.004
0.004
0.004
Resistance
720
980
1280
Charge Time
2.88
3.92
5.12
Table 12 Capacitor Bank Charge Time
88
Standard 75 W Light Bulb vs. 125 W Flood Light
Bulb
A light bulb was chosen to be used as a charge resistor. It was simple to choose
one as the charge resistor because of its great characteristics. Another problem
that arose was which bulb to choose. There are many different type of light bulbs
with different power ratings. The choices were narrowed down to two bulbs
shown in table 12 below. One bulb is a standard household bulb that is 75 W and
found in almost every lamp in the house. The second was a 125W flood light bulb
that is used primarily for flood lights. Below are the standard power equations
that were used to choose the right bulb for the coil gun.
𝑉2
𝑃=
𝑃 = 𝑉𝐼𝑃 = 𝐼 2 𝑅
𝑅
Voltage:
Power
75
125
300.00
Current
0.25
0.42
V
Resistance
1200
720
Voltage:
Power
75
125
350.00
Current
0.21
0.36
V
Resistance
1633.33
980
Voltage:
Power
75
125
400.00
Current
0.19
0.31
V
Resistance
2133.33
1280
Table 13 Standard Bulb vs. Flood Bulb
The table 13 above depicts each individual light bulb at the specified voltages.
The table gives a better understanding of which light bulb is best for the charge
resistor. The flood bulb allows more current to travel through the circuit and to the
capacitors. An important concept to note is that as the voltage increases the
current gradually decreases. The standard light bulb decreases at a slower rate
when voltage is increased. In a coil gun with a significantly higher voltage then
400V the standard light bulb may have a better advantage. A major item in the
table is the resistance. The standard bulb has an extremely high resistance than
the flood bulb. This is pertinent because it means the bulb with the higher
resistance takes longer to charge. The longer charging is relative to the equation
89
for the RC time constant. It is clear which bulb will be used in the coil gun. The
125 Watt flood light bulb is perfect for this 400V coil gun.
Digital Voltmeter
In most coil gun projects the designers choose to connect a voltmeter or
multimeter to measure the voltage the gun is operating at. In this coil gun it was
decided that it would be better to have a more permanent solution to measure
voltage across the capacitor banks while they are being charged. A digital
voltmeter will be included in the gun’s circuit to easily display voltage. An
additional feature may be to have a current meter added as well. For the coil gun
it is a necessity to monitor the voltage being used at all times to make sure the
maximum voltage is not exceeded. The voltmeter ensures that the voltage across
the capacitors does not exceed the maximum voltage rating of 400V. As a
precaution the banks voltage will be well below the maximum voltage rating.
Keeping a significantly lower voltage than the maximum voltage rating maximizes
the life and quality of the capacitors. The digital voltmeter will be placed in
parallel with the capacitor bank. The amp meter if added will be positioned in
series with the parallel capacitor bank and the solenoid.
Digital Thermometer
In the coil gun it is a good idea to keep track of the amount of heat being
generated. Similar to other electronics, a certain temperature should not be
exceeded or electrical failure will occur. Keeping track of the temperature of the
coil gun is not a hard issue to address. A digital thermometer can be used to
keep a record of how hot the gun is getting. In testing the coil gun the digital
thermometer can be a quick reference in seeing if the gun is functioning
correctly. If for some reason temperatures keep raising at fast rates it can be
implied that there is a problem in the circuit. This means the circuit is on the
verge of failing and the current through the gun should be immediately turned off.
When the temperature that the coil gun cannot operate is reached it will be
recorded. This temperature value will be important to know because then the coil
gun can be tuned so it will never reach that temperature again. From that
temperature value the circuit can be upgraded so that it will turn off before it
reaches that temperature. This safety feature will save the coil guns components
from ever being shorted or fried. The temperate value at which the projectile
reaches its maximum speed will also be an important value to keep in mind. The
circuit might be able to be tuned to only shoot at that temperature value as well.
This is perfect and helps to identify any problems when the voltage being
inputted into the circuit is being varied.
90
The digital thermometer in conjunction with the fans and heat sink will be the key
to dealing with power dissipation. The digital thermometer helps to pick the
appropriate fan speeds. The fan speeds can be orientated on the output being
displayed on the digital thermometer when the gun is in use. If the digital
thermometer in the system is chosen to be used it will like come already
preassembled. The circuit construction of one is fairly simple but making sure the
value being displayed needs is the main priority. If for some reason the value
being displayed was wrong because the digital thermometer circuit was not
constructed properly it would be detrimental in the design of the heat dissipation
process. The preassembled digital thermometer will be tested before it is actually
used in the guns circuit. To make sure it is function properly it ambient
temperatures will be taken from inside and outside. Once temperatures are
verified only then will the digital thermometer be used. The display only uses
small amounts of current to run. It will likely use the same power that is being
used by the computer fans. The digital thermometer is a pertinent addition to the
circuit. It may even be just as important as the voltmeter and current meter. It
allows the user to easily analyze circuits and the problems that are arising. If the
thermometer is added to the circuit it will provide valuable information to the user.
It will also be able to be used for troubleshooting other circuit components.
Testing Current
The current throughout the circuit should be tested. It is important to know the
current through each electrical component to make sure that device can handle
the current at that point. The peak current along with the time the peak stays
constant should be identified too. Identifying this gives the designer the ability to
modify the circuit to handle the peak currents. Knowing peak currents helps
because then the proper electrical components can be switched out for higher or
lower values depending on what the multimeter displays. Each component
chosen was picked for either its current rating or voltage rating.
The
manufacturer spread sheet usually only designed for normal use. It doesn’t give
specifications for high currents at short periods in time for such devices as a coil
gun. Therefore a lot of the devices current and voltages rating at short durations
were computed to choose the appropriate device. It is still possible that some of
the devices may fail when in use. It is very difficult to estimate the peak currents
and voltages with respect to time. To measure the current through the circuit a
portable digital multimeter will be used. A multimeter will display the currents
through that portion of the circuit if it is orientated in series with the device’s
current it is trying to read. This will have to be done before the circuit is
permanently connected. Once the circuit is permanent is will be difficult, if not
impossible to measure the current at every single piece of the circuit. The testing
91
of the circuit should be at the maximum voltage that the circuit is being designed
to operate. That voltage is of course four hundred volts. This voltage is only after
the capacitor bank is discharged. The portion of the circuit before the capacitor
bank will stay the same even when the voltage is varied at the capacitor bank.
The solenoid will have the highest peak current so that is the device that will get
the most attention when the current is being measured.
The current will first be measured from the beginning of the circuit which is the
power source. Any outlet that the gun is to be used with will need to be tested. In
order to test the outlet the two probes of the multimeter will be connected to each
terminal of the outlet. This will measure the current that will be traveling from the
outlet to the full-wave bridge rectifier that is converting alternating current into
direct current. The current going into the bridge rectifier should not be more than
40 amps because the diodes can only handle a current of 20 amps each. Since
the full-wave bridge rectifier has a split at the initial portion of the circuit it can
handle twice its current rating, which is where the number 40 amps come from.
The current should be measured from the top portion of the full-wave bridge
rectifier and the lower portion as well. The wire resistance and the diode internal
resistance should be extremely minor. The current through the charge resistor
should be the same as the two values of the full-wave bridge rectifier combined.
This is because the charge resistor is just before the full-wave bridge rectifier and
is in series with the AC power source as well. When the capacitor bank switch is
turned to the on position and the capacitor bank is charged to the max the
current through the rest of the circuit should be zero. The current stays zero until
the silicon-controlled rectifier receives a current through its gate to release the
holding current. The current released from the capacitor can then be measured
after its discharge. The current released from the capacitor bank is extremely
large. This current travels to the solenoid. At this point it is important to measure
the time at which the current stays at its peak value. The current for the whole
discharged of the capacitors should be measure through the solenoid. From this
a relationship of the current with respect to the time formulation can be derived.
The current through the silicon-controlled rectifier should not be an issue. The
device is designed to handle high currents. The only thing that should be looked
at is the voltage across the electrical component. The current traveling through
the gate may be measured to make sure enough current it being supplied to
operate the device properly. There is a damping resistor connected to the
cathode of the silicon-controlled rectifier. The current through the resistor should
be the same as the current through the solenoid and silicon-controlled rectifier
since all those components are in series.
Testing Voltage
It is necessary to measure the voltage throughout the circuit. The voltage will be
measured using either a voltmeter or multimeter. It is fairly easy to measure the
92
voltage in any circuit. The voltage in the coil gun circuit can be measured by
simply putting the voltmeter in parallel with any electrical component. This has a
big advantage over measuring the current. The voltage can be tested even if the
circuit is in its final design and permanent.
The voltage of the electrical AC outlet will first be measured. The output voltage
should range from 110V to 120V. From that point current flows through the
charge resistor to the full-wave bridge rectifier. The voltage at the charge resistor
should be measured. This is important because it can give the designer the
ability to calculate the amount of heat that is being dissipated. Since the charge
resistor is a light bulb there is not too much concern over the amount of heat
being generated. Light bulbs are designed to deal with high power. If the charge
resistor was a normal resistor there would likely be an issue with the high
temperatures being generated. The voltages at the diodes of the full-wave
bridge rectifier are measured as well. Diodes can handle high voltages because
they have little internal resistance. The voltages at the capacitor bank are the
most important voltages to measure. Even though there will be a digital
voltmeter to measure the voltage that the capacitor banks are being charged to a
voltmeter will be connecting there too. The capacitors are the most expensive
part of the coil gun. If any one of those capacitors reaches above 400V or close
it can permanently damage a capacitor. Replacing the capacitor would take
some time. Having extra capacitors would be beneficial but the cost is too high.
A lot of precaution will be taken to make sure the capacitors do not fail or
malfunction. After measuring the voltage across the capacitor bank, the voltage
of the solenoid will be measured. It should be 300 volts to 400 volts as well but it
will still be measured to be sure. The current might have a large effect on the
amount of voltage being produced. The voltage over the silicon-controlled
rectifier will also be interesting to observe. There is no worry about the voltage
because the one chose has a very high voltage rating. The voltage across the
damping resistor should be real low. The value of the resistor is small so that is
where that assumption comes from.
Knowing the voltage may not be as important as knowing the current but none
the less is pertinent. Voltage differences and changes in the circuit can also lead
to better analysis of problems that may be present in the circuit. Every device
needs a voltage present to operate. If no voltage is present in a device it can lead
to the assumption that there is not enough power going through the circuit for it to
properly operate. Too much voltage is the circuit can cause problems that deal
with overload. There being excessive voltage can short electrical components
and may even produce too much energy from power dissipation. Voltages can
also affect the way the magnetic field is produced in the solenoid itself. The
voltage when the coil gun is turned off should be measured as well. The peak
voltages that might occur can damage the capacitor bank, silicon-controlled
93
rectifier, or coil. The peak voltages can create an undesired back electromagnetic
field.
Circuitry Protection
A high powered coil gun requires protection in the circuit. There needs to be a
way the circuit deals with negative effects that will occur during operation. Some
negative effects that can occur in a circuit are back electromagnetic fields,
overheating, shorting, ringing, and power loss. Electrical components can also be
negatively affected by these symptoms. Special precautions must be taken to
rectify these symptoms or even keeping them from occurring all together.
Keeping these negative effects from occurring can increase the safety of the coil
gun and user operating it. Every circuit design needs some type of protection
against things that could negatively occur.
The capacitor bank has a number of different added electrical components. To
keep the capacitors from warming up too much each will have cap coolers. The
cap coolers will keep the capacitor bank at lower temperatures. The capacitor
banks are the most costly part of the coil gun so every precaution will be taken to
keep them from any type of damage. The temperatures of the capacitors are not
expected to reach high temperatures even without the cap coolers. The cap
coolers will simply add more protection and help the capacitor bank to operate
normal through the entire discharge process. The negative effects of the
magnetic fields can harm the capacitor bank when the device is off. After a
discharge there can be back electromagnetic field that can damage the capacitor
bank, silicon-controlled rectifier, and solenoid. The back electromagnetic field
will be reducing, if not eliminated by adding a diode. The diode better known as a
flywheel diode can help silicon and other electrical devices from being damaged
by the field that is generated.
The solenoid will undergo the most abuse in the circuit. The coil will experience
large amounts of current through its wire. The wire must be able to handle
extensive amounts of heat being generated from high currents for short duration
times. The heat that is generated from the current through the coil will be
minimized by using heat dissipating electronics.
The heat dissipating
components will be by coating the wire with insulation. The surface area of the
wire is cooled by using fans and heat sinks. The solenoid experiences a majority
of the electromagnetic field that is being generated to push through the projectile.
The solenoid was also given a flywheel diode. This will protect it from the peak
voltages that occur as a result of the back electromagnetic fields that are exerted
on the wire. The wire will experience only small duration of current to protect it
from overheating which it can easily do. The manufacturer specification sheets
94
for the wire only give specifications for maximum current for normal use, not for
short durations. The short pulses help to alleviate the amount of heat that can be
generated if the coil were to have high current flow through for an extended
amount of time.
The silicon-controlled rectifier requires little protection from negative effects than
can occur in the circuit. Most silicon-controlled rectifiers already come ready to
function with circuits that have high currents and voltages. There is one thing
that the silicon controlled-rectifier does react negatively with. That is the back
electromagnetic field. A flywheel diode could be inserted in parallel with the
silicon-controlled rectifier but that could cause another problem. To protect the
silicon-controlled rectifier a resistor is connected to the cathode in series. The
resistor in series makes the circuit critically damping. The estimated cost to build
the Power Subsystem is $20 for various passive components.
Controls and SoftwareSubsystem
The purpose of the Controls and Software Subsystem is to design the software to
drive the coil gun, to determine how the coil gun is going to communicate with the
software, and to choose devices that achieve these feats. Throughout this
subsystem many ideas and options will be discussed on how to mesh these
ideas to allow the coil gun to move and communicate with all the other
subsystems correctly and efficiently. Some of the systems within the Controls
and Software Subsystem that will be discussed are the position determination,
which discuss different types of hardware and how it will communicate through
other devices, the FPGA, the Microcontroller, the Software Architecture, and
finally the User Interface, which will discuss how the user will interact with the coil
gun’s software as a whole. The following diagram, Figure 60, is a function block
diagram for the Controls and Software Subsystem and how it acts with the other
subsystems in this coil gun project as a whole:
95
VIDEO CAMERA
(FROM SEPARATE
SUBSYSTEM)
F
P
G
A
USER
INTERFACE
SOFTWARE
ARCHICTECTURE
MICROCONTROLLER
POSITION
TRANSDUCER
FIELD
GENERATION
SUBSYSTEM
DC MOTORS(FROM
SEPARATE SUBSYSTEM)
MOTION SUBSYSTEM
Figure 60 Controls and Software subsystem
Position Determination
There are two viable options that will best serve the need to determine the
position of the targeting of the coil gun. The first is an analog system called a
resolver and the other is a digital component known as rotary encoder. Since the
choice has been made to use an FPGA, the rotary encoder has been chosen
over the resolver because it is digital versus analog and there would be no need
to purchase an analog to digital converter.
A rotary encoder is an electro-mechanical device which takes an angular position
of a plane and converts it to digital code. Thus this would make a rotary encoder
also an angle or position transducer. Rotary encoders are used frequently on
servo motors of mechanical devices to track the position of the motor shaft.
96
There are two types of rotary encoders. First there is an absolute rotary encoder
and then there is the incremental rotary encoder. Determining the difference
between these two types of rotary encoders is essential before we are able to
make a decision on a model of any sorts.
The absolute rotary encoder creates a digital code for each specific angle of the
plane. Absolute rotary encoders also do not lose their position after the device is
powered down, they can give absolute position once powered back up. There
are four main types of absolute rotary encoders we have to choose from. The
four different absolute rotary encoders are known as the mechanical absolute
rotary encoder, the optical absolute rotary encoder, the fiber optic absolute rotary
encoder, and finally the magnetic absolute rotary encoder.
The fiber optic absolute rotary encoder is a non-metallic position sensor that
works in extreme electromagnetic fields. Fiber optic absolute rotary encoders
are ideal for mechanic and industrial inspection systems, such as an MRI due to
the ability to operate in electromagnetic fields with ease. Though we are using an
electromagnetic field to fire a projectile, it may not be extreme enough for us to
be concerned about the effects on the rotary encoder. Fiber optic absolute rotary
encoders also seem to be far out of our price range.
The magnetic absolute rotary encoder is an accurate angular measurement
device for over a full turn of 360 degrees. A two pole magnet rotates over the
center of the chip to measure the angle. By this process the absolute angular
position is given instantaneously. There is also no need for calibration in most
cases for a magnetic rotary encoder. The magnetic absolute rotary encoder is
highly reliable and ideal for applications in harsh environments. This would be
considered closer to our price range for such an item, but I feel we could create
problems since we will be creating an electromagnetic field with our coil which
could affect the encoder.
The mechanical absolute rotary encoder uses an insulated disk complete with
concentric rings of opening, all being fixed to the plane. As the disc rotates, there
are contacts, which are connected to separate electrical sensor. The metal of
the disk is connected to a source of electrical current. A pattern is created, due
the contacts coming in contact with the metal disk, in a binary sequence of every
possible position. The mechanical absolute rotary encoder seems to be an
efficient device and some model can be found that will meet our budget.
Optical absolute rotary encoder has a disc like its mechanical counterpart, but
instead of being made from metal it is made of either glass or plastic. To
determine the disc’s position at any time, a light source and detector creates and
reads an optical pattern. To read the code you will some sort of controlling device
to determine the angle of the shaft; we will be using a microcontroller.
97
Also known as a quadrature encoder, the incremental rotary encoder is another
option we have to determine position of our plane. An incremental encoding will
essentially generates pulses proportional to the position of the plane. The pulses
are represented as a binary sequence, i.e. 5V would be a 1 and 0V would be a 0.
The pulses are then counted, the counter will add or subtract the signal while
turning. The incremental rotary encoder is most common rotary encoder due to
low cost. For an incremental rotary encoder there are two outputs, known as
quadrature outputs. The fact that they are only two outputs is reasoning for the
incremental rotary encoder’s low cost. Though the incremental rotary encoders
only use two sensors the accuracy the device is not compromised. Like the
absolute rotary encoder, the incremental rotary encoder can be either
mechanical or optical.
Mechanical incremental rotary encoders do have problems that may affect our
system. What has been found is that because mechanical incremental rotary
encoders’ switches require debouncing they are limited to the rotational speeds
they can handle.
Optical Incremental rotary encoders are chosen over there counterpart, the
mechanical incremental encoder, when there is a need for a high degree of
precision or for a tolerance to higher revolutions per minute is a desired
commodity.
Choosing the ideal rotary encoder for this project is a difficult task because of the
similarity is effectiveness and cost of some of the models. The team will most
likely evaluate the choice of the specific rotary encoder later on into the project
design.
Software Architecture
For this coil gun project, the software architecture involves all of the functions
and structures used to control the movement and targeting. The functions
involved in the software architecture are the basic C functions and the other
functions are function that will be created while designing the programs.
Main():
The first function that will be discussed is the ‘main’ function call. The ‘main’
function is one of the standard C programming functions. The execution of the
program controlling all of the functions of the controlling of the coil gun will all go
through the ‘main’ function. The program will stay within the ‘main’ function while
98
being executed. At the end of the function instead of terminating it will return the
beginning. To exit out of the ‘main’ the user will have to manually do so by using
the proper protocol. While in the ‘main’ function the program will go through the
other functions within the program. The program will also end after the program
runs through the function which fires the coil gun. So either to ultimately end the
program the user will either do so manually or the program will automatically do
so after the program fully executes all its purposes. Inside the ‘main’ all of the
local variables we need to be declared at the beginning of the function.
Menu():
The ‘menu’ function will be the first time the user interacts with the program using
the ‘main’ function. The ‘menu’ function will prompt the user with a menu of
choices. The choices that the ‘menu’ function will prompt the user with are first
the manual targeting option or second the automatic targeting option. Looking at
the software architecture block diagram shown below depending on the option
selected the path through the ‘main’ function will be different. Once either option
is chosen the program will then proceed to the ‘position’ function.
Position():
The ‘position’ function is the next step through the ‘main’ program. The ‘position’
function will be used to determine the position of which the barrel of the coil gun
is directed at. To achieve this feat the function will communicate with the rotary
encoders on the horizontal plane and the vertical plane. The encoders will tell
the program at which angle each of the planes are at respectively with the
horizontal axis and vertical axis. Once the position is determine through using
the rotary encoders the function will save the information about the position into
the functions variables. The ‘position’ function will be needed to be called later
on in some of the other functions to access the information in the variables.
Sensor():
The next function to be discussed is the ‘sensor’ function. The ‘sensor’ function’s
main purpose is to grab the data from the camera/sensor. The ‘sensor’ function
will be accessed and used when the user has selected to use the automatic
targeting choice at the ‘menu’ function earlier on in the program. The
sensor/camera will located at the end of the barrel of the coil gun itself. To save
power the camera/sensor will not be active at all times. Within the ‘sensor’
function the camera/sensor will be turned on and off when the data from the
camera/sensor is needed or when the camera/sensor is no longer needed for the
calculations.
UserInput():
99
The ‘user input’ function is a function is created to allow the user to decide where
he/she would like to direct the coil gun’s barrel with respect to the location at
which the coil gun’s barrel is currently at. The process for the ‘user input’ function
will allow the user to input coordinates at which he would like for the coil gun to
target or to use the directional keys to move the coil gun.
PosDiffCalc():
To determine how far to move the coil gun from left to right and up and down, the
function ‘position difference calculator’ is called. The ‘position difference
calculator’ will take the data from the ‘position’ function for the initial position and
the then take the data from either ‘sensor’ function or the ‘user input’ function
whether the user chose the manual targeting system or the automatic targeting
system. After both of the variables are taken in from the other two functions the
‘position difference calculator’ function will be able to make its calculations. The
‘position difference calculator’ function will then take the absolute value of the
subtraction of the variable from the ‘position’ function with either the variable of
the ‘sensor’ function or the ‘user input’ function, as stated above to know whether
or not the ‘sensor’ function variable or the ‘user input’ function variable is used is
the choice the user made between the manual targeting system or the automatic
targeting system.
DCMotor():
The ‘DC Motor’ function is the next function to be discussed. The ‘DC Motor’
function will be using the data from the variables in the ‘Position Difference
Calculator’ function. The variables from the ‘Position Difference Calculation’
function will be used to determine the voltage that will be needed to be applied
the DC servo motors and for how long that voltage will be needed to be applied
for. Doing these calculation will allow for the DC servo motors to move the
correction distance horizontally and vertically.
Charge():
The next function, the ‘Charge’ function, will begin the process of the actual firing
of the coil gun. The ‘Charge’ function will turn the circuit on that will begin
charging the capacitor banks. The ‘Charge’ will be called when the coil gun’s
barrel directed at the specific target chosen by the user. Once the charging
process has begun the ‘Charge’ function will begin receiving data from a
measuring device telling the function what the value of each capacitor is.
FireStandby():
100
The ‘Fire Standby’ function will call the ‘Charge’ function. The ‘Fire Standby’
function will loop around taking in the capacitor values from the ‘Charge’ function.
When the capacitor values are at the desired levels for firing the program will
then proceed to the ‘Fire’ function. If the capacitor values are not at the desired
levels then the ‘Fire Standby’ will loop back through acting as a standby.
Fire():
The final function that will be discussed is the ‘Fire’ function. The ‘Fire’ function
will call to the ‘Fire Standby’ function. The ‘Fire’ does this call to make sure that
the capacitor values are when we want them to be before unloading them. If the
go ahead is given by the ‘Fire Standby’ function then the capacitor banks will be
unloaded into the coil. Once the ‘Fire’ function unloads the capacitor bank into
the coil the projectile will be fired. After the projectile is fired then the ‘Fire’
function sends the user back to the beginning of the ‘Main’ function and start over
with the ‘menu’ function, allowing for the user to decided if he wants to continue
to choose to the automatic targeting system or the manual targeting system or
exit out of the program.
A Block Diagram of the Software Architecture is shown in Figure 61 below:
101
MenuSelect()
auto
Position()
MAIN
manual
Position()
Sensor()
Userinput()
PosDiffCalc()
PosDiffCalc()
DCMotor()
Charge()
FireStandby()
Capacitors
charged?
Fire()
Figure 61 Software Architecture Block Diagram
Microcontroller
102
A microcontroller is essentially a computer on an integrated circuit. Just like any
other computer, a microcontroller has memory, a processor and is a
programmable device. Microcontrollers are specially designed for embedded
systems, rather than just general systems. Some systems, or products, that use
microcontrollers are remote controls, power tools, home appliances, etc. Real
time response is a necessity for an embedded and microcontrollers in turn
provide real time response, which is why microcontrollers are a perfect fit of
embedded systems. To allow for real time response, microcontrollers use
interrupt routines. When an interrupt routine is processed the microcontroller
completes that specific task before it continue can on with its next instructions.
Since microcontrollers understand machine code, compilers and assemblers are
used to turn high level programming languages, such as C and Java, into just
that.
Some other features of the microcontroller are the GPIO pins.
Each
microcontroller has several of these GPIO, and the purpose of these pins is to be
programmed to be an input or an output state by the developer. Pins
programmed to be an input pin purpose would be to read in outside signals and
sensors. Pins programmed to be an output pin purpose would be to drive motors
and other devices. Sometimes there is a need for a microcontroller to take in or
output an analog signal, to do these processes an analog to digital converter or
digital to analog converter would be needed. Timers are another tool that can be
programmed into a microcontroller. One example is a PIT, programmable
interval timer. A PIT has counter, once that counter reaches zero an interrupt is
sent to the processor and indicating that the counter has finished. There are
other timers such as a TPU, time processing unit, which would be used detect
input events, generate output events, and so on. Another process that would be
very beneficial to this project in general is the PWM, pulse width modulation. The
PWM allows the CPU to control power converters, motors, etc. The last block
researched would be used to transmit and receive data allowing communication
with other devices in the system; this feature is known as the Universal
Asynchronous Receiver/Transmitter, or the UART.
As discussed above, shown below Figure 62 is a description of the pins of a
microcontroller. This microcontroller is the PIC16F84A of the PIC microcontroller
family.
103
Figure 62 Courtesy of Microchip, permission given
The next figure62 is the actual location of the pins on a microcontroller, or the pin
diagram. Again this is the PIC16F84A 8 bit microcontroller of the PIC
microcontroller family.
104
Figure 63 Courtesy of Microchip, permission given
Programming the Microcontroller
To program our chosen microcontroller C will be the chosen programming
language. The program as a whole will consist of multiple classes for each of the
microcontroller’s functions. The first class will be used to tell the microcontroller
how to control each of the DC motors. This will control the voltage sent to each of
the motors to controls the arms of the targeting system. The voltage applied lets
the motors when and how far the move the arms. Another class will be used to
tell the microcontroller to receive the information from each of the rotary
encoders. The information from the vertical encoder will be used to determine the
angle the coil gun’s barrel is pointing with respect to the horizontal plane. The
information from the horizontal encoder will be used to determine the angle the
coil gun’s barrel is pointing with respect to the vertical plane.
Choosing a Microcontroller
105
To choose the right microcontroller research on many devices will need to be
done. There are a few concerns as far as the hardware goes. For one, will the
microcontroller chosen be able to handle the amount of devices going through it.
Can the chosen microcontroller output instructions to the DC servo motors while
possibly also receiving input from a video camera, which would be showing what
the coil gun is targeting. There are multiple families of microcontrollers that will
be decided between.
Atmel AVR
The first family of microcontrollers that will be discussed is the Atmel AVR. The
Atmel AVR was first developed in 1996. This microcontroller family is a modified
Harvard architecture 8-bit single chip microcontroller. The Atmel AVR is also
known for being one of the first microcontroller families to have a program
storage system that used flash memory. To rid the need for external memory in
most cases the Atmel AVR has flash, EEPROM and SRAM all integrated onto a
single chip. The Atmel AVR’s program instructions are stored within the flash
memory. Each of the program instructions usually take one or two 16 bit words.
The size of the program memory for each of the models within the Atmel AVR
family is described in its name. So if the model’s name contains 64 in it the model
then has 64 kb of flash, and if the model’s name contains 32 in it the model then
has 32 kb of flash. Within the internal data memory lies the register file, I/O
registers and the SRAM. These can be found in the data address space. The
Atmel AVR family of microcontrollers have 32 single-byte registers, and are
within the 8-bit RISC (Reduced instruction set computing) devices classification.
The working registers are the first 32 memory addresses and then are followed
by the 64 I/O registers. In hex the first 32 would be 0000 to 001F and the next 64
would be 0020 to 005F. The SRAM would follow then follow he registers at 0060.
The addressing and opcodes for the register file and I/O registers can be
addressed as if they were in the SRAM.
Internal EEPROM is common throughout all of the Atmel AVR family. EEPROM
is usual because like flash memory when electrical power is lost the EEPROM
can retain all its content. The EEPROM can only be accessed in the same
process at which an external device would be accessed. This would be done by
using point registers and read/write instructions. The reasoning for this is
because the EEPROM memory is not mapped in to the addressable memory
locations within the microcontroller. Having to use these methods to access the
EEPROM, causes the access of the EEPROM to take more time than the other
internal RAM devices. Writing to the EEPROM is a limited process. According to
Atmel the EEPROM in this family of microcontrollers the number of writes an
EEPROM can handle is one hundred thousand.
106
The Atmel AVR family of microcontrollers uses a two stage, single level pipeline
design. This design allows for the device to execute an instruction while also
taking in the next instruction. The Atmel AVRs are fast when considering 8 bit
microcontrollers. They are fast because for the most each instruction takes only
one to two clock cycles. Rather than most 8 bit microcontrollers, the instruction
set of the AVR family is more of an orthogonal style. Describing this orthogonal
style can be done by going through a few examples. For instance the pointer
registers X, Y, and Z all have different addressing capabilities from one another.
Another example of this orthogonal style is that the register locations R0 – R15
addressing capabilities are different than the addressing capabilities of the
registers at the locations R16 – R 31. Additionally the I/O ports 0 to 31 differ in
addressing capabilities the I/O ports 32 to 63.
Below is a block diagram, Figure 64, of one of the products in Atmel AVR family:
Figure 64 Atmel AVR Family
107
Most commonly the AVR family can handle clock speeds up to 20 MHz, while
there are some models within the AVR that can handle upwards to 32 MHz from
some devices. The more recent models of the AVR family no longer have the
need for external clocks or resonator circuits, due to having an oscillator on the
chip itself. A few of the AVR models contain a system clock prescaler, this
prescaler can be configured during run time allowing optimization of the clock
speed.
The development tools for the AVR family can be inexpensive, in some cases
even free. This could be a decided factor towards these models if the other
families of microcontrollers have pricing development tools. The inexpensive
development tools would include the development board, while the free
development tools would be the development software. Some of these
development tools and evaluation kits for the Atmel AVR consist of numerous
starter kits. Some of these starter kits are the STK600, STK500, the AVR dragon,
the AVR Butterfly, etc.
On whole the Atmel AVR family of microcontrollers is a formable foe when
deciding on the microcontroller of choice for this project. The Atmel AVR family of
microcontrollers is efficient, fast, inexpensive, and can handle many processes
that the project would need.
PIC Microcontrollers
The PIC family of microcontrollers is a family of Harvard architecture
microcontrollers developed by the Microchip Technology company. This family of
microcontrollers was derived from a microcontroller known as the PIC1640,
which was developed by General Instruments. Initially the name PIC was meant
to refer to the Programmable Interface Controller. The PIC family’s low cost,
availability of the device and its development tools, and the ability to be reprogrammed with flash memory makes the family intriguing to the industry while
also being intriguing to the everyday hobbyist.
The architecture of the PIC family of microcontrollers mainly consists of the
following:
-
The Harvard architecture that is commonly used with microcontrollers
outside of the PIC family.
Fixed length instructions
Single cycle execution type of instructions, taking about 4 clock cycles.
RAM locations functioning as registers
Hardware stack for return addresses
108
-
Data space mapped with CPU, port, peripheral registers, as well as the
program counter
The RAM acts as the memory and registers, thus the memory space is not
distinguished from the register space. Because the RAM acts like this it is
commonly referred as just plainly the register file. All the microcontrollers in the
PIC family use data and addressing in 8 bit lengths, in turn are called 8 bit
microcontrollers. The PIC family of microcontrollers saves return addresses in a
hardware call stack. The hardware stack is software accessible, unlike earlier
microcontrollers. The PIC family can handle any from 30 instructions to 80
instructions depending on the model. The instruction set involves instructions to
execute many operations on the registers directly. Below is a block diagram,
Figure 65, of a microcontroller in the PIC family:
109
Figure 65 Microcontroller in the PIC family
Courtesy of Microchip (permission given)
TI MSP430
The final microcontroller family that will be discussed is Texas Instruments’ TI
MSP430 family of microcontrollers. The MSP430 family is built around a 16 bit
processing unit. Low power consuming embedded systems and low cost was the
main focus when developing the MSP430 family of microcontrollers. Current
110
drawn during the idle mode can be lower than 1 mA and 25 MHz is the top CPU
speed, this allows for the low power consumption. There are six different LowPower modes within the MSP430, as well. As seen below, the functional block
diagram for the TI MSP430 is somewhat similar to other functional block
diagrams of the other families of microcontrollers.
A functional block diagram, Figure 66, of microcontroller from the TI MSP430
Figure 66 Microcontroller from the TI MSP430
Courtesy of Texas Instruments
Due to cost of using both a FPGA and a microcontroller, and we must use a
FPGA to deal with video processing, the group has decided to do all processing
on a FPGA.
111
FPGA
A FPGA, field programmable gate array, is an integrated circuit that is configured
by the user using a HDL, or hardware description language. Using a FPGA
allows for the user to design a system based on logic functions. Basic logic
functions would be AND, OR, XOR, NOT, etc. These logic functions are created
using gates of logic and bits of memory. The HDL language that is most common
is known as Verilog VHDL. Verilog VHDL is a programming language similar to
the C language, making the development and debug processes similar as well.
Early single chip microprocessors had around ten thousand logic gates and ten
thousand bits of memory, compared to present day FPGAs which could possible
contain close to ten million logic gates and ten million bits of memory shown as
Figure 67.
Figure 67 FPGA
Another big difference in the FPGA versus its old counterpart, the
microprocessor, is the logic gates are manipulated by the programmer. Even
though this is the case, the development of the FPGA and a microprocessor are
similar enough that if someone knew how to develop one or the other they could
112
easily learn the other. A step by step comparison of the two design processes is
shown in Table 14 below:
Microprocessor
Architectural design
Choice of language
(C, JAVA)
Editing programs
Compiling programs
(.DLL, .OBJ)
Linking programs
(.EXE)
Loading programs to ROM
Debugging P programs
Documenting programs
Delivering programs
FPGA
Architectural design
Choice of language (Verilog, VHDL)
Editing programs
Compiling programs
Synthesizing programs
(.EDIF)
Placing and routing programs
(.VO, .SDF, .TTF)
Loading programs to FPGA
Debugging FPGA programs
Documenting programs
Delivering programs
Table 14 Microprocessor FPGA Comparison
FPGA Design Process
Editing, Compiling, and Synthesizing:
Because Verilog is very similar to C type programming, any editor can be used.
Though we can use any environment to code, I would prefer to use the
environment suggested by the manufacture of the FPGA we choose.
The compiling process for a FPGA is similar to any other compiling process. The
process involves building the files created in the editing process to create the
logic sequence desired. When the code is compiled the logic gate data are
written to the registers, latches, out ports, buses, etc. of the FPGA.
The synthesizing process, though seems similar at times, is different than the
compiling process. Synthesizing takes your program logic and maps it into logic
gates, rather than the processing instructions of the FPGA. The FPGA programs,
once compiled, must eventually be embedded into the actual FPGA. This is done
113
by using the IDE software to upload the program into the FPGA using an USB
cable from the PC.
Choosing an FPGA
There are a few manufactures of FPGAs and each of those manufactures have
many models that can be used. Some of the manufactures that will be chosen
from are Xilinx and AMTEL. The biggest concerns in choosing a FPGA to use for
this project are cost and sophistication. When I say sophistication, I mean how
many gates of logic and bits of memory are we going to need for the processes
we will executing.
One of the devices that will be looked at is the AMTEL FPSLIC. The device is a
combination of both a FPGA and a microcontroller. Since the AMTEL FPSLIC is
a FPGA and a microcontroller in one device, this allows for a hastier design and
developing time versus having to integrate a microcontroller to an FPGA that are
not molded together. The cost of the AMTEL FPSLIC is more than using a Xilinx
FPGA and the Xilinx IDE, but I believe using the AMTEL FPSLIC for its advanced
features could be the better choice. With the AMTEL FPSLIC we receive the
development board and the IDE that is used to program the device. Having to
learn the IDE that is used for the AMTEL FPSLIC could create a problem though
not a major one. The biggest problem is the time constraint of trying to learn to
use the development and debug process of a new environment, where as if the
Xilinx FPGA was to be used it would be an IDE that the group already has
experience with the software.
Another device that will be considered is a FPGA of the Spartan series from
Xilinx. One of the advantages of using a Xilinx FPGA is the fact that IDE is free to
use, thus we would only need to purchase the development board instead of a
starter kit. Since we would only need to buy the board from Xilinx we are
probably looking at a cheaper solution than if we were to buy the starter kit.
Within the Xilinx family the FPGA that would be chosen would be from the
Spartan 3 family of FPGA. The eight family Spartan 3 can contain anywhere from
50,000 gates to five million gates, and is meant to meet cost sensitive projects.
Below is a functional block diagram, figure 67, of a Spartan 3 FPGA
114
Figure 68 Spartan 3 FPGA Courtesy of Xilinx
So between the two FPGA systems discussed there advantages and
disadvantages. With Xilinx we would be using a system that we are familiar with,
though Xilinx system does not have a microcontroller embedded. With AMTEL
FPSLIC it has an IDE that would need to be learned to use in a short period of
time, but to its advantage it has a microcontroller embedded in its systems which
in turn could end up being very useful.
After further consideration due to the cost and familiarity of the device, a FPGA
from the Xilinx Spartan 3 family of FPGAs will be chosen for this project. It has
been found that a Spartan 3 FPGA and evaluation board can be found for the
price of 129.00 on some sites, such as a store site on Ebay.com.
User Interface
For the user interface there are multiple functions that we would like for it to
cover. We would like for the user to have the choice of a manual and automatic
setting. We would also like for the user interface to contain a password input. The
115
motion control of the coil will be programmed into the microcontroller/fpga
software and will be discussed in another section.
For the manual setting, the idea would be to have two sub settings. The first
setting would allow the user to input coordinates. When the coordinates have
been input the coil gun’s targeting system will aim the coil gun directly towards
the target. The way this would work is the vertical and horizontal coordinates
would go into separate variables. For the vertical coordinate input we would take
the difference of the input versus the current position. The code would be then
executed differently depending on whether the new coordinate is less than the
current or the new coordinate is greater than the current coordinate. The current
position would then be increased or decreased by an index in a loop until the
current position is equal to the inputted (new) coordinate.
Pseudo code:
Current position -> Xcurrent, Ycurrent
Inputted position -> Xnew, Ynew
Horizontal difference = Xcurrent – Xnew
Vertical difference = Ycurrent – Ynew
If Horizontal difference < 0, Increase Xcurrent by index
If Xcurrent not equal to Xnew stay in loop else continue
If Horizontal difference > 0, Decrease Xcurrent by index
If Xcurrent not equal to Xnew stay in loop else continue
If Vertical difference < 0, Increase Ycurrent by index
If Ycurrent not equal to Ynew stay in loop else continue
If Vertical difference > 0, Decrease Ycurrent by index
If Ycurrent not equal to Ynew stay in loop else continue
The second setting would allow the user to use a controller to aim the coil gun.
The likely outcome will be using the computer as the controller, thus using the
arrow keys on the keyboard of the keyboard. The user would be given the option
to input from the keyboard an 8(up), 2(down), 4(left), 6(right), 0(exit). Then
depending on the input the current vertical and horizontal position will be
increased or decreased by an index.
Pseudo code:
Current position -> Xcurrent, Ycurrent
Ask user for input 8,2,4,6, or 0
116
If input is 8 increase Ycurrent by index
If input is 2 decrease Ycurrent by index
If input is 6 increase Xcurrent by index
If input is 4 decrease Xcurrent by index
If input is 0 return to main menu
For the automatic setting we will be using sensor to detect light/color. If the user
selects to use the automatic setting the sensor will be turned on and begin saving
data. When the sensor detects the light/color the program will compare the
coordinates and compare with the current position of the coil gun. If the
coordinates of the detection differ from the current position the coil gun will move
to the coordinates of the object.
Pseudo code:
User chooses automatic setting
Turn sensor on
Current position -> Xcurrent, Ycurrent
Read Sensor
If sensor detects color/light grab coordinates
Horizontal difference = Xcurrent – Xnew
Vertical difference = Ycurrent – Ynew
If Horizontal difference < 0, Increase Xcurrent by index
If Xcurrent not equal to Xnew stay in loop else continue
If Horizontal difference > 0, Decrease Xcurrent by index
If Xcurrent not equal to Xnew stay in loop else continue
If Vertical difference < 0, Increase Ycurrent by index
If Ycurrent not equal to Ynew stay in loop else continue
If Vertical difference > 0, Decrease Ycurrent by index
If Ycurrent not equal to Ynew stay in loop else continue
For access of the coil gun controller we would like for there to be a password that
the user must enter. The password would consist of a number characters or
digits. For simplicity the password chosen will most likely be hard coded into the
program. To check for correctness of the password inputted a program will go
through each character or digit one by one.
117
Psuedo Code:
Ask user for password
Check character 1
If True continue
Else prompt user for password again
Check character 2
If True continue
Else prompt user for password again
…..continue
Check character 7
If True, Prompt user with main menu (manual/auto)
Else prompt user for password again
At this point and time the user will be granted access to the program allowing the
user to decide between using an automatic or manual control system.
To access the User Interface will be using a PC. The idea of using a separate
LCD screen with buttons was discussed. It is felt that the idea of using a PC is
viewed as the better option, for simplicity for the developer and the simplicity for
the user.
Controls and Software Budget
The Controls and Software budget had to be thought out because it is a wide
known fact that finding a FPGA for a low price is not going to be very easy. The
microcontrollers and rotary encoders are going to be a little lower on the cost
side. First for the FPGA the group was able to find a dealer online at ebay.com
that sells the Xilinx Spartan 3 FPGA from one hundred and twenty nine dollars.
After shipping and handling from the dealer on ebay.com the FPGA will approach
approximately one hundred and fifty dollars. Though a specific rotary encoder
has not been chosen by the group the group still has a pretty good idea about the
price range. Most of the basic Rotary Encoders, excluding the fiber optic rotary
encoders and the magnetic rotary encoders because they are quite pricy, seem
to be in the range of around five dollars apiece. The group will be looking at
purchasing two Rotary Encoders, thus the total price of two rotary encoders will
be around ten dollars. Finally the last device that will need a pricing from the
Controls and Software budget is the microcontroller. Since the group has decided
118
to go with a microcontroller from the PIC family of microcontrollers, the price of
the microcontroller will not be very expensive. The group is looking at a price of
about five dollars for just about any of the models we choose from the PIC family
of microcontrollers.
Summary of Final Software Design
User Interface
In the end the group ended using the onboard buttons, switches and sevensegment display as the User Interface, instead of integrating the coil gun through
a laptop. The User in the end uses the switches to charge the coil gun’s capacitor
bank and the fire the coil gun (discharge the capacitor bank). The mussel velocity
is outputted to the User by way of the seven-segment display.
FPGA
The FPGA that the group decided to use to embed the software this project is a
FPGA of the Spartan series from Xilinx. One of the advantages of using a Xilinx
FPGA is the fact that IDE is free to use, thus we would only need to purchase the
development board.
Within the Xilinx family the FPGA that would be chosen would be from the
Spartan 3 family of FPGA. The eight family Spartan 3 can contain anywhere from
50,000 gates to five million gates, and is meant to meet cost sensitive projects.
Below is a functional block diagram, FIGURE 8.1, of a Spartan 3 FPGA:
119
Figure 69 Spartan 3 FPGA, Permission from Xilinx LLC
Some of the features of the Spartan 3 board that project used directly is the
seven segment display to display the mussel velocity and the I/O pin
characteristics to input voltages and out voltages for different tasks. One of the
tasks that will done by the FPGA using the input voltages and output voltages is
the control of the servo motors this is done by sending out pulse width
modulations through the output pins, which was discussed briefly in an earlier
section. The second task that will be done using the FPGA is displaying the
mussel velocity of the projectile using the provided seven segment display on the
FPGA board.
SOFTWARE
For this project all of the existing software is written in Verilog HDL. All of the
Verilog code was written and implemented using the Xilinx 9.2 ISE Webpack.
The main software tasks were to implement the speed trap to capture the mussel
velocity and display that velocity on the seven segment display. The speed trap
uses two optical sensors separated by 4 cm, that when on maintain a low voltage
~ 1.5V and as the projectile impedes the sensors view the voltage drops to zero.
120
To calculate the velocity will count the number of clock cycles from when the first
optical sensor drops to 0V up until the second optical sensor goes low. An
example of the Verilog code is shown below in Figure 70:
always@ (posedge clk )
if(!sensorIn)begin
Startcount <= 1'b1;
end
else if(!sensorOut) begin
Startcount <= 1'b0;
end
always@ (posedge clk)
if(Startcount == 1'b1)
count = count + 1;
else if(Startcount == 1'b0)begin
velocity = (0.04/(count*0.00000002))*3.3;
end
Figure 70 Speed Trap Code
To display the velocity using the built in seven segment display it took some
manipulating of the fpga’s onboard 50 MHz clock. There are seven Pins for the
seven segment display and four Pins to switch between each digit, therefore
theoretically you can only display one digit at a time. Though we have the ability
to display four digits, only three will be shown since the ideal mussel velocity will
be around one hundred feet per second making the forth digit obsolete. To get
around this a counter was created to allow the display to toggle between digits
for a specified amount of clock cycles. Even still none of the digits are actually
displayed at the same time, though to the human eye it appears to be. Another
difficulty that was run into was the fact that we were trying to display a calculated
integer. To extract each digit (hundred place, ten place, one place) the integer
was first divided by one hundred and then mod by one hundred, then result from
the mod function was used to calculate the next digit by dividing by ten and the
one digit is then calculated by taking the mod of the ten digit. The code in
Fdemonstrates how to extract each digit that is to be displayed from a calculated
integer:
121
always @(posedge clk)
DIG2 = velocity/100;
REM2 = velocity%100;
DIG1 = REM2/10;
REM1 = REM2%10;
DIG0 = REM1/1;
Figure 71 Digit Extraction
Below is Figure 72 which includes example code of how to execute the toggle of
the display:
always @(posedge clk)
if(count < 15020)begin
count = count +1;
if(count > 5000)begin
toggle <= 4'b1101;
DISPOUT <= DISP0;
end
if(count > 10000)begin
toggle <= 4'b1011;
DISPOUT <= DISP1;
end
if(count > 15000)begin
toggle <= 4'b0111;
DISPOUT <= DISP2;
count = 0;
end
end
Figure 72 Digit Output with Toggle
122
Executive Summery
The overall design of a coil gun is complete. The coil gun proved to be harder to
design than originally expected. It required very in depth research of magnetic
fields. The concepts and theories of magnetic fields took the most time to
comprehend. There were numerous equations and formulations derived to have
the capability of even starting the research and design process. The magnetic
fields were on a similar but different level of comprehension than normal
electrical circuits. The comprehension required research in magneto-motive
forces and fluxes. Theories from the Biot-Savat ‘s laws, Ampere’s laws, and
Maxwell equations were the basis of the relevant research to magnetic fields.
They were the theories and laws that were used to grasp the conceptual idea of
getting the projectile to actually move. They were also the theories that put
constraints and prohibited many different methods to be used by different
electrical components. The main conceptual idea for the coil gun was to get a
magnetic field to create via current through a coil to have a projectile shot from a
barrel. That design of the magnetic proved difficult because it affected many
devices that would help the configuration of the coil gun.
The back
electromagnetic fields became troublesome because they would negatively affect
the operation of the coil guns circuit components.
Choosing the capacitor bank was the most important part in the development of
the coil gun. The development of currents, induction, and voltages were centered
on the capacitor bank. They are the reason all of those values were chosen.
Since the gun was built on certain goals and specifications the need for the
capacitor bank to be the correct value was pertinent in the design process. The
capacitors were the first and last electrical components chosen. They were
continuously being adjusted to deal with the speed at which the projectile needed
to be fired at. By using a RLC simulator the values of the capacitor bank were
chosen on the bases of current pulses needed in order to achieve different
projectile speeds. The capacitors took a lot of research to find its appropriate
values. They were the most tedious to find because picking the wrong capacitors
would be very expensive. After careful research the selected values were found.
The capacitors like the magnetic fields presented many obstacles to overcome.
Choosing the switching system to switch on the current from the capacitors was
not too hard. The difficultly of the process was due to the numerous choices.
There were many different transistor choices to be chosen for the design of the
coil gun circuit. All of the choices presented valuable advantages to the circuit
and were carefully considered. The disadvantages and advantages of the
transistors were weighed to pick the most appropriate choice. The choice leaded
to the selection of a silicon-controlled rectifier. The silicon-controlled rectifier met
all of the circuit requirements and integrated into the circuit without having to
compensate for any changes to circuit elements. This was very important
123
because when the design is actually constructed currents and voltages may be
varied.
The type of wire was pertinent to the magnetic field generation. Without the
proper selection of the wire the required magnetic field will not be able to be
created. The wire selected was selected because it was capable of handling the
transfer of large currents and voltages at short durations for peak current pulses.
The selection of wire proved to be very tedious because of thermal sustainability.
The coil needed to have wire that can support extremely high temperatures. The
type of wire chosen is ready to withstand anything thrown at it. The wire being
used will contribute to the coil gun operating successfully. The development of
the magnetic field will be exerted through the wire winded around the barrel.
The barrel was probably the easiest part of the design process. The barrel was
designed on the basis that it would be able to allow a small projectile to be shot
from. After the projectile was chosen it was easy to design the barrel for the coil
gun. There was only one real requirement in picking the barrel and that was the
material. For the magnetic field to be exerted properly on the projectile the barrel
needed to be non-conductive. None the less the barrel was very important in the
development of the coil gun design. Without the barrel the projectile would have
no path to be shot. Without that extra non-conductive material between the coil
and the projectile the gun would not be able to shoot.
Partial Circuit Schematic Overview
The main circuit shows how the coil gun circuit will be set up. The circuit begins
with the power source. The power source being used is a standard United States
120V 60Hz alternating current power. The current from the power supply travels
through a wire to the full- wave bridge rectifier. There are four diodes that are
rated with 20 amps each. From this full-wave bridge rectifier the alternating
current from the power source is converted to DC power. The current from the
full-wave bridge rectifier is then sent through a 125 watt flood light bulb to the
capacitor bank.
The 125 watt bulb is used as the charge resistor for the
capacitor bank. At the capacitor bank the current begins to charge until the
maximum charge is reached. There is a switch connected to the capacitor bank
to allow it to begin charging. The capacitor bank is connected in parallel with a
diode, that diode is acting as the flywheel diode. The cathode end of the diode is
facing upwards. When the capacitor bank is discharged the charges are
released from the capacitor bank. Then there is a large current sent to the coil
that is winded around the barrel. The capacitor bank has a capacitance of four
millifarads. There is also a resistor connected in parallel to the capacitor bank.
The resistor connected to the capacitor bank is the bleed resistor. The function of
124
the bleed resistor is to drain the energy from the capacitors when it is stored or is
not in use. The value of the bleed resistor that was chosen to be used is a
hundred and fifty kiliohms. The solenoid is attached in parallel with a diode just
like the capacitor bank is. This is the flywheel diode for the coil or inductor. In
series with the end of the coil there is a silicon-controlled rectifier attached. This
silicon-controlled rectifier is turned on only when a current is supplied to its gate.
The gate of the silicon-controlled rectifier will be attached to a triggering system
that is not shown in the figure. The cathode end of the silicon-controlled rectifier
is attached to a resistor. This resistor is the damping resistor and keeps the
circuit critically damped. The value of the damping resistor is .593 ohms. The
circuit does not need to be grounded since it’s hooked up to an AC power outlet.
It is automatically grounded through the outlet. Figure 68 below shows the main
coil armature.
Circuit Schematics and PCB board
Figure 73 Circuit Schematics
Figure 73 is the Schematic of the whole circuit of the coil gun. It constructs by
four parts which are voltage transform, voltage rectifier, charge circuit and fire
circuit. And the corresponding PCB design looks like as shown on Figure 74.
125
Figure 74 PCB for Coil Gun
Subsystem
Field Generation
Sensors and Motion Control
Power
Controls and Software
Project Total
Cost
$343.98
$165
$20
$165
$693.98
Table 16 Coil Gun Total Budget Breakdown
126
Milestones
The following Gantt chart shows the schedule of the project. The Gantt chart
displays the dates on the top of the chart. On the right side of the Gantt chart
shows each of the milestones for the project in their own specific subsystem.
Field Generation(Research)
Field Generation(Design)
Field Generation(Procurement)
Field Generation(Build)
Field Generation(Test)
Power Systems(Research)
Power Systems(Design)
Power Systems(Procurement)
Power Systems(Build)
Power Systems(Test)
Controls (Research)
Controls (Design)
Controls (Procurement)
Controls (Build)
Controls(Test)
Sensor(Research)
Sensor(Design)
Sensor(Procurement)
Sensor(Build)
Sensor(Test)
User Interface(Research)
User Interface(Design)
User Interface(Procurement)
User Interface(Build)
User Interface(Test)
Automation(Research)
Automation(Design)
Automation(Procurement)
Automation(Build)
Automation(Test)
Figure 75 Milestones
127
User Manual
1. Make sure all switches are in the OFF position, and load the projectile into
the barrel of the Coil Gun.
2. Connect the load the program into the FPGA, the Seven Segment Display
should read 0000, and then connect the FPGA to the PCB.
3. Plug in the AC power cable into the wall outlet.
4. Locate the switch box on the back of the capacitor bank. Turn on both
switches; one is for the AC Power and the other for the DC Power. You
should hear a buzz from the transformer when the AC switch is in the ON
position and a LED should light up when the DC switch is in the ON
position.
5. Now locate the two switches on the side of the capacitor bank box. The
red switch is for the DC Voltage Meter; turn it to the ON position. The DC
Voltage should then be reading about 0V. The white switch is to the
Charging circuit; Turn it to the ON position.
6. Now that the Charging switch is in the ON position we can Charge the
capacitor bank by using switch 6 on the FPGA board. Once Switch 6 is in
the logic 1 position, the DC Volt meter should display an increasing
Voltage.
7. When the DC Volt meter display 400V turn Switch 6 to the logic 0 position,
the Voltage should then stop increasing and start to very slowly decrease.
8. At this time the Coil Gun is ready to be fired. The Firing circuit is controlled
by Switch 7. When you are ready to Fire turn Switch 7 to the logic 1
position. The Coil Gun will then Fire, at this time turn Switch 7 back to the
logic 0 position.
9. You will then see the mussel velocity of the projectile displayed on the
Seven Segment Display. To Reset the Velocity press Button 0 on the
FPGA.
10. Turn all the switches to the OFF position. If the User would like to use the
Coil Gun immediately reload the projectile into the barrel and continue at
Step 4, else unplug the AC cable.
128
Appendix A
Date: Mon, Aug 2, 2010 at 11:57 AM
Subject: Re: Permission to reference your web site.
To: Josef Von Niederhausern <josef.von@gmail.com>
Hello Josef,
Yes you may use it.
Thanks,
Donnie James
----- Original Message ----From:Josef Von Niederhausern
To:admin@anothercoilgunsite.com
Sent: Monday, August 02, 2010 7:47 AM
Subject: Permission to reference your web site.
Sir or Madam,
For our senior design project we are building a coil gun. Your website is very
informative and helpful we are grateful to you for that. We would like your
permission to use your images as figures in our write up. Of course we will credit
your work and list your website, www.anothercoilgunsite.com, as a reference.
Thank You,
Josef Von
Directed to use:
http://www.ti.com/corp/docs/legal/termsofuse.htm?DCMP=TIFooterTracking&HQ
S=Other+OT+footer_terms for permission from TI

RE: Permission to use content
7/30/10
Ronald Wilson
ron.wilson@ubm.com
129
To brian.hoehn@knights.ucf.edu
From:Ronald Wilson (ron.wilson@ubm.com)
Sent: Fri 7/30/10 12:52 AM
To: brian.hoehn@knights.ucf.edu (brian.hoehn@knights.ucf.edu)
Brian:
Please do so.
ron
From: brian.hoehn@knights.ucf.edu [mailto:brian.hoehn@knights.ucf.edu]
Sent: Tuesday, July 27, 2010 4:28 PM
To: Ronald Wilson
Subject: Permission to use content
Ron Wilson,
My Electrical Engineering senior design group at the University of Central Florida
is doing a project which includes a FPGA device. I would like to ask for your
permission to use some figures and images from the following article.
http://www.eetimes.com/design/programmable-logic/4006429/FPGAprogramming-step-by-step
All of the information used will be credited to your website.
Thank you,
BrianHoehn
University of Central Florida

RE: Permission requests
7/28/10
Marc.McComb@microchip.com
130

Marc.McComb@microchip.com
Add to contacts
To brian.hoehn@knights.ucf.edu
From:Marc.McComb@microchip.com
Sent: Wed 7/28/10 7:58 PM
To: brian.hoehn@knights.ucf.edu
Attachments, pictures and links in this message have been blocked for your
safety. Show content | Always show content from this sender
Hi Brian:
Thanks for your email. Please visit this webpage that should answer your
questions:
http://www.microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=4
87&param=en023282
If you have any questions please let me know.
Regards,
Marc McComb
Academic Program Sales Engineer
2355 West Chandler Blvd., Mailstop 9-B
Chandler, AZ85224-6199
office: 480-792-4391
mobile: 480-478-5676
Need more information on Microchip's Academic Program or would like to
become a Partner? Visit www.microchip.com/academic for more information.
131
From: brian.hoehn@knights.ucf.edu@MICROCHIP
Sent: Tuesday, July 27, 2010 3:51 PM
To: University; Eric Lawson - C11906
Subject: Permission requests
Hello,
My Electrical Engineering senior design group at the University of Central Florida
is designing a coil gun for our project. We may be using one of your products and
would like to ask for permission to include some of Microchip's data sheets,
including figures and diagrams within the data sheets. All information used in the
report from your website and data sheets with be credited to Microchip.
Thank You,
BrianHoehn
University of Central Florida

RE: Permission to reference your web site.
7/26/10
To 'Josef Von Niederhausern', chic0lindo@hotmail.com,
brian.hoehn@knights.ucf.edu, gorvyng@gmail.com
From:Barry Hansen (barry@coilgun.info)
Sent: Mon 7/26/10 4:07 PM
To: 'Josef Von Niederhausern' (josef.von@gmail.com)
chic0lindo@hotmail.com; brian.hoehn@knights.ucf.edu;
Cc:
gorvyng@gmail.com
Attachments, pictures and links in this message have been blocked for your
safety. Show content | Always show content from this sender
Hi Josef (and everyone!),
Thanks for checking, you are welcome of course to reference my coilgun website.
Are you at the Univ of Central Florida? I’m not very familiar with it, but it must be
132
Have a great day,
Barry Hansen
www.coilgun.info
From: Josef Von Niederhausern [mailto:josef.von@gmail.com]
Sent: Monday, July 26, 2010 8:03 AM
To: barry@coilgun.info
Cc: chic0lindo@hotmail.com; brian.hoehn@knights.ucf.edu;
gorvyng@gmail.com
Subject: Permission to reference your web site.
Mr. Hansen,
For our senior design project we are building a coil gun. Your website is very
informative and helpful we are grateful to you for that. We would like your
permission to use your images as figures in our write up. Of course we will credit
your work and list your website, www.coilgun.info, as a reference. The carbon
copies listed are my co-designers.
Thank You,
Josef Von Niederhausern
HI Brian,
You can accept this email as permission to use the datasheet in your project.
Thank you,
Cynthia
Cynthia Zamorski Legal Counsel
Xilinx, Inc. 2100 Logic Drive San Jose, California 95124
(408) 879-4638 Mobile (408) 892-189
133
Quoting gorvy <gorvy@knights.ucf.edu>:
>
>
>
>
>
I am really interest in the content you mention in " Control of
Stepping motors" and some of the graphs are very helpful for me to
explain the how to control the stepping motor. so I want to ask you
kindly if it's possible to use some of the graphs on my paper for my
project. let me know. thank you.
You are welcome to selectively quote and selevtively borrow
illustrations from my material on the web, on the condition that
you properly cite your sources and give appropriate credit in
the tradition of good scholarship. Thank you for asking.
-- Doug Jones
-- jones@cs.uiowa.edu
134
Appendix B - References
[1]Xilinx (2010) FPGA and CPLD Solutions, www.xilinx.com
[2]MicroChip Technology Inc. (2010) PIC family of Microcontrollers Research ,
www.microchip.com
[3]Texas Instruments Inc. (1995-2010) TI MSP430,
http://focus.ti.com/mcu/docs/mcuprodoverview.tsp?sectionId=95&tabId=140&familyId=
342
[4]Atmel Corporation (2010) Atmel AVR http://www.atmel.com/products/avr/
[5]Steve Trahey (2008) Choosing a code wheel: A detailed look at how encoders work
http://www.smallformfactors.com/articles/id/?3039
[6] Douglas W. Jones “Micro stepping of Stepping Motors”
http://www.cs.uiowa.edu/~jones/step/micro.html
[7] Jieun Kwon and Yuncheol Baek “Real-time Interactive Media Design with Camera
Motion Tracking” http://www.sd.polyu.edu.hk/
[8] Ted T.Lin “Understand Stepper Parameters Before Making Measurements”
http://www.tmworld.com/article
[9] Orientalmotor “AC or DC? Brushed or Brushless?”
http://www.orientalmotor.com/MotionControl101/AC-brushless-brushed-motors.html
[10] Andrew Caballero & Ian Stine “Auto-Targeting Sentry Gun with Friend/Foe
Recognition”
http://www.eecs.ucf.edu/seniordesign/fa2008sp2009/g08/Group8SeniorDesignDoc2.pdf
[11] Pridgets Products for USB Sensing and Control (2010), http://www.phidgets.com/
[12]Servo Motor Control (2009), http://www.digilentinc.com/
[13] “Programming-Computer Vision Tutorial”
http://www.societyofrobots.com/programming_computer_vision_tutorial.shtml
135
Appendix C- M files or Code
function [S] = fuse(AWG)
% given A and S
% Wire Sizes used in Fuses
%
% The Standard Handbook for Electrical Engineers lists the following
formula:
%
% 33 * (I/A)^2 * S = log( (Tm - Ta) / (234 + Ta) + 1 )
%
* I = current in Amperes
%
* A = area of wire in circ. mils
%
* S = time the current flows in seconds
%
* Tm = melting point, C (copper's melting point is 1083 C)
%
* Ta = ambient temp, C
A = (5*92^((36-AWG)/39))^2;
Ta = 27;
Tm = 1083;
rs= log10((Tm - Ta)/(234 + Ta) + 1 );
srs = rs*(A^2)/33;
S = (srs/.001)^.5;
function [ dn ] = AWG_dn( AWG )
% finds the diameter of AWG in mm
dn = .127*92^((36-AWG)/39);
function [ area ] = AWG_An_Meter( AWG )
area = (pi/4)*(.001*AWG_dn( AWG ))^2;
function [ B ] = B_tesla_midcoil_j(l,r1,r2,i,AWG)
%
B_tesla_midcoil_j(l,r1,r2,i,AWG)
u0 = 1.26*10^-6; % permability of free space
n1=sqrt(r2^2+(l/2)^2)+r2; % neumerator
d1=sqrt(r1^2+(l/2)^2)+r1; % denominator
dn= AWG_dn(AWG)*.001; % diameter of wire (meters)
rn = dn/2;
% radius of wire (meters)
layers=floor((r2-r1)/2/dn); % number of layers
turnlayer = floor((l)/dn); % number of turns (linear)
n = layers*turnlayer;
% total turns
% disp('winding density')
n_density = 1/l*n;
% turn density per meter
% disp('Feild')
wire_cross_area = pi*(rn)^2; % in meters^2
coil_cross_area = l*(r2-r1); %coil cross sectional area
j = i*n/coil_cross_area; %current density for coil cross sectional area
B = u0*j*l/2*log(n1/d1);
function [ An ] = AWGturnmeter( AWG )
% A is cross sectional area d is diameter in inches
% d = .005*92^((36-AWG)/39);
136
% A = pi/4*d^2;
D = (0.000127*92^((36-AWG)/39));
An=1/D;
function [B] = B_tesla_midcoil(l,r1,r2,i,AWG)
%
B_tesla_midcoil(l,r1,r2,i,AWG)
%
Finds the on axis strength of magentic field in the midoimt of a
coil
%
also finds: inducatnce, turn density,
u0 = 1.26*10^-6; % permability of free space
n1=sqrt(r2^2+(l/2)^2)+r2; % neumerator
d1=sqrt(r1^2+(l/2)^2)+r1; % denominator
dn= AWG_dn(AWG)*.001; % diameter of wire (meters)
rn = dn/2;
% radius of wire (meters)
layers=floor((r2-r1)/2/dn); % number of layers
turnlayer = floor((l)/dn); % number of turns (linear)
n = layers*turnlayer;
% total turns
% disp('winding density')
n_density = 1/l*n;
% turn density per meter
% disp('Feild')
B = u0*i*n/2/(r2-r1)*log(n1/d1);
% streangth of field in tesla's
% z = i*n/(r2-r1) %%%Checker
% lh = u0/2*log(n1/d1)
function [ B ] = B_tesla2(l,r1,r2,x,i,AWG)
%
B_tesla2(l,r1,r2,x,i,AWG)
%
l length of coil, r1 inner radius, r2 outter radius, x2 begining
coil
%
to point B, x1 end of coil to point B, n number of turns, i current,
%
rho is conductor resisitivity in units of ohms-length must match
that
%
of r1, lambda is cross sectional area of wire, P is Power
%
G is maximum when alpha = 3 and beta = 2
% rho= 1.724*10^-8; % resistivity of copper per
hyperphysics.phyastr.gsu.edu/hbase/electric/resis.html
rho= 1.68*10^-8; % resistivity of copper per wiki
u0 = 1.26*10^-6;
x1 = x-l/2;
x2 = x+l/2;
B_alpha = r2/r1;
B_beta = l/(2*r1);
B_gamma = (x1+x2)/(2*r1);
dn= AWG_dn(AWG)*.001; % diameter of wire (meters)
rn = dn/2;
% radius of wire (meters)
layers=floor((r2-r1)/2/dn); % number of layers
turnlayer = floor((l)/dn); % number of turns (linear)
n = layers*turnlayer;
% total turns
n_density = 1/l*n
% turn density per meter
%%%%%%%%%%%%%%%%%%% finds average radius of layers
137
count = 1;
atotal = -rn+r1;
while count ~= layers+1;
atotal = atotal + dn;
a_tot(1,count) = atotal; %finds the average radius of the layers
from center axis
count = count+1;
end%finds the average radius of the layers from center axis
average_radius_of_turn = sum(a_tot)/layers;
%finds the average
radius of the layers from center axis
c= r2-r1;
induct
= .8*average_radius_of_turn^2*n^2/(6*average_radius_of_turn+9*l+10*c) %
inductance of coil in H
length_of_wire = sum(2*pi*turnlayer*a_tot(1,:));
wire_cross_area = pi*(rn)^2 % in meters^2
% resistance = rho*length_of_wire/wire_cross_area
% linear resistance
P=i*i*resistance
%power consumed by coil
lambda = (r2^2-r1^2)/r2^2; % end cross area
% lambda = (r2-r1)*l; % length wise crossarea
n1= B_alpha + (B_alpha^2 + (B_gamma + B_beta)^2)^.5; %numerator 1
d1 = 1 + (1 + (B_gamma + B_beta)^2)^.5;
%denominator 1
n2= B_alpha + (B_alpha^2 + (B_gamma - B_beta)^2)^.5; %numerator 2
d2 = 1 + (1 + (B_gamma - B_beta)^2)^.5;
%denominator 2
C1 = (8*pi*B_beta*(B_alpha^2-1))^-.5;
G2 = (B_gamma + B_beta)*log1p(n1/d1);
G3 = -(B_gamma - B_beta)*log1p(n2/d2);
G = C1*(G2+G3);
B1=((P*lambda)/(r1*rho))^.5;
B = u0*B1*G; % feild strength in teslas
% factor 1
% factor 2
% factor 3
% unitless G factor
% coefficient 2
function [ B ] = B_tesla(r1,r2,l,x,n,i)
%
B_tesla(r1,r2,l,x,n,i)
u0 = 1.26*10^-6;
x1 = x-l/2;
x2 = x+l/2;
n1=sqrt(r2^2+x2^2)+r2;
d1=sqrt(r1^2+x2^2)+r1;
n2=sqrt(r2^2+x1^2)+r2;
d2=sqrt(r1^2+x1^2)+r1;
n = 1/l*n;
g = x2*log(n1/d1)-x1*log(n2/d2)
B = u0*i*n/2/(r2-r1)*g;
138