Investigation of A High-Power Virtual Cathode Oscillator

INVESTIGATION OF A HIGH-POWER
VIRTUAL CATHODE OSCILLATOR
by
XUPENG CHEN, B.E., M.S.
A DISSERTATION
IN
APPLIED PHYSICS
Submitted to the Graduate Faculty

of Texas Tech University in
Partial Fulfillment of
the Requirements for
the Degree of
DOCTOR OF PHILOSOPHY
Approved
Accepted
Dean of the Graduate School
December, 2003
ACKNOWLEDGEMENTS
I would like to express my deepest appreciation to Dr. J. Dickens, Dr. L. L.
Hatfield, Dr. M. Kristiansen, Dr. H. Krompholz, and Dr. A. Neuber for then- support and
invaluable technical advice throughout this project. I would like to thank Dr. J.
Mankowski for his technical advice throughout the experimental part of this project. I
would like to thank Dr. E. Choi for his technical advice throughout the theoretical part of
this project. I would like to thank Chris Hatfield and Daniel Garcia for their efforts in
repairing the coaxial VIRCATOR.
I would also like to express my gratitude to Dr. D. Hemmert, Dr. J. Kim, Mr. M.
Lara, and Mr. J. Spears who shared their experiences and their ideas with me.
I am gratefiil to Dr. W. Jiang for his introduction to this project in my start-up
period.
I would like to thank my fiiends for their helpful discussions and comforts when I
had troubles during this project.
Finally, I would like to th£tnk my wife, Chunyan Fu, for her support and
understanding throughout the duration of this project. I would like to thank my son,
James Chen, for his bringing a lot of enjoyment to my life.
This work was supported by AFOSR and FMV, Sweden.
u
TABLE OF CONTENTS
ACKNOWLEDGEMENTS ii
ABSTRACT v
LIST OF TABLES vi
LIST OF FIGURES x
CHAPTER
1. INTRODUCTION 1
2. SPACE-CHARGE LIMITED CURRENT FOR COAXL\L DIODES 4
2.1 Poisson's Equation for Planar Diodes and the Child-Langmuir Law 4
2.2 Poisson's Equation for 1-D Cylindrical Diodes 5
2.3 Poisson's Equation for 1-D Cylindrical Diodes Simplification and

1-D Current Density 8
2.4 Comparison of Langmuir-Blodgett Law with Our Approximation

Result 13
2.5 Relativity Corrected Space-Charge Limited Current for the

Cylindrical Diode 17
2.6 Space-Charge Limited Current for 2-D Coaxial Diodes 24
3. APPLICATION OF ANALYTICAL CURRENT EXPRESSION IN

DIODE DESIGN 30
3.1 Current Evaluation of Coaxial Diodes 30
3.2 Virtual Cathode Position Determination and Virtual Cathode
Oscillation Frequency 31
4. EXPERIMENTAL CONFIGURATIONS AND VIRTUAL CAVITY 36
4.1 Experimental Configurations 38
ui
4.2 Experimental Data Analysis 41
4.3 Experimental Results 43
4.4 Cavity Resonance Effect and Virtual Cavity 46
4.5 The Function of the Reflectors 48
5. PROPAGATING MODE DETERMINATION 54
5.1 Cutoff Frequencies of the Waveguide 54
5.2 Field Distribution for Different Modes in a Circular Waveguide 56
5.3 Wave Combinations 56
5.4 Theoretical Analysis for Microwave Modes 60
5.5 Analysis of Experimental Data 62
5.6 Discussions on Propagatmg Modes 80
6. CONCLUSIONS AND SUGGESTIONS 84
REFERENCES 87
APPENDIX 89
IV
ABSTRACT
In this dissertation, three topics are investigated in different degrees: space-charged
limited current for a vacumn cylindrical diode, microwave resonance effects in coaxial
diodes and microwave propagating mode identification.
By introducing a physics approximation method into analyzing the nonlinear
Poisson's equation, an analytical expression for the space-charge-limited current density
for a one-dimensional (1-D) cylindrical diode has been investigated and developed.
Applying the approximation method, we get the relativistic theory corrected current for
1-D cylindrical diodes. This 1-D result is extended to 2-D.
Some experiments at TTU show that the interaction between the virtual cathode
oscillation and the cavity is a key to determine the microwave frequency and propagating
mode. Particularly, we observe that the E-beam plays an important role in cavity
formation. Some possible explanations are proposed.
The methods to determine microwave propagating modes are reviewed and some
suggestions are proposed to identify microwave propagating modes.

LIST OF TABLES
3.1. Virtual Cathode Position Predictionfi-omDifferent Models 32
4.1. Examples of Frequency Record in Cavity Effect Research of a Coaxial

Vircator at TTU 44
4.2. Typical Experimental Records for a Certain Diode Geometiy 45
4.3. An Example Experunental Records for a Certain Diode Geometiy 45
4.4. Frequency List for Geometries without Reflectors 46
4.5. Frequency List for Geometries with a Donut Reflector 49
4.6. Frequency List for Geometries with a Strip Reflector 50
5.1. x„„ Array 55
5.2. x'„m Array 55
5.3. Cutoff Frequencies (GHz) for TMnm Mode 55
5.4. Cutoff Frequencies (GHz) for TEnm Mode 56
A.l. Frequency Record 1 in Cavity Effect Research at TTU 90
A.2. Frequency Record 2 in Cavity Effect Research at TTU 90
VI
A.IO. Frequency Record 10 in Cavity Effect Research at TTU 94
A.l 1. Frequency Record 11 in Cavity Effect Research at TTU 95
A.16. Frequency Record 16 m Cavity Effect Research at TTU 97
A.31 Frequency Record 31 in Cavity Effect Research at TTU 105
vu
A.43. Frequency Record 43 in Cavity Effect Research at TTU Ill
A.44. Frequency Record 44 in Cavity Effect Research at TTU Ill
A.49. Frequency Record 49 m Cavity Effect Research at TTU 114
vni
IX
LIST OF FIGURES
2.1. Cross-section of a Planar Vacuum Diode Geometry 4
2.2. Cross-section of a Cylindrical Vacumn Diode Geometry 6
2.3. Static Field Enhancement Correction for (a) Rc>Ra and (b) Rc<Ra 10
2.4. Comparison between Langmuir-Blodgett Law and Analytical Expression for

Case 1: Cathode Radius=0.132 m,Rc>Ra,Voltage=500kV 16
2.5. Comparison between Langmuir-Blodgett Law and Analytical Expression for
Case 2: Cathode Radius=0.132 m, Rc<Ra, Voltage=500 kV 17
2.6. Non-relativistic and Relativistic Normalized Impedance Comparison 24
3.1. Current Predictions from Different Mathematical Models 31

3.2. Election Reflection Frequency Evaluation for TTU Coaxial Vircator.
Rc=0.132m, Ra=0.098m 33
3.3. Comparison of Vircator Frequency with Election Reflection Frequency 35
4.1. Overall Sketch of the Experimental System 37
4.2. Experimental Configuration 38
4.3.E-probe Schematics 39
4.4. Waveforms Directly Obtained by the E-probes 41
4.5. Fast-Fourier Transformation of the Waveforms Shown in Figure 4.4 42
4.6. Schumann Cavity around the Earth and in the Diode Region 47
5.1. Transverse Field Distribution for a Circular Waveguide (First 15 Modes) 58
5.2. Combination of Three Time Harmonic Traveling Waves with Slightly

Different Frequencies: 1-10 s 59
5.3. Raw Horizontal and Vertical Probe Waveforms for a Shot: -20-180 ns 63
5.4. FFT Spectia for Waveforms Shown in Figure 5.3 64
5.5. Filtered Waveforms for the Horizontal Probe Shown in Figure 5.3: 0-100 ns 65
5.6. Filtered Waveforms for the Vertical Probe Shown in Figure 5.3: 0-100 ns 67
5.7. Filtered Waveforms for the Fields Shown in Figure 5.3: Phase Difference 69
5.8. Filtered Waveforms for the Horizontal Probe Shown m Figure 5.3 with Two
Narrower Band Filters: 0-100 ns 71
5.9. Filtered Waveforms for the Vertical Probe Shown in Figure 5.3 with Two
Narrower Band Filters: 0-100 ns 72
5.10. Filtered Waveforms for the Fields Shown in Figure 5.3 with Two Narrower
Band Filters: Phase Difference 73
5.11. Raw Waveforms for a Shot: 50-350 ns 74
5.12. FFT Spectia for Waveforms Shown in Figm-e 5.11 75
5.13. Filtered Waveforms for Horizontal Probe Shown in Figure 5.11: 50-350 ns 76
5.14. Filtered Waveforms for the Vertical Probe Shown in Figure 5.11: 50-350 ns 78
XI
CHAPTER 1
INTRODUCTION
The coaxial vircator has been studied as a potential high-power microwave source
[1,2]. It has a very simple geometry, which makes it very easy to fabricate, to operate
and to maintain. It does not need an external applied magnetic field. It can have a high
power level of operation, currently operating in the hundreds of megawatts to several
gigawatts range for a single shot. The vircator has many prospective appHcations, such
as high energy particle accelerators, electionic warfare and motor engines.
We have to overcome its fatal drawback, low power efficiency, before it can be
put into applications. Currently, the best power efficiency of the coaxial vircator is - 5 %
[2]. Recent research on the coaxial vircator primarily focuses on its physical mechanisms
to improve its power efficiency.
It is believed that the vircator generates microwaves by two mechanisms, one is
the election reflection movement between the real and virtual cathodes and the other is
the virtual cathode oscillation of the collected electrons. The physical foundation of the
two mechanisms is the bremsstiahlung radiation or braking radiation: the energy loss
through electiomagnetic waves fi-om the braking of relativistic particles. To set up the
braking environment for the relativistic election beam in a vacuiun diode, the space-
charge limited current is intioduced. It is supposed that the virtual cathode forms by the
accelerated elections passing through the mesh anode once the diode current exceeds the
space-charge limited current. After the virtual cathode is set up, the election reflection
mechanism is setup and the virtual cathode itself will oscillate [3]. Some experiments
show that the election reflection mechanism contiibutes little to the microwave
generation and they designed a device called a redition to minimize the election
reflection movement. Reported efficiency from the redition ranges fi-om 5.5 to 6.0
percent [4]. The vutual cathode oscillator is so called, to some degree, because the
virtual cathode oscillation mechanism dominates m microwave generation. If the
electron reflection mechanism dominated in microwave generation, it should not be
necessary to form the virtual cathode and a physical reflection cathode should be enough
to maintain a working device.
In addition, the space-charge limited current is determined by the diode geometry
and the applied voltage. The Child-Langmuir law [5] is used to predict the space-charge
limited current for the planar vacuum diode. The Langmuir-Blodgett law [6] is used to
predict the space-charge limited current for the cylindrical vacuum diode. However, the
Langmuir-Blodgett law is a nvunerical solution from a truncated power series solution of
the nonlinear Poisson's equation. This makes it difficult in optimizing the diode design.
A good prediction for the space-charge limited ciurent is helpful in evaluating the virtual
cathode oscillation frequency and election reflection frequency. The frequency
evaluation becomes very important in improving the power efficiency.
The resonance cavity is a basic technology in the microwave area. Research on
the resonance effect in an axial vircator was done by J. Benford, [7]. They found that the
interaction between a vircator microwave generator with an enclosed resonant cavity will
determine the output microwave propagating mode and the non-resonant frequencies are
suppressed. This is common in a resonance cavity. An idea to design a similar cavity in
a coaxial vircator comes up immediately. Since the coaxial vircator has a different
structure, in experiments we find that the coaxial vircator shows a different resonance
behavior. These experiments tell us that the E-beam plays an important part in forming a
cavity in the coaxial diode region. This cavity is not an enclosed structure. This
characteristic is different from conventional cavities. We defme it as a virtual cavity.
Before these experiments were conducted. Dr. Jiang had used similar structures,
with reflectors, to enhance the power efficiency of the coaxial vircator and got some
obvious achievements [8]. However, Dr. Jiang did not pay much attention to the
microwave frequency and propagating mode determination mechanism. One of reasons
was that he did not have a fast oscilloscope that time. A strip reflector and a donut
reflector characteristic are investigated and it is found that they have different influences
on the frequency determination.
To clarify the propagating mode, we do several attempts and get some hints.
However, it is still undetermined since the propagating modes are very complicated. We
believe that the propagating microwaves are a combination of different modes and
different frequencies, which make the mode identification very hard. Because the
propagating mode is not very clear, so far, we cannot calculate the microwave power
accurately. Although it appears that the microwaves are a mixture of TEn and TMQI
modes [1], we have no direct evidence to support this result. The propagating mode
determination limits oiu" resonance cavity research. Based on the above reasons, the
mode determination will be one of our possible fiiture research directions.

CHAPTER 2
SPACE-CHARGE LIMITED CURRENT FOR COAXL^L
DIODES
2.1 Poisson's Equation for Planar Diodes and the Child-

Langmuir Law
Figure 2.1 is the cross-section of a planar vacumn diode. The parallel electiodes
have infinite area. Because the minimum potential is a thin layer very close to the
cathode, we can assume that its position is located at the cathode, neglecting the potential
difference from the actual cathode. Another assumption is that the emitted elections have
a vmiform zero initial velocity at the cathode. In other words, the emitted election initial
temperature is zero.
x=0 x=d
Figure 2.1. Cross-section of a Planar Vacuum Diode Geometry
For the planar diode, the Child-Langmuir law gives the space-charge limited
ciurent. Because we will refer to the differential equation for the planar diode and its
solution in the future, let us list them here [5].

d^V(x) J \rn I \ •,,•>
dV(x)
=0 (2.2)
dx 1=0
f^(0) = 0 (2.3)
nd) = V, (2.4)
4^0 |2e^^3/2
•^ = 7 ; ^ . — ^ o ' ' ' (2.6)
9^/' V w
where w, e, A Ax and J represent the election mass, the election charge, the gap distance,
the distance of any point in the diode region from the cathode, and the current density,
respectively. Here we assume that the plasma in the diode region is fully non-neutial and
the Z-pinch from the self-magnetic field is negligible.
2.2 Poisson's Equation for 1-D Cylindrical Diodes
Figure 2.2 is the cross section of a cylindrical diode, where the outer conductor is
the cathode. The planar diode and the cylindrical diode have the same physical
mechanism and the difference between them is just the geometry structure. Here we
accept all physical assumptions mentioned above for the planar diode. Those are. the
length of the electiodes is infinite, the E-field at the cathode is zero and the election
initial velocity is zero at the cathode.

Cathode
V(Rp)=0
|Rp-RJ=d
Anode
Figmre 2.2. Cross-section of a Cylindrical Vacuum Diode Geometry
First, let us use the geometry illustiated in Figure 2.2 to derive Poisson's equation.
We assume that the potential at the cathode is zero, directly based on our experience with
the planar diode. We define E(r), V(r), J(r), p(r) and v(r) as tiie E-field, the electiic
potential, the current density, the election density, and the election velocity at the
position r, respectively. We have the following relations
£(r) = r£(r) = - r ^ ^ . (2.7)

dr
Neglecting the ion contiibution to the current because of their heavy mass, we have
J{r) = rJ{r) = fpir)v{r) , (2.8)
where both p(r) and v(r) are negative values everywhere because of the election carrier.
Based on the current continuity or the charge conservation, we have

1 ri
V • J ( r ) = 0 =^ - — ( ^ ^ W ) = 0. (2.9)
r or
That is,
J{r) = (2.10)
where C is a constant of integration. Assuming that the injected current density from the
C
cathode '\sJ{RJ = J^, since J^ = — , we get
R„
J{r) = J . ^ . (2.11)
r
Substituting Eq. (2.8) into Eq. (2.11), we get
R^
p{r) = J, (2.12)
rv{r)
Based on energy conservation, we have
.1/2
^2e,
v(r) = -{V(r)) (2.13)
ym
Substitiiting Eq. (2.13) mto Eq. (2.12), we get
-1/2
pir) = r-jAi-nr)'
\ m
(2.14)
Then the one dimensional Poisson's equation can be written as
1/2
dr dr (2.15)
-111
with k = J^-^\ —
SoVm
The bovmdary conditions for Eq. (2.15) are
V{R^ = 0, (2.16)
dV{r)
=0
dr
(2.17)
nRJ = V,. (2.18)
So far, we have set up Poisson's equation, Eq. (2.15), and the boimdary conditions,
Eqs. (2.16)-(2.18) for tiie cylindrical geometiy illustiated m Figure 2.2, in which the outer
conductor is the cathode. Eqs. (2.15)-(2.18) are also valid for the case in which the outer
conductor is an anode.
2.3 Poisson's Equation for 1-D Cvlindrical Diodes

Simplification and 1-D Current Density
Let us consider a cylindrical diode with geometry of Rc>Ra- We can see that Eq.
(2.15) is a nonlinear differential equation and it is unpossible to get a simple analjrtical
solution. Here we intioduce an approximate method to solve it. Eq. (2.17) tells us that
dV(r)/dr is very small and close to zero in the region near the cathode, while
R. 2e
kV {r) = J^—^\— V (r) is muclt greater than zero. For example, Jc is
generally of the order of 10^~^ A/m^ when a potential of several hundreds of kilovolts is
applied to a vacuum cylindrical diode with 3.34 cm gap [3]. Thus, in the region near the
cathode, we can neglect the second term on the left side of Eq. (2.15). It then becomes
r^^^^ = kV-"\r). (2.19)

dr
Actually, m the region near the cathode, using r « i?^ and substituting k back, we can
fiarther simplify Eq. (2.19) as

-1/2
(2e\
V-"\r), (2.20)
dr^ s. ymj
and Eq. (2.20) is valid in die region near tiie cathode. At the same time, V(Rc)^0,
dV(r)/dr\r=Rc=^Q m Eq. (2.16) and (2.17) are valid boundary conditions for Eq. (2.20).
Comparing Eq. (2.20) with Eq. (2.1), we can see that they share a similar form. This
indicates that we can use a similar procedure to solve Eq. (2.20) and get a similar result.
•^^=ÂF(^(^^-^^))" f°^ ^^>^- (2.21)

9Ar V m.
Comparing Eq. (2.21) with Eq. (2.5), we can see that they have a similar form. The
difference is that Ajc in Eq. (2.3) can be extended arbitiarily within the A-K gap because
Eq. (2.3) is valid for the whole diode region, while Ar in Eq. (2.21) is restricted to an
infinitesimal range from the cathode because Eq. (2.21) is valid only in the region near
the cathode.
To get an explicit relation like Eq. (2.4), we have to find a correlation between
V{R^ - Ar) and Ar. Suppose that we have two diodes sharing the same gap distance and
with infinite cathode emitting area. One is a planar diode and the other is a cylindrical
diode and if we can set up a relation between V{R^ - Ar) and V(/^), we can set up a
relation between V{R^ - A r ) andArforEq. (2.21).
In electiostatics, the electric field in a parallel-plate capacitor with mfinite area is
uniform, while the electrical field in a cylindrical capacitor with infinite length is
enhanced by its geometiy. Now we construct two capacitors, a parallel-plate capacitor
witii infinite area and a cylindrical capacitor witii infinite length, sharing the same gap
distance, D, and applying tiie same potential difference, Vo. We can tiien investigate the
potential distribution along the electric field lines. If we let the distance from the cathode
change at the same rate for both capacitors, which means at the same step length, we can
find a gain ratio, a, between the potential m the cylindrical capacitor and in the parallel-
plate capacitor, at the position sharing the same distance, A;c or Ar, from the cathodes.
Figiu-e 2.3 is helpful in imderstanding this relationship. This ratio is for a static electric
system, while the diode systems we address are filled with moving elections and even
with different bovmdary conditions, which refers to dV(r)/dr\r=Rc=Q or dV(r)/dr\x=o=^ in
the diode system. However, we can still use the gain ratio, a, for the two-diode system
because both ratios are from piu-e geometry field enhancements and should be close to
each other.
^.=Vo
Cathode
(a)
Figure 2.3. Static Field Enhancement Correction for (a) Rc>Ra and (b) Rc<Ra
10
Using tiie Rc>Ra geometry illustiated in Figtire 2.3 (a) and assuming tiie cathodes
are at zero potential, for tiie parallel-plate capacitor we get
F(Ax) = ^ A x (2.22)
and for the cylindrical diode we get the potential distiibution as
Considering that Ax and Ar change at the same rate, we can define Ajc=Ar for Eqs. (2.22)
and (2.23). Equation (2.23) divided by Eq. (2.22) gives a relationship between V(Rc-Ar)
and V(Ax),
^^^^-^^ = ^ H^^^). (2.24)

F(Ax) AxhxiRJRJ K
Because Eq. (2.21) is valid only at the region near the cathode, we are interested m the
ratio, a, in the region near the cathode. When Ax and Ar go to zero with the same step
length, Eq. (2.24) gives us the ratio, a, at the region near the cathode, i.e.,
a= lim ^ H^^^^) = . (2.25)

^=^rôAx\n(RJRJ R^ i?>(i?,//?J
Eq. (2.25) tells us that, in the region near the cathode, the V(Rc-Ax) in the cylindrical
geometiy can be represented with the V(Ax) in the planar geometry. That is
ViR^ - Ar) = aV(Ax). (2.26)
Using Eq. (2.26) to approximately cormect the relation in the diode systems and
substituting Eq. (2.26) into Eq. (2.21), we get
11
for R.>R„. (2.27)
9Ar V m
Comparing Eq. (2.27) witii Eq. (2.3) and (2.4), we get Jc for the cylindrical diode in terms
of J in the planar diode by letting Ax =Ar.
4s, lie V,3/2 1 A3/2

/ =
1/2 D 3 / 2
for /?„ > R„ (2.28)
m D"^R HRclRa)j
It is noted that the current density in Eq. (2.28) is approximately proportional to D~^
when the gap distance, D, is very small compared with the cathode radius. Re.
Matiiematically, ]n{RJRJ«D/R^ '^D/R^ in Eq. (2.28) when D = \R^-R^\ is very
small. That means that Eq. (2.28) will reduce to the Child-Langmuir law when the gap
distance is much smaller than the radii.
Now we can see that the cathode current density in the cylindrical diode can be
expressed with its own geometry and applied voltage. With a similar procedure as for
Rc>Ra, we can get the current density for Rc<Ra
3/2 f ,3/2
^ ^4s, \le V, 1 for R.<R„. (2.29)
9 \m D"'Rl" V HRalRc)
Based on charge conservation
V . J ( r ) = 0 ^ J(r) = J , ^ , (2.30)
we can get the expression for the current density, J(r), at any point in tiie cyUndrical
diode
3/2 >3/2
4^0 2e V, 1 (2.31)
J(r) for K>R^,
9 ^ m rd"'R:" .HRJRa)y
12
3/2
4£ IJP V^'^ f 1
for R,<R,. (2.32)
9 V m rd"'R;" y
MK/Rc),
With tiie static field approximation method, we get an approximation expression
of the voltage-current relation for the cylindrical diode. The current density in Eq. (2.28)
and (2.29) is the steady state space-charge limited current density.
2.4 Comparison of Langmuir-Blodgett Law with Our

Approximation Result
The Langmuir-Blodgett law gives the space-charge limited current, J(r), in a
cylindrical diode. We manipulate the Langmuir-Blodgett law [6] to give the current
density in a SI system
1£L per'^'(r)
•^W^ÂI—^-I^'
9 Vm r^/3'
(2-33)
where yf is a geometry correction factor. Langmuir and Blodgett used a very subtle
method to get a truncated power series for y9 and tabulated it in their paper. In Eq. (2.33),
it is assumed that the potential at the cathode is zero and that Eq. (2.33) is valid for the
whole diode region. As we know, any function can be expanded as a power series only
within a small range. A power series for fi is valid only in a certain range. Basically, Eq.
(2.33) is a ntunerical solution from a mathematical approximation method, truncated
power series ^pethod.
Our results in Eq. (2.28) and (2.29) are from a physical approximation. We set up
a potential relation near the cathode between the planar diode and cylindrical diode.
Because the current is limited by the zero potential point near the cathode due the space
13
charge and tiie current is determined only by the potential distribution near the cathode, it
is very reasonable to use the above relation to get a current solution.
Based on the above analysis, we cannot determine whether the Langmuir-
Blodgett law or our analytical expression is a better approximation for the real current
because tiie two results are from different approximation methods. It is for sure tiiat our
expressions have an explicit relation to the geometiy parameters and are helpful m
optimizmg tiie design of cylmdrical diodes. Even so, it is necessary to compare the
Langmuir-Blodgett law with our approximation result for some physical cases.
We use the following cases to investigate the similarity and the difference
between Eq. (2.33) and Eqs. (2.28)-(2.29).
Case 1: Catiiode Radius=0.132 m;
Anode Radius/Cathode Radius=0-1.0;
Applied Voltage=500 kV;
Catiiode Widtii=0.032 m.
Case 2: Cathode Radius=0.132 m;
Anode Radius/Cathode Radius=1.0-18.0;
Applied Voltage=500 kV;
Catiiode Widtii=0.032 m.
Assuming the emission is uniform, we will compare the currents and the current
percentage differences from the two different models. In these two cases, we fix the
cathode radius at 0.132 m and the applied voltage at 500 kV and vary the anode radius.
14
We do not consider if these configiu-ations are physically possible, that is, we compare
the results mathematically only.
Figure 2.4 shows the comparison between Eq. (2.33) and Eq. (2.28) for Case 1.
In this case, the cathode radius is larger than the anode radius. The anode radius varies
from zero to the cathode radius, 0.132 m. In Figtu-e 2.4, the horizontal axis represents the
gap distance divided by the cathode radius or normahzed gap distance; the left vertical
axis represents the space charge limited current in these configurations and the right
vertical axis represents the percent difference of these configurations. The currents from
the Langmuir-Blodgett law are nearly identical to our approximation results. In
particular, the percentage difference between the two results is less than 5% when the
normalized gap distance is less than 0.6, which is the design configuration of a usual
cylindrical diode.
15
1 • 1 • 1 • 1 • 16
10M
—+—Current (Langmuir-Blodgett law) / - 14
—X— Current (Analytical Expression) /
1M Current Percentage Difference / ^-. " 12
'• V / € •
, ^ - X
/ $ - 10
< : kV / c J
ent(
100k
8
1—
"
o 10k =•
6
^"x-x^ / •£ •
1
4
1
Ik
iiiiii 1 1 I
100 - / ^ — - r ' 1 . 1 . 1 . 1 .
0
0.0 0.2 0.4 0.6 0.8 1.0
Normalized Gap Distance (D/R )
Figure 2.4. Comparison between Langmuir-Blodgett Law and Analytical Expression for
Case 1: Cathode Radius=0.132 m, Rc>Ra, Voltage=500 kV
Figure 2.5 shows a comparison between Eq. (2.33) and Eq. (2.29) for Case 2. In
this case, the cathode radius is smaller than the anode radius. The anode radius varies
from the cathode radius, 0.132 m, to 20 times the cathode radius, 2.64 m. In Figure 2.5,
the axes represent tiie same quantities as in Figure 2.4. From Figure 2.5, the currents
from Langmuir-Blodgett law and our analytical result again match very closely. The
maximum percentage current density difference between the two results is about 7% for a
normalized gap distance less than 20.
Given the comparisons for the above two cases, it can be concluded tiiat our
expression for the space-charge lunited current density in the cylindrical diode is a
16
comparable analytical result to Langmuir-Blodgett law. The importance of this result is
that it is the first expression with a simple ftmction for the space-charge lunited current
density in a cylindrical geometry diode.
10M
O— Current (Langmuir-Blodgett Law)

+— Current (Analytical Expression)
— Current Percentage Difference <u _
o
c
100k I -2
o
0)
c O)
O
Q-
-6
•«>•$'4>,
5 10 15 20
N o r m a l i z e d G a p Distance (D/R )
Figure 2.5. Comparison between Langmuir-Blodgett Law and Analytical Expression for
Case 2: Cathode Radius=0.132 m, Rc<Ra, Voltage=500 kV
2 5 Relativity Corrected Space-Charge Limited Current for

the Cvlindrical Diode
When the applied voltage on the diode reaches a certain value, we have to make a
relativity cortection for the election mass. The space charge hmited current has to be
17
corrected by tiie relativistic factor. Based on energy conservation in the theory of
relativity, we have
mjt:^ -eV{r) = mroC^ -eV(RJ, (2.34)
where
2 / „ \ > -1/2
,J,-Z^ (2.35)
1-
v\R.) (2.36)
n^
and c is the speed of light in the vacuum. Because we define the cathode as the zero
potential point and the initial velocity of the election at the cathode is zero, Eq. (2.34) can
be simplified as
m]^^-eV(r) = mc\ (2.37)
Combining Eq. (2.35) and Eq.(2.37), for the geometiy illustiated in Figure 2.2 we get
1/2
v(r) = -c<l- l + ^ ( F ( r ) ) (2.38)

mc
Then Eq. (2.38) and Eq. (2.12) can be combmed to obtain
-2 -1/2
p(r) = -J^^\\- l + -^{V(r)) (2.39)

re mc
We can rewrite Poisson's equation after tiie relativistic correction as
-1 -1/2
1 dfdVjr)]^ R^l_ (2.40)
1 - l + _£^(K(r))
r dr\ dr J " SQ re mc
18
To get a solution for tiie current density from Eq. (2.40), we can use an approximation
metiiod similar to tiiat of section 3. Near the cathode region, we have
dVir) ^ ^
dr (2.41)
r^R
(2.42)
We use Eqs. (2.41) and (2.42) to simplify Eq. (2.40), keeping the relativistic term on tiie
right side of Eq. (2.40)
-111
d'Vjr) . 1 eV(r)
1 - .1+
dr c£n mc (2.43)
From Eq. (2.37), we have
1 + —V- = /
mc (2.44)
Substituting Eq. (2.44) mto Eq. (2.43), we get
d'V{r) J,
dr' ce.
{i-y-r'
(2.45)
Eq. (2.45) has the same form as the relativistic corrected Poisson's equation for the planar
geometry and the solution process is given by Jory and Trivelpiece [9, 10, 11].
dv
Intioducing the variable :v = —^, Eq. (2.45) reduces to
dr
dy ^ L (2.46)
1/2
dr
where
19
eJ.
K= (2.47)
Sf^mc
Integration of Eq. (2.46) yields
y' =2K{r'-\y"+C (2.48)
( e \dV
The constant of integration C is zero because y = dyldr = \ —r- — = 0 at the cathode
\mc J dr
where / = 1. Taking the square root and integrating from R^ to 7?^ + Ar , Eq. (2.48)
becomes
\{:^'-^'" dy = {1KY' \dr = {lKTAr. (2.49)
From standard integral tables, the integral over / may be expressed as
"\[:f^-xY'dy = F S^r,- -IE (2.50)

V ^ J
where
5,^ = cos"' (2.51)
and r^ is the relativistic factor correspondmg to F(Ar), F(^,V2/2)and E{d,4ll2)
are the first and second kind of elliptic mtegrals defined by
sin(f)
F{^,k)^ \ ii-Ai-k^x^)rdx, (2.52)
Ei^,k)= I (i-.vr(i-x^r^.
sm(^)
(2.53)
20
Combining Eqs. (2.47), (2.49) and (2.50), we get the expression for tiie current density.
f
J„ = s^mc (2.54)
2eAr^
Remember that Eq. (2.54) is valid only for the region near the cathode. That means we
caimot substitute gap distance, d, for Arand the anode voltage F(i?^) = V, for F(Ar)to
get the y in Eq. (2.44) directly. However, we can use a static electric field
approximation method to connect F(Ar) and Ar, similar to the method used in the non-
relativistic correction case. Actually, we can still use the ratio, a , in Eq. (2.25). Based
on the constant current along the election movement direction in the planar diode, we can
use Ar = d and V = aV, for / to get J^ in Eq. (2.54). That is.
-i2
J = s^mc s ^ -2E (2.55)

2ed' ^" 2
1^0^-ir
with
^0=1+—r (2.56)
mc
and
SQ =COS" (2.57)
Witii tiie relation of Eq. (2.30), we get
J(r) =
£„mc R^
s ^ •2E s.,^
' 4^]2rM-ir
+ — (2.58)
2ed^r i-irl-r
21
Altiiough tiie above correction procedure is for the geometiy illustiated in Figure
2.3(a), tiie corrected results, Eqs. (2.55)-(2.58), are valid for the geometiy illustrated in
Figure 2.3(b).
To compare the difference between the non-relativistic corrected and tiie
relativistic corrected space charge limited current, we compare the normalized
impedances [10] for a cylindrical diode with R,>R^. In a planar geometry diode, the
normalized impedance is defined as
R.or,n=U4-\ (2.59)
with
c/ = ^ (2.60)
mc
e =\rnc So J je
(2.61)
where V, is the applied potential for the planar diode. Because the current density in a
cylindrical diode is not constant any more, we change the definition of Eq. (2.61) to
( j(r)e^fd'r^
4 = (2.62)
ymc s, J V ^ ^ J
With these definitions, we get the normahzed unpedances for both the non-relativistic
corrected and the relativistic corrected cases as
r mc 2\"^ 3/2
R non-rel KHRJRa) (2.63)

4^f2
22
t j^\ r -^ / - xw.-i-2
rel
2eV
2 F So,f\.J.
-2E ^].2_rM^r_
S,,^ + - (2.64)
mc 2 i + (ro^-ir
The normalized impedances for the planar geometry are independent of the geometry,
while the normalized impedances for the cylindrical geometiy are still dependent on the
geometry because of the geometric correction factor, a , involved. This difference will
give a constant factor for the impedance once the geometry is specified, so it has no
influence on the normalized impedance comparison.
We use the geometry of a vircator at TTU with a cathode radius of 0.132 m and
anode radius of 0.098 m. Figure 2.6 shows the normalized impedance comparison
between Eq. (2.63) and Eq. (2.64), both of which are for the geometry withi?^ >Ra-
From Figure 2.6, we can see that the relativistic correction does not matter when
/ < 5 - 6. This means we can still use the non-relativistic relations, Eq. (2.28) and Eq.
(2.29), to optimize the parameters when we design the cylindrical diode. Actually, the
space charge limited current density and the respective impedance value should be
corrected by the relativistic factor when /, is over 2, which is shown in the planar diode.
Because tiie cylmdrical geometiy field enhancement, /, in Eq. (2.56) is greater than
what is expected without the geometry correction, we should consider the relativistic
correction effect at a smaller applied voltage in the cylindrical geometiy, /o^'^' than in
the planar geometry.
In Figure 2.6, we can see that the normalized impedance for the coaxial diode at
TTU is about 2 when the normahzed applied voltage is 1, which is about 500 kV.
23
• • I I I I I I 1 1 1—1—1—1—n-| 1 1—1—1—1
-^ h On-KeiailVISuL UOrrytLbQ INUMiiaii.ccu luifjcuam^c

Relativistic Corrected Normalized Impedance
<D •
O '^•'
C
CO
•g 1 ; - ; !-;--;--:-: . _. _ i... _. . 1. . ^Ni^^"*v^
£ 1
' ' ' ^^'^^K... • "
^*Ss.^
' • - - — - - ,
'r \--~\~-\ -; ; ; ;
(D
N ^^'-''v.; 1 :
"co
E
i_
o
. . . . . . .
0.1
10 100
Normalized Applied Voltage (U=eV /(mc ))
Figure 2.6. Non-relativistic and Relativistic Normalized Impedance Comparison
2.6 Space-Charge Limited Current for 2-D Coaxial Diodes
In reality, we always use a finite cathode ratiier tiian an infinite cathode. This
makes it necessary to extend our current density to two dimensions. We have noticed
tiiat Y. Y. Lau [12] used a simple theory to extend tiie Child-Langmuir law to 2-D for the
planar diode. Here we tiy to use tiie same idea to extend our 1-D result to 2-D. Before
tills fiirther derivation, let us analyze the difference m tiie zero net E-field point forming
mechanism between the planar diode and the cylindrical diode.
In a planar diode witii infinite electiode area, botii tiie current and current density
are conserved and the elections are accelerated from tiie cathode to tiie anode. The
24
election density is constant from tiie cathode to the anode. These moving elections and
their image charges produce a maximum E-field at the cathode, in a direction opposite to
the applied E-field and this E-field will cancel tiie applied static field to get a zero net E-
field at the cathode when the moving election number reaches a certain size. The E-field
from the image charges just enhances or reduces and does not change the du-ection of the
E-field from the moving space charges. In the cylindrical diode, the situation is a little bit
different.
In a cylindrical diode filling with moving elections, the current is conserved and
the current density is not. The movmg elections between two concentric cylindrical
conductors wdll have positive image charges inside the inner conductor. The number of
these positive image charges is less than that of the moving elections. The E-field at one
point due to space charge is determined only by the charges surrounded by a cylindrical
surface with this pomt on it and has nothing to do with those charges outside of this
surface. In the R^ > R^ case, the applied static E-field is a divergent field and the
maximum convergent E-field due to the space charge, including the moving elections and
image charges is at the outer conductor. The image charges do not change the E-field
direction due to their origmal sources. At this point, it is similar to a planar diode.
However, in tiie R^ < R^ case, the applied static E-field is a convergent field and the
maximum divergent E-field due to the space charge, only from tiie image charges, is at
the inner conductor. That is, tiie positive image charges of the moving elections limit tiie
current.
25
We consider a cylindrical diode with a cathode strip with a finite width, W.
Elections are emitted from the cathode ship uniformly. We assume W » R^. The
charge density, p{r,9,z) , is assumed to be independent of z and 0 . We have
J ( r ) * r = p(r)v(r)r = J^R^ and we can get p{r) = J^R^ l[rv{r)). Because the image
charge density is proportional to its sources, we can define the image charge density
yo'(r) as J^R^ l{rX{r)), where X{r)'\s a fimction of r , reflecting the image effect and
the velocity of the moving elections. Now we are ready for our derivation.
We define the center line of the cathode as z-Q. Let us focus on the E-field at
the point of (r = i?^,^ = 0,z = 0)in the cylindrical coordinates due only to the space
charge. The E-field due to the space charge m the current limit will cancel the static E-
field due to the external applied potential at (r = i?^, ^ = 0, z = 0). The E-field from the
space charge can be expressed as
^ = PiryAG^^r ^^.65)
4;rf0Pc - r cos 6')^ + (rsm ^)^ + z^)
On the center line, z=0, we just consider the r component of the electric field by
symmetry and this component is obtained by multiplying Eq. (2.65) by the directional
cosine, (i?^ -rcos(9)/^(/?^ -rcos<9)^ +(rsm(9)^ +z^ . Summing over tiie moving
elections within the gap, the E-field on tiie center line, z=0, of the emitting ship is given
by
F = j ^ A ^ L / f , iK-rcos9)dz (,,,^
(4ns,v{r) { J.^^[(R^ - r C O S ^ ) ' + (rsin^)' + Z ' /
26
Summing over the image charges inside of tiie inner conductor, tiie E-field on the center
line of the emitting strip is given by
^... ] J S M L - 1 , , 7 ^ iR.-rcosO)dz ^
i47rs,X(r)i in[{^R^-rcoêy+{rsmef+z')"
For tiie case of R,>R^, the E-field at (r = i?^,^ = 0,z = 0)due to the space
charge, E = E'-E", can be expressed as
E = M^ f ^^ Y j ^ T (R,-r cos 0)dz
47rs,iM(r)l Jn((R,-rcos^f+irsm0y+z'f''
where M{r) is a segmental function defined by
, . fv(r) R<r<R^
M(r) = \ for a c . ^ .
l-Z(r) 0<r</?„ ^ ^
The integration over z in Eq. (2.66) and can be performed to yield
. , 2(^.-.cos4l-^(<^--"°^^^'"'°">')l
47rsQ
I^^R^ r-^ i ^^^1 -• (2.70)
JM(r)J p,-rcos^)'+(rsin^)')
In Eq. (2.70), we have used the first order approximation
4((J?,-rcos^)^+(rsm^)^) ^ V" ^ 2{iR^ -rcos0)'+irsin0)')
w' J ~ W
In Eq. (2.70), considering the 1-D case with the infinite emitter width and 2-D case with
the finite emitter width, the E-field in Eq. (2.70) in both cases will cancel tiie external E-
field due to the applied potential, where the space-charge limited current reaches tiie
27
maximum pomt. We assume that botii v(r) and X(r) are the same for the 1-D and 2-D
cases using tiie same argument as for approxunating p(z,r,0) by p(r). Thus we have
^ _ 2((/?^-rcos^)'+(rsin^)^)'|
0 .
2. 2iR^-rcos0)
-^r^nR. f dr W
E= \d0 ,—^ ^, ^ C2 72)
4^ô i^(f)i ((R,-rcos0)'+irsm0fj ' ^' ^
and
0 J 2a-
^ c m ^K]
. r dr
rfr y ^ 2(R^-rcos0)
47rSo (Mir) I [(R^-rcos0f-i-irsinGf)' ^'
The E-field m Eq. (2.72) and Eq. (2.73) will cancel the applied static E-field at the pomt
of (r = i?,,^ = 0,z = 0), whichisFo /{R^ \n(R^ /Rj).
Taking the ratio of Eq. (2.70) and Eq. (2.71), we get
J.20 ,. (^(^)l W-
= 1 +1 \. • (2-74)
JcXD \ J L J
r dr i,^ L 0 ^K^c-r-o.a,
2{R^-rcos0)
i M{r) { [{R^ -rcos0f+ (r sin 0)')
From the integral over 0 in Eq. (2.74), we get
".AiR,-rcose)Jj^ (2.75)
W W
and
y 2(R^-rcos0) 4. ^^.76)
J {(i?,-rcos^)'+(rsin^)') R,
Thus we substitiite Eq. (2.75) and Eq. (2.76) mto Eq. (2.74) to get
28
4 ^ = 1+^ . (2.77)
c df
Fortunately we can cancel the integral of in the denominator and numerator of
iM{r)
Eq. (2.74) because we get expressions independent of r in Eqs. (2.75) and (2.76).
Otherwise we could not get Eq. (2.77) because we do not know the velocity profile, the
potential profile and the image relationship in the cylindrical diode.
For the case of R^ < R^, we can follow the above procedure with
Too R<r<R^
M{r) = \ for " \ (2.78)
Then we find tiiat Eq. (2.77) is valid for tiie case of R,<Ra-
29
CHAPTER 3
APPLICATION OF ANALYTICAL CURRENT
EXPRESSION IN DIODE DESIGN
3.1 Current Evaluation of Coaxial Diodes
Since our analytical 1-D current expression is comparable with the Langmuir-
Blodgett law, the current evaluation results from these two expressions are close to each
other. On this level, our analytical expression is of little use except for its beautiful
appearance. Actually, it is more complicated than this. An approximate analytical
expression gives us an explicit relation with physical parameters. To some degree, it is a
correction to Langmuir-Blodgett law, although we have no evidence to show which one
is closer to the real current density.
The 2-D current density is based on the assumption that W»Rc. Y. Y. Lau
claims that his 2-D result for planar diodes works for W = D although it is derived with
an assumption of W»D [12]. Perhaps the 2-D result for coaxial diodes in this work can
be extended to W = R^.
Figure 3.1 shows the prediction for the current from the TTU typical coaxial
vu-cator geometiy from different mathematical models. We can see that the difference
due to the different models is not very large. Generally, the 2-D result is larger than its
corresponding 1-D result.
30
60
1-D Langmuir_BlojJgett Law, Eq. (2.3-1)
2-D Langmuir_Blo Jgett taw, Eq. (2.3;i) and Eq. (2.77)
50
1-D, Eq. (2.31)
-1- 2-D, Eq. (2.31) anip Eq. (2.77)
40 -•- 1-D Relativistic, (2.58)
•£ 30
O 20
10
200 400 600 800 1000

Applied Voltage (kV)
Figure 3.1. Current Predictions from Different Mathematical Models
3.2 Virtual Cathode Position Determination and Virtual

Cathode Oscillation Frequency
Traditionally, a coaxial vircator has two frequencies, tiie election reflection
frequency and the virtual cathode oscillation frequency [2].
The election reflection frequency is determined by tiie election drift distance and
election drift velocity. In a planar diode, tiie vutual cathode forms as a mirror image of
the cathode on the otiier side of tiie anode. However, in a coaxial diode, tiie virtiial
catiiode position can not be determined by the this symmetiy because the current density
is not conserved in a coaxial diode. Witii tiie analytical current density in Eq. (2.28) and
31
(2.29), we can determine the vktual cathode position by setting an equal current for two
back-to-back coaxial diodes, sharing the same mesh anode. For the coaxial geometry at
TTU, we can the virtual cathode position in Table 3.1.
Table 3.1. Vutual Cathode Position Prediction from Different Models

Cathode Radius Anode Radius Virtual Cathode Radius* Virtual Cathode Radius**
0.132 m 0.098 m 0.067 m 0.064 m
* Based on our analytical result.
** Based on simple symmetry.
The virtual cathode positions in Table 3.1 refer to a steady-state virtual cathode,
neglecting the virtual cathode oscillation. Once the vutual cathode position is determined,
we can reevaluate the election reflection frequency. Assuming that the election is
emitted with zero initial velocity and that the election moves at an average velocity of a
half of peak relativistic corrected velocity, we have the following frequency for 500 kV
1/2
v(i?J = c l - 1 . ^ (3.1)
mc
v(i?J (3.2)
Anotiier important equation for the election reflection frequency is [13],
r \
flcosh(Kô))- (3.3)
frreflexl
27r{R,-RJ^
Figure 3.2 is the election reflection frequency from tiie above different models. We can
see tiiat tiie difference in pretty large. Originally, Eq. (3.3) is for a planar diode. It is
predictable tiiat the reflection frequency bandwidth is not small smce the virtiial catiiode
position itself oscillates over the time.
32
N 2.4G
c
5. 2.0G
o
0)
Qi
—•— From the model for planar diodes
—^— From the new model for cylindrical diodes
2 1.2G
(D
^4-y^-^-^-^-ii-^-^-^-^
LU )^)K-5I^->I«-
-)K->I«-^-^-
800.0M
300 400 500 600 700 800 900
Figure 3.2. Election Reflection Frequency Evaluation for TTU Coaxial Vhcator.
Rc=0.132m,Ra=0.098m
We are concerned with the virtual cathode frequency more than the election
reflection frequency because we believe that the virtiial cathode oscillation contiibutes
more to tiie microwave generation than election reflection. The virtiial cathode
oscillation frequency depends on a uniform plasma oscillation frequency, which is a
fimction of tiie particle density. In a diode system, the particle density varies witii tiie
potential distiibution because of current conservation. Generally, the particle density in
tiie anode region is used to estimate the plasma frequency. Disregarding the theoretical
correctiiess, we follow the usual procedure to compare the election frequency and tiie
virtual cathode frequency for the coaxial diode at TTU.
33
Based on tiie analysis in Chapter 2, we can determine tiie charge density, p, and
tiie particle density, n, in tiie region near tiie anode
^(^»)-^^ (3.4)
«(-J = ^ (3.5)
where J(Ra) and v(Ra) are from Eq. (2.31) and Eq. (3.1), respectively. Then we determine
tiie virtual cathode oscillation frequency as [14]
/v,«.../_c.<*o.e =(1-924-2.309) */p„.„„ (3.7)
Figure 3.3 is the comparison between the virtual cathode oscillation frequencies and the
election reflection frequency from Eq. (3.3). We can see that the election reflection
frequency falls within the range of the relativistic corrected virtual cathode oscillation
frequencies. I would like to point out that this reflection frequency is from the planar
diode approximation model. It cannot be correct to use the anode particle density in Eq.
(3.6).
Here I have some comments on the relations between physical phenomena and
physical models. We know that any physical model is a description of a physical
phenomenon and generally is not a complete description. A good physical model is a
good description for a phenomenon within certain physical limitations. This is why we
get different results with different physical models in this chapter. Any physical model
just provides some guidelines for engineering designs.
34
2.8x10
Blectron Reflection Frequency

+ Minimum Virtual Cathode Oscillation Frequerlcy
- O • Maximum Virtual Cathcjde Oscillation Frequeficy
8.0x10°
200 400 600 800
Figure 3.3. Comparison of the Vircator Frequency with the Election Reflection
Frequency
35
CHAPTER 4
EXPERIMENTAL CONFIGURATIONS AND VIRTUAL
CAVITY
Figure 4.1 is an overall experimental configuration, including a Marx generator, a
pulse forming line (PFL), a vacuum system, a waveguide and anechoic chamber. Our
experimental goal is to obtain data with which to analyze the frequency determination
mechanisms. For the experimental configuration illustiated in Figure 4.2 without any
reflector, tiaditionally it is accepted that microwave frequencies are determined by the E-
beam property because no enclosed cavity forms in this type of geometry. Some
previous data hint that this statement does not always hold. To investigate the frequency
determination mechanism in-depth, we designed a group of experiments to get a
complete set of data.
The coaxial vircator has been investigated as a potential high power microwave
source because of its advantages over the planar vircator. The primary goal is to improve
its power efficiency, for which theoretical research on the coaxial vircator is important.
Although some efforts have been made [2], the theory of coaxial vircators has not been
studied as extensively as for planar vircators.
Cavity resonance is a basic technology for microwave generation. It can be used
to narrow the frequency band and to increase the power output concurrently. J. Benford
mvestigated the resonance effect in a planar vircator [7]. Because the coaxial vircator has
geometry different from the planar vkcator, our experimental data on a coaxial vircator at
36
TTU shows a different resonance behavior from an planar vircator. This research pomts
us in tiie right direction to improve its power efficiency. Generally, the vutual cathode
oscillator working frequencies are considered to be determined by tiie election reflectmg
frequency andtiievirtiial catiiode oscillation frequency. Thesefrequenciesdepend on tiie
diode gap distance and tiie appUed diode voltage and current [2, 14]. For our
experimental vacuum diode with a 500 kV applied pulse, tiie tiieoretical prediction
frequencies are: vutual catiiode oscillation, 1.8-2.5GHz, election reflecting frequency,
1.80GHz. Actually,tiiesepredictions offer only some design guidance; they are different
from the experimental data because different approximate models ignore different
physical effects. We always tiy to design tiie diode to make tiiese two frequencies
coincident to get maximum microwave power output [1].
Marx Tank
Vacuum System
Waveguide
Oil Berm
Lead Shielding.
Anechoic Chamber.
Figure 4.1. Overall Sketch of the Experimental System
37
4.1 Experimental Configurations
The coaxial vircator at Texas Tech University is powered by a Marx bank with a
pulse forming line pulsed power system, witii a 70 ns pulse lengtii and a -500 kV and 40
kA range. Figure 4.1 is tiie system sketch. The core experimental configuration
including a coaxial diode, a circular waveguide, two diagnostic E-probes and a
microwave absorber is shown in Figure 4.2. The schematic of tiie E-probe is illustiated
in Figure 4.3.
Cathode > Microwave

ve <-'^
E Probe 1 > Absorbe
/: -^ g
Mesh Anode î^ ^
D, D i! Reflectors 0
E Probe 2
.}a.
Circular Waveguide e >
:AAAA/"SAA/\?
Velvet Emitter
Figure 4.2. Experimental Configuration
The core of the experimental configuration is the coaxial diode. The stainless
catiiode is connected to the output of the pulse generator and the stainless mesh anode is
grounded to tiie aluminum waveguide. The catiiode has a velvet ship with a widtii of 3.5
38
cm working as an explosive emitter. Most of the emitted elections from the emitter can
pass through the mesh anode to enter the inner anode region, where the virtual cathode
forms and oscillates. The cathode radius is 13.1 cm and the anode radius is 9.9 cm. The
waveguide radius is 9.8 cm, close to that of the anode. The 2.5 m waveguide is
connected to the mesh anode at one end and to the microwave absorber at the other end.
Two E-probes are installed in the waveguide wall, 1.5 mfromthe anode end, 90° apart on
the azimutii, so that one is referred to at the horizontal probe and the other at the vertical
probe.
SMA Bulkhead
O-ring
Screws
Probe Mourit
Waveguide
Insulator
Inner Conductor
Figure 4.3. E-probe Schematics
To investigate the resonance effect in the cylindrical diode, we changed tiie
emitter position, Di, to 4 cm, 5.08 cm, 6 cm and 7 cmfromthe cathode back wall. For a
certain emitter position, we used three different reflector arrangements near tiie
eguide. no reflector (1), a stiip reflector (2), or a donut reflector (3) as shown in
way
39
Figure 4.2. The reflectors are made of copper plate. The stiip reflector has a width of
2.56 cm and tiie donut reflector has an inner radius of 8 cm and outer radius of 9.8 cm.
The reflector position relative to tiie cathode emitter, D2, is also changed. For any fixed
geometry sharing the same emitter width, emitter position, reflector type and reflector
position, we vary the pulse voltage over a range. For a different velvet emitter width of 1
cm, we changed Di, D2 and the appHed pulse voltage and got another set of data.
Two high-frequency cables carry the two E-probe signals to oscilloscopes in a
screen room. The cables are calibrated by a TDR and a HP8719C network analyzer. We
use the TDR to make sure that the electrical lengths of the two cables are almost the same,
within 20 ps. The network analyzer is used to calibrate the attenuation of the two cables
to ensure that they are close to each other. We use two oscilloscopes to sample the E-
probe signals, an Agilent Infiniium oscilloscope with 2.25 GHz bandwidth, 8 GS/s
sampling rate and a Tektionix TDS6604 with 6 GHz bandwidth and a 20 GS/s sampling
rate. In fact, this configuration was also used to study the power efficiency in earher
investigations. What we do now is to use faster oscilloscopes to monitor the raw signals
from the E-probes. We obtain the frequency spectiiun by applymg a fast Fourier
tiansformation (FFT) to the raw data recorded by the fast oscilloscopes.
In addition, we have a capacitive probe, a B-dot probe and a Rogowski coil to
monitor the diode voltage and diode current, respectively. Since we are most concemed
with the microwave frequency, we obtain these diode voltage and current signals just for
reference and only normalized values are recorded.
40
4.2 Experimental Data Analysis
We tiy to get the frequency and propagating mode information from the
experimental data. Then we can determine tiie frequency and mode dependence on the
geometry, applied voltage and current.
1.2
H E-Probe
-0.6
Z)
<,
d.
E
<
-0.6
20 40 60 100
Time (ns)
Figure 4.4. Waveforms Directly Obtained by the E-probes
The first step to analyze our data is to apply a fast Fouriertiansformation(FFT) to
tiie horizontal and vertical field signals. We can get thefrequencyinformation for both
horizontal and vertical fields. Figure 4.4 shows typical horizontal and vertical field
41
signals for tiie same shot. Figure 4.5 shows frequency distiibutions from Figure 4.4.
From Figure 4.5, we can seetiiattiiepeakfrequenciesof the horizontal and vertical fields
are of a little difference, however, their bandwidtiis overiap considerably. Theoretically
we can considertiiattiiefrequenciesoftiiehorizontal and vertical fields for the same shot
coincide.
0.03
H E-Probe 3.165 GHz, 2.24
0.02
0.01
<
® 0.00 iMltiyi«JH|i«ai i*irmJ»aWyiii«i6»*ii^ ^^Ml^^^Mihtî
•Q.
I 0.03
0.02
0.01
0.00
Frequency (GHz)
Figure 4.5. Fast-Fourier Transformation of the Waveforms Shown in Figure 4.4
42
4.3 Experimental Results
We finished a complete group of experiments. Based on the analysis method
above, what we can get from the data analysis is tiie horizontal and vertical fields. We
intioduce some attempts to get the propagating mode. These attempts are helpfiil m
understanding the microwave generation. We still cannot determine the propagating
mode because of some practical problems.
Table 4.1 is typical frequency information for some shots. Generally the first
column, shot number, represents the charging voltage of two back-to-back capacitors in
the Marx bank and has some relation to the diode voltage. Columns two to five are FFT
results for the raw horizontal and vertical fields, where the peak frequencies and their
corresponding amplitudes are shown. From this table, we can see that the horizontal and
vertical fields share similar peak frequencies. We will consider that peak frequencies
which are close are the same for the horizontal and vertical fields, for example, 3.06 GHz
and 3.09 GHz can be considered as any frequency from 3.06 to 3.09 GHz. Sometimes,
there are two or more peak frequencies shown simultaneously, for example Shot 60 and
Shot 64 in Table 4.1. In this case, we consider their relative powers to determine the
dominant peak frequency and the secondary peak frequency. Column six and seven
represent the measured diode voltage and current. The voltage is measured by a Pearson
coil and the current is measured by a B-dot probe. Since these probe performances are
not as good as expected, we just list their original recorded value here for reference.
43
Table 4.1. Examples of Frequency Record in Cavity Effect Research of a Coaxial
Vircator at TTU
Shot No. H E-Probe Peaks V E-Probe Peaks Diode V-I
Freq. (GHz) Amp. (A.U.) Freq. (GHz) Amp. (A.U.) V. (A.U.) I. (A.U.)
Shot 60 2.92 19.0 2.92 21.5 2.56 1.32
Shot 66 2.90 15.8 2.91 18.0 2.57 1.24
Shot 76 3.16 18.0 3.19 12.8 2.82 1.89
Shot 80 3.25 21.3 3.16 17.0 2.95 1.96
Shot 62 3.08 3.17 3.07 6.60 2.56 1.26
Shot 80 3.06 3.25 3.09 3.47 2.52 1.51
Shot 60 1.93/2.89 2.10/2.03 1.92/2.89 9.95/4.48 2.5 1.29
Shot 64 2.0/2.91 4.8/1.76 1.96/2.91 9.5/2.3 2.60 1.38
Shot 80 2.05 8.03 2.07 15.6 2.91 1.71
Table 4.2 is a typical record for a certain diode geometry. For all these shots, the
diode physical structures are the same. We vary the charging voltage, resulting in a diode
voltage and current change. From Table 4.2, we can see that the microwave frequency
increases slightly with the diode voltage or current increasmg, which is predicted by the
frequency dependence on the plasma density. For our research on the cavity effect, we
will consider these frequencies as a smgle frequency for simplicity. For the geometiy in
Table 4.2, we consider the microwave frequency as approxunately 3.16 GHz. Table 4.2
shows good consistence with the tiieoretical prediction. Not all experimental data looks
as neat as Table 4.2. In some diode geometiies, with a diode voltage or current change,
tiie microwave frequency will "jump" from one band to another band or tiie dominant
peak frequency shifts from tiiis one to another one. Table 4.3 is an example. In Table 4.3,
we can see tiiat the dominant peak frequency is about 2.0 GHz when the diode voltage or
current is lower and it shifts to about 3.10 GHz with higher diode voltages or currents.
Generally, we will choose one dominant peak frequency depending on the reflector type
44
in a geometry for future research. For example, we will choose 3.10 GHz as the
dominant peak frequency for Table 4.3 since this geometiy has a donut type reflector.
The reason is that most of the geometiies with a donut reflector have a 3.10 GHz
dominant peak frequency.
Table 4.2. Typical Experimental Records for a Certain Diode Geometry

Shot 60 3.06 3.74 3.06 12.8 2.55 1.27
Shot 62 3.08 3.17 3.07 6.60 2.56 1.26
Shot 64 3.09 2.47 3.09 5.95 2.47 1.36
Shot 66 3.07 2.50 3.06 8.75 2.52 1.30
Shot 68 3.14 2.33 3.14 7.30 2.58 1.48
Shot 70 3.16 4.49 3.17 6.81 2.62 1.53
Shot 72 3.16 4.19 3.16 6.04 2.61 1.75
Shot 74 3.18 4.53 3.17 6.88 2.66 1.61
Shot 76 3.16 5.28 3.17 4.36 2.67 1.95
Shot 78 3.16 8.12 3.17 5.06 2.75 1.67
Table 4.3. An Example Experimental Records for a Certam Diode Geometry

Shot 60 1.90/3.0 3.27/1.20 1.89/3.17 3.35/3.09 2.36 1.27
Shot 64 1.99/3.15 10.9/1.1 1.99/3.16 7.0/2.32 2.56 1.28
Shot 68 2.02/3.16 1.66/2.04 2.01/3.16 1.92/3.48 2.62 1.35
Shot 70 3.08 1.71 3.08 2.91 2.32 1.56
Shot 74 3.18 3.17 3.19 5.37 2.40 1.72
Shot 78 2.02/3.19 1.48/3.34 2.01/3.16 1.47/3.27 2.57 1.72
Shot 80 3.0 1.71 3.09 5.71 2.39 1.70
Shot 82 2.02/3.19 1.1/1.98 3.19 4.2 2.62 1.88
Based on tiie explanation above, we are ready to list one dominant peak frequency
for every coaxial diode stinctiire. We would like to list the frequency for those diode
45
stiiictiu-es m tiu-ee tables, tiie fu-st for stiiictiu-es without any reflector, the second for
stiaictiires with a donut reflector and the third for stinctiires with a stiip reflector.
4.4 Cavity Resonance Effect and Virtiial Cavity
Table 4.4 is a summary list for the diode stiictiire without reflectors. From Table
4.4, we can see that tiie dominant peak frequency decreases as tiie velvet cathode is
moved farther from tiie back wall of tiie catiiode. These results caimot be explamed by
conventional cavitytiieorybecause tiie geometiy illustiated ui Figure 4.2 doesn't form an
enclosed structure. Even though some parts of tiie diode structure form partial enclosed
structures, for different D], those structures are fixed and should not cause the output
microwave frequency to change. That is, the emitter positions are responsible for the
frequency change.
Table 4.4. Frequency List for Geometries without Reflectors

Cathode Width (cm) DI (cm) Dominant Peak (GHz) Secondary Peak (GHz)
3.5 4 3.17 1.95
3.5 5.08 3.16 _*
3.5 6 2.70 2.0
3.5 7 2.68 1.95
1 4 3.0 2.10
1 5 3.0 1.85
1 6 2.93 1.82
1 7 2.60 1.90
*: "-" means that there is no relativefrequencythere.
One possible explanation is that a cavity is formed by the collected elections or E-
beam, the back wall of the cathode and the mesh anode. This is reasonable since the E-
beam itself acts as a conductor. A similar cavity illustiated in Figure 4.6 forms between
46
tiie earth surface and tiie ionosphere, which is called the Schumann resonance [15]. We
calculate tiie resonancefrequencyfor a cylindrical cavity with a radius of 9.9 cm and find
TMoH=3.17 GHz witii a lengtii of 5.08 cm and TMon=2.76 GHz with a lengtii of 6 cm.
For Di=4 cm or 7 cm, we get a propagating mode, with a frequency close to the
experimental frequency.
Cathode
Mesh/^node
Cavity E-beam
Velvet Emitter
Figure 4.6. Schumann Cavity around the Earth and in the Diode Region
Another possible explanation is that a maximum or minimum value of certain
fields, E field or B field, is determined by the E-beam. This field forms a mandatory
boimdary condition, which constitutes a non-conventional cavity with other physical
boundaries. It is harder to analyze, because we cannot determine how the mandatory
boundary forms. We consider the region mentioned in thefirstpossibility as a half cavity
and calculate the cavity resonance frequency. We find that. TMn2=3.05 GHz,
TE3i2=3.16 GHz with a length of 6.30 cm; TMi,2=2.63 GHz, TE3i2=2.76 GHz witii a
length of 8.25 cm. The length of 6.30 cm represents the distance between the back wall
of tiie cathode and the middle point of tiie common region occupied by the velvet emitter
47
witii Di=4 cm and 5 cm. The length of 8.25 cm represents the same point position
occupied by the emitter with Di=6 cm and 7 cm.
The above possibilities are only some guesses about the cavity in the diode region.
We can call this kind of cavity a virtual cavity since it has no obvious physical boimdary.
It is too early to comment on that because, so far, we cannot determine the propagating
mode exactly. Although we have some evidence showing the possible TMoi and TMn
modes based on the directivity measurement at the open end of the waveguide in another
set of experiments, we still need more solid evidence to clarify them, which is a goal for
our future research.
4.5 The Function of the Reflectors
Table 4.5 is a frequency list for the donut reflector at different positions. From
Table 4.5, we can see that the donut reflector has little influence on the microwave
frequency, which means that the donut reflector reflection is not stiong enough to change
the cavity. We do observe tiiat the microwave power generally is decreased by the donut
reflector.
48
Table 4.5. Frequency List for Geometiies with a Donut Reflector
Cathode Width (cm) DI (cm) D2 (cm) Dominant Peak (GHz) Secondary Peak (GHz)
3.5 4 3.08 3.14 2.07
3.5 4 4.08 3.20 2.0
3.5 4 5.08 3.16 2.0
3.5 4 6.08 3.13 2.0
1 4 3 2.90 1.40
1 4 4 2.90 1.36
1 4 5 2.90 2.07
1 4 6 2.90 2.69
3.5 5.08 3 3.16

3.5 5.08 4 3.16 -
3.5 5.08 5 3.10 2.0
3.5 5.08 6 3.16 2.0
1 5 3 3.0 1.85
1 5 4 3.0 1.82
1 5 5 2.95 -
1 5 6 3.0 -
3.5 6 3.08 2.70 -

3.5 6 4.08 2.68 2.05
3.5 6 5.08 2.70 2.0
3.5 6 6.08 2.70 2.0
1 6 3 3.60 2.70
1 6 4 2.95 3.66
1 6 5 2.90 -
1 6 6 2.80 -
Table 4.6 is a frequency list for the stiip reflector at different positions. From
Table 4.6, we can see that the stiip reflector changes the microwave dominant frequency
for 3.5 cm velvet catiiode from 3.16 GHz in geometiies without reflectors to 2.0 GHz.
However, the dominant microwavefrequencyfor 3.5 cm velvet catiiode witii Di=6 cm or
7 cm has not been changed by a stiip reflector. This phenomenon is not expected. A
49
similar odd behavior shows up m other structures. It is not such a surprise because the
physical mechanism offrequencydetermination intiiiscoaxial vircator is not understood.
We do observe that the microwave power generally is increased by the strip reflector.
Table 4.6. Frequency List for Geometries with a Strip Reflector

Cathode Width (cm) DI (cm) D2 (cm) Dominant Peak (GHz) Secondary Peak (GHz)
3.5 4 3.08 2.0 3.16
3.5 4 4.08 2.0 3.95
3.5 4 5.08 1.91 3.80
3.5 4 6.08 1.92 3.80
1 4 3 2.70 2.95
1 4 4 1.64 3.05
1 4 5 1.64 3.10
1 4 6 2.95 1.60
3.5 5.08 3 2.0 2.90

3.5 5.08 4 1.95 3.90
3.5 5.08 5 1.93 3.86
3.5 5.08 6 1.95 3.90
1 5 3 1.90 3.0
1 5 4 1.90 3.0
1 5 5 1.90 3.80
1 5 6 1.84 3.70
3.5 6 2.08 2.70 2.0

3.5 6 3.08 2.70 2.0
3.5 6 4.08 2.75 2.0
3.5 6 5.08 2.70 2.0
3.5 6 6.08 2.70 2.0
1 6 3 2.85 -
1 6 4 1.90 3.80
1 6 5 3.68 1.82
1 6 6 1.80 3.60
3.5 7 3.08 2.69 -

3.5 7 4.08 2.69 -
50
We notice tiiat the microwave frequencies for a 1 cm velvet cathode are slightly
lower than those for a 3.5 cm velvet cathode, with a sinular diode structure. A reason for
this is that we measure the cathode position from its middle point to the back wall of the
catiiode. Altiiough the Di is of tiie same value botii for the 1 cm and 3.5 cm velvet
cathode, the actual E-beam edge position for the 1 cm velvet cathode is different from
that for the 3.5 cm cathode. This difference leads to different cavity sizes. This different
frequency proves that the cavity formation in the diode region depends on the E-beam
positions. Another phenomenon is that the odd frequencies out of our expectation for the
1 cm velvet cathode occur more often than for the 3.5 cm velvet cathode. One possible
reason is that the E-beam has a smaller width for the 1 cm velvet cathode. The reflection
effect from a narrower E-beam as a conductor is not as stiong as from an E-beam with a
greater thickness. Or the mandatory boundary field stiength difference has a different
microwave reflection effect.
Reviewing the frequency Usts in Table 4.4, 4.5 and 4.6, we can find that different
microwave frequencies exist simultaneously for the same diode structure without
reflectors, with a donut reflector and with a stiip reflector. Different reflector sets just
change the microwave power or shift a dominant peak frequency to the secondary peak
frequency. That is, different reflector sets enhance one microwave frequency and
suppress another one. This is a general function for any microwave reflector, but it is not
the only function.
51
Another important phenomenon we observe is that the microwave frequencies are
almost independent of the reflector position. The reflector at different positions leads to
different naicrowave reflection intensities. This means that the microwave reflection by
tiie reflector is not as stiong as those by tiie E-beam or the mandatory boundary. The
microwave frequency is still determined by tiie vutual cavity mentioned above. This
phenomenon was not expected. From another point, it proves that the virtual cavity
exists in the diode region. The following analysis will answer this question.
The different performances of the donut and strip reflector have to be explained.
We may be able to explain the different effects of the donut and the strip reflector by
their different microwave reflection intensities. It has also been observed that the
reflector positions have little influence on the microwave frequencies. To some degree,
this phenomenon reduces the stiength of the argument for different microwave reflection
uitensities. Another important phenomenon is that the middle part of the strip reflector
has an obvious damaged spot after a lot of shots. If this damaged spot is caused by the
microwave reflection, the spot should fade away gradually along the radius direction,
which is not the experimental case. We understand that a large fraction of the elections
collected in the virtual cathode move along the waveguide. We believe tiiat the damaged
spot is caused by those escaping elections. In this way, we consider the strip reflector as
an election collector. This election collection changes the field stiiicture in tiie diode
region, leading to a propagating mode shift. However, the microwave reflection is a
basic fimction of the reflectors.
52
Anotiier phenomenon worth mentioning is tiiat the dominant peak frequencies
may be half or double tiie secondary peak frequencies in some cases. We believe tiiat
this is a harmonic.
Given the above analysis, we can conclude tiiat the E-beam plays an important
part in forming a cavity m tiie geometry illustiated in Figure 4.2. The donut reflector has
little influence on the microwave frequency and the strip reflector has an obvious effect
on the microwave frequency. Although tiie reflectors reflect the microwaves, we believe
that coUectmg the elections has an overwhehning effect on the field stixicture in the diode
region.
We tmderstand that the virtual cathode oscillates in a certain frequency range,
which primarily depends on the gap distance of the diode and the apphed voltage.
Microwaves with that frequency range are produced by the vircator. Once a cavity forms
in the diode region, microwaves of a certam frequency in this range are enhanced and
non-resonant frequencies are suppressed. The frequency enhanced by the resonance
becomes the dominant peak. Because different mode resonant frequencies of a cavity
sometimes coincide, different modes may propagate in the waveguide sunultaneously,
which makes the mode measurement more complicated.
53
CHAPTER 5
PROPAGATING MODE DETERMINATION
The propagating mode is still a big problem. To evaluate the microwave power,
we need to know the propagating mode. To investigate the virtual cavity resonance
behavior, we need to know the propagating mode. We understand that the propagating
mode in a circular waveguide is determined by n, m in expressions of TMnm or TEnm,
where n represents the number of half waves along the radius direction and m represents
the number of half waves along the azimuth dnection. Once we determine the n and m,
we have to use relative information to determine TE or TM. Generally, the relative
information includes the cutoff frequency of the waveguide and the microwave source
resonance cavity.
5.1 Cutoff Frequencies of the Waveguide
We use an alummum waveguide to couple out the microwaves from the coaxial
diode. This wavegiude is 2.5 m long and 9.8 cm in radius. The cutoff frequencies for
specific modes are given by following equations [16]:
(/•) = \ ^ (5.1)
\JC)TM^ _ I '
2n-^ps a
if) = ^ ^ (5.2)
UC)TE^ f— '
54
where a represents tiie waveguide radius and jc„„ and x;^ represent the mth root of the
/ith order Bessel fimction and second kind of Bessel function, respectively. Tables 5.1
and 5.2 give these values. A cutoff frequency is the minimum frequency to set up a
specific mode in a waveguide. The microwaves above this frequency can tiavel in the
waveguide and the microwaves below thisfrequencywill die away soon.
Table 5.1 • x„„ Array
^nm
1 2 3
0 2.405 5.520 8.654
1 3.832 7.016 10.173
2 5.136 8.417 11.620
3 6.380 9.761 13.015
4 7.588 11.065 14.372
Table 5.2 • ^nn, Array
^nm
1 2 3
0 3.832 7.016 10.173
1 1.841 5.331 8.536
2 3.054 6.706 9.969
3 4.201 8.015 11.346
4 5.317 9.282 12.682
Table 5.3 and 5.4 show the cutofffrequenciesfor the waveguide.
Table 5.3. Cutoff Frequencies (GHz) for TMnm Mode

nxm 1 2 3
0 1.19 2.732 4.283
1 1.897 3.473 5.035
2 2.542 4.166 5.751
3 3.158 4.831 6.442
4 3.756 5.477 7.113
55
Table 5.4. Cutoff Frequencies (GHz) for TEnm Mode
nxm 1 2 3
0 1.897 3.473 5.035
1 0.9112 2.639 4.225
2 1.512 3.319 4.934
3 2.079 3.967 5.616
4 2.632 4.594 6.277
5.2 Field Distribution for Different Modes in a Circular

Waveguide
Observing the field lines for different modes in a circular waveguide is helpful in
determining the propagating mode. Figure 5.1 [17] shows 15 tiansverse field
distributions for a single mode in a circular waveguide.
From Figure 5.1, we can see the periodic E-field over both azimuth and radius
direction. If the polarization of a single frequency microwave is fixed, we can observe a
clear pattern like Figure 5.1. We have two E-probes with a 90° separation in azimuth and
Figure 5.1 shows the amplitudes for the two probes to be equal to each other for those
modes with «=0, 2, 4, ....
5.3 Wave Combinations
A single frequency signal is easy to analysis. Experimentally, we often get an EM
signal resultmg from a combination of signals with different frequency, and different
modes. To analyze this kind of a combined signal is very difficult. A basic method is to
separate these original signals from the combined signal.
56
Consider a wave packet consisting of two tiaveling waves with the same
amplitude [18], the same propagating mode, and slightly different frequencies o), + A«
rcv,v
and OQ - Aco. The phase constants, being functions of frequency as /3- —.,|l -
yea J
will also be slightiy different: /?„ + AJ3 and P^-Afi. We have
E{t, z) = E, cos[(co, +Ao})t- (yô + Ap)z]+E, cos[{co, -Aco)t- (JS, - Aj3)z]

= 2EQ cos{tAa) - zAP) cos{o)Qt - fi^z)
If the amplitudes are different, the situation will become more complicated. Figure 5.2 is
a waveform combination of W(t) = A(t) + B(t) + C(t), where A(t) - 0.5cos(<y,? -P^z),
B{t) = -1.5cos((Jt)2t-P2z) and C(0 = 2.5 cos(ftJ3^-/?3z) with z=1.5 m. We define
CO 1=2000 rad/s, co2=2001 rad/s, (03=1999 rad/s and pi, P2, Ps are the phase constants
corresponding to (Oi, C02, C03, respectively for (Oc=1500 rad/s. In Figure 5.2, we can see
the long-term result of the combination. Based on this example, we can separate the
different frequency signals using digital band pass filters.
57
TM„
TM„
TM„ TM.
TM,
Figure 5.1. Transverse Field Distiibution for a Curcular Waveguide (Fkst 15 Modes)
58
Based on the above analysis, a sum of multi-frequency microwaves with the same
mode or different modes has phase shifts from its original components and forms a wave
packet, while a sum of single frequency microwaves with different modes only yields a
phase shift. These combinations have the same effect on the detected horizontal and
vertical field signals. By applying band passfiltersto the detected horizontal and vertical
signals, we can get single-frequency signals.
Figure 5.2. Combmation of Three Time Harmonic Traveling Waves with Slightly
Different Frequencies: 1-10 s
59
5.4 Theoretical Analysis for Microwave Modes
All analysis methods are based on the electiomagnetic theory of cavities and
waveguides. The theoretical basis is the microwave functions tiaveling in a circular
waveguide. Here we list the wave fimctions for TM and TE modes separately [16].
TM Mode
E^ = -JPTU,J„ (Tp) cos(n,/>)e-^ê^'^ (5.4)
E, = -Û,J„ (Tp)sinin,^)e-^ê^'^ (5.5)

P
E^ = TÛoJ„ {Tp)cos{n^)e-^ê''^ (5.6)
^ ^ _Zif^^^j^(j^)sin(n^)e-^ê^'" (5.7)
P
H^ = -j(DsTUj„ {Tp) cos(n^)e-^ê^'* (5.8)
H, = 0 (5.9)
TEMode
^ m£Lv,JATp)sin{n<l>)e-^ê^'' (5.10)
" P
E^ = j(opTVj„ (Tp) cos{n<l>)e-'ê^'^ (5-11)
£^=0 (5-12)
H^ = -jjnVj„{Tp)cos{n(l>)e-^ê^'^ (5-13)
H. =J^V,J„(Tp)sin(n^)e-Jê''^ (5-14)
P
H, = TX^n (Tp)cos(n^)e-^ê''^ (5-15)
60
where J^Tp) and J'„(Tp) are tiie Bessel function and Bessel function of tiie second kind,
respectively; T = ^ for TM modes and T = ^ for TE modes with a as the radius of

a a
tiie waveguide; /? = — 1 - - ^ witii the cutoff frequency m ; UQ and Vo are

c\ \(o)
coefficients depending on field intensities. At the inner wall of the waveguide, only E
and H^ are non-zero components since J„ (To) = 0 for TM modes and J„ (Ta) = 0 for TE
modes, which are the boundary conditions on the surface of a conductor.
We detect E^ at two positions, where the ^ difference equals 90° with the same z
coordinate. Supposing a single-frequency microwave is stably generated in the diode
region and propagates as a single mode in the waveguide, the phase factor, cot-Pz,
yields a zero phase difference for two perpendicular E-probes. Define one E-probe
azimuth position as (po and the other E-probe azimuth position as ô+90°. Considering
the TM mode as an example, we find the ratio of E^ for the two E-probes
^ = ^°^("ô^ . (5.16)
cos(nô +n7r/2)
This value will determine the phase difference and the amplitude ratio. Eq. (5.16)
shows that the possible phase difference is 0° or 180° and the amplitude ratio varies,
depending on ôand n. We find tiiat // = 1 when n=0; fi^ -cot(^) when n=l; ^^-\
when n=2; // = cot(^) when n=3. Obviously, the same amplitude should be shown in
the detected horizontal and vertical fields when n=0, 2. For n=l, 3, the amplitiide ratio
61
can be any value, varying witii (po- If we consider the sign difference as a phase
difference, we can see tiiat it is 0° and 180° for n=0 and 2, respectively; for n^l or 3, the
phase difference could be 0° or 180°, varying with (po- For the TE mode system, we can
get a similar statement. A point worth mentioning here is that nK/2 itself doesn't
constitiite a phase difference. Once n and (po are determined, cos(n(!Jo),cos(«ô +nn:l2)
and p will be fixed and do not change over time.
Because the possible phase differences for two E-probes with 90° azimuth apart
are 0° or 180°, we cannot get any conclusion from it.
5.5 Analysis of Experimental Data
Our experimental data involves two frequency systems, one without reflectors or
with a donut reflector, the other with a strip reflector. We guess that these two frequency
systems have different propagating modes. So we choose one shot from each frequency
system to do more in-depth analysis.
5.5.1 An Example with a Donut Reflector or without Reflectors
For the frequency systems with a donut reflector or without reflectors, we choose
another shot in the same diode structure, from which we got Figures 4.2 and 4.3. This
diode structure has no reflector and Di= 5.08 cm. Because we recorded data with a
Tektionix TDS6604 witii 6 GHz bandwidth and a 20 GS/s samplmg rate, a better result is
expected.
62
-20 0 20 40 60 80 100 120 140 160 180
Time (ns)
Figure 5.3. Raw Horizontal and Vertical Probe Waveforms for a Shot: -20-180 ns
Figure 5.3 is the raw waveform for the shot mentioned above. From Figure 5.3,
we can see that most of the microwave power happens in the period of 0-100 ns. Figure
5.4 is the FFT spectia for the horizontal and vertical fields. We can see that the
microwave dominant peak is 3.16 GHz, which falls in the band of 3.10-3.20 GHz. So,
we use the following narrower bands to filter the horizontal and vertical waveforms
witiiin 0-100 ns. 3.10-3.1 IGHz, 3.11-3.12 GHz, 3.12-3.13 GHz, 3.13-3.14 GHz, 3.14-
3.15 GHz, 3.15-3.16 GHz, 3.16-3.17GHz, 3.17-3.18GHz, 3.18-3.19 GHz and 3.19-3.20
63
GHz. We get Figure 5.5 and Figure 5.6 fortiiesefiltereddata. Figure 5.7 is these fihered
data in a higher resolution, where we can see tiie phase difference between the horizontal
and vertical field with the samefrequencyband clearly.
0.03
E-Probe 3.16 GHz, 2.40
0.02
0.01
Z)
< kMbtMMMliifiaiiiaMMMrill rtWlVf>i.>>|WlitftHh*^ WAI'MW^ \Uii^.^m ,1
3
V E-Probe 3.16 GHz, 2.70
a.
E
<
0.02
0.01
0.00
Frequency (GHz)
Figure 5.4. FFT Spectia for Waveforms Shown m Figure 5.3
64
< 1 A&l.
.1
^'hLl ' d'l . ixriF J I HI J
III l < ^ II
Aftlllijijj.hll
ir*i%ii
1 f
i' • I I — I — I —
40 60 80 100
Time (ns)
Figure 5.5. Filtered Waveforms for tiie Horizontal Probe Shown in Figure 5.3: 0-100 ns
65
3.15E9 to 3.16^9 Hz Band Pass ffilter on ShortFieldiHReld
-0.03 -
40 60
Time (ns)
Figure 5.5. (Continued)
66
0.08
^ 3 1E9'to 3 11E^ Hz Band Pass Rlêr on ShortField îêld
0.04 [-,
0.00
-0.04
-o.(ie H \ ^ } \ \ ^
— — 3.11E9 to 3.12E9 Hz Band Pass Filter onShortFieldlVRdd
on Sl-
0.08
-0.08
Q. 0.06 ^ —U312Ebto3.13E9HzBai'idPassHllteronskTtF
Held IVH eld
< 0.03
0.00
-0.03
-0.0)3
0.06
0.00
-0.06
"• I ito3.15E9HzBafidPassliilterondkxtReldlvField '
0.05
0 20 40 60
Time (ns)
Figure 5.6. Filtered Waveforms for tiie Vertical Probe Shown m Figure 5.3: 0-100 ns
67
Q- 0.0
< -0.1
0.04
0.00 IllllWIMIIIIM
-0.04
40 60
Time (ns)
68
0.12
• 3.1E9 to 3.11E$ Hz Band Pass Rker on ShortRdd HReld
0.08 =ittert3ftShortReli1«1_A iO—1*J_.
0.04
0.00
-0.04
-OM
• 3.11E9 to 3.12f9 Hz Band Pass Filter on ShortFieldHReld
0.08 • 3.11F9 tn 3,t9lta Hz Band Pass F liter on ShnitRfiIri"
VReld
0.00
-0.08
Q. 0.12 -3.12Ebto3.1313IHzBartdPass Pa"CXI shortReld.HReld
E 0.08 -31?F9tn3 13iaIHzBandJBass-îlta^jnShortFidd VRfJri
<
0.04
0.00
-0.04
-OM
0.10
0.05
0.00
-0.05
ai4dbto3.15IE9HzBalid Pass Miter on dkxtReldFFiae-'-
0.15 "334B to 3:i5Bl^SnaP^"Firfa^ni9TQffFie& VFieia"
0.10
0.05
0.00
-0.05 ^
-0.10
-0.15
50.0 50.2 50.4 50.6 50.8 51.0
Time (ns)
Figure 5.7. Filtered Waveforms for the Fields Shown in Figure 5.3: Phase Difference
69
0.6
0.4
0.2
0.0
-0.2
-0.4
-3.16E9to3.1713I Hz Band PassFilter on ShortReld. HReld
-3.16E9to3.1713I Hz Band PassFilter on ShortReld VReld
0.2
3
<
0.0
c± -QA + +
E 3.17E9to3. 18E3 HzBandPassIR ter on ShortReld_HReld
< 3.17E9to3. 18E3 HzBandPassIRter on ShortReld i/Reld
0.2
0.0
0Q12 + +
3.18E9ito3.19E6 Hz Band Pass Rfter on ShortReld hReld
0.08 3.18E9 to 3.19E»tizreanctPa^frter on ShortReld 'mm
0.04
0.00
-0.04
-0.0§ + + +
3.19E9to3.2E9 I t Band Pass RIt 3-on ShortReld HReld
0.06
0.03 -
0.00
-0.03
-0.06
50.0 50.2 50.4 50.6 50.8 51.0
Time (ns)
70
0 .2
0 .0
3
<
- 0 .2
0 .0
-0 ,2
4 0 6 0 1 0 0
Time (n s )
Figure 5.8. Filtered Waveforms for the Horizontal Probe Shown in Figure 5.3 with Two
Narrower Band Filters: 0-100 ns
From Figure 5.5 and 5.6, we can get horizontal and vertical waveforms with
constant amplitudes for each band except the band of 3.15-3.16 GHz. Based on the
example in Chapter 4, we understand that the waveforms for the band of 3.15-3.16 GHz
are composed of two or more slightly different frequency microwaves. We use two
narrower bands, 3.15-3.155 GHz and 3.155-3.16 GHz, to filter the original waveforms
within 0-100 ns. We get the filtered waveforms as shown in Figures 5.8, 5.9 and 5.10.
From Figure 5.7, we can see tiiat some phase differences are 180°, some are 0° and some
are between the two of tiiem. Obviously those witii 180° or 0° phase difference are smgle
frequency signals. We have no idea why some of them are not 180° or 0° phase
difference although constant amplitiides show them a singlefrequencysignals. Because
71
tile microwavefrequenciesm the diode region are very complicated, we pay our attention
to tiiose witii 180° or 0° phase difference. Most of tiiese phase differences are constant
from tiie beginning to tiie end. Based on our examples in Chapter 4, if tiie ratio of the
horizontal and vertical field amplitudes varies over time, there are two or more slightly
different frequency microwaves.
0.2
0.0
d. -0.3
E
<
0.0
-0.3
20 40 60 100
Tim e (ns)
Figure 5.9. Filtered Waveforms for the Vertical Probe Shown in Figure 5.3 witii Two
Narrower Band Filters: 0-100 ns
72
From Figures 5.8 and 5.9, we notice that microwaves witii a varying amplitude
ratio are filtered to two sUghtly different frequency microwaves. Reviewing Figure 5.10,
we can seetiiattiiephase difference for tiie two bands is 180° and the amplitiide ratios for
the two bands are close to 1.
0 .0
3
<
5 0.4 5 0.6
Time (n s )
Figure 5.10. Filtered Waveforms for the Fields Shown in Figure 5.3 with Two Narrower
Band Filters: Phase Difference
Now our original raw waveforms have been filtered into sub-waveforms with
different constant amplitudes. Most of them show 0° or 180° phase differences, which is
consistent with the theoretical analysis above. Based on the above analysis, we cannot
determine the n value m the mode determination from the two E-probes because there are
two unknown variables m Eq. (5.16). Another problem is to determine the m value (in
some books, it is referred to as the p value.) in the mode determination. We understand
73
tiiat n refers to the number of half-wave field variations in the ^-direction and m refers to
tiie number of half-wave field variations m the r-direction. An easy way to get the
information on m is to get a pictiu-e from a gas bulb array at the open end of the
waveguide. These gas bulbs light up when illuminated by the microwaves and show
where tiie microwave power is largest. With these improvements, we can get the
propagating mode and recalculate tiie microwave power in a better way.
5.5.2 An Example with a Stiip Reflector
150 200 250 300 350

Time (ns)
Figure 5.11. Raw Waveforms for a Shot: 50-350 ns
74
For tiie frequency system witii the stiip reflector, we choose a shot with a diode
stiiicture, which has Di= 5.08 cm and witii tiie stiip reflector at D2- 4 cm. For tiiis shot,
we recorded data with an oscilloscope witii an Agilent Infiniium oscilloscope with
2.25GHz bandwidth, and an 8 GS/s sampling rate.
0.000
0.0 0.5 1.0 1.5
Frequency (GHz)
Figure 5.12. FFT Spectia for Waveforms Shown in Figure 5.11
75
0.50
1.95Ktd1.96E9HzBandPassRlteraiData1 HReld
50 100 150 200 250 300 350

Time (ns)
Figure 5.13. Filtered Waveforms for Horizontal Probe Shown in Figure 5.11: 50-350 ns
76
01E9 Hz Band Pass RIterdn Dalai HFIek
-0.25 - ^ " f
200
Time (ns)
77
200
Time (ns)
Figure 5.14. Filtered Waveforms for the Vertical Probe Shown in Figure 5.11: 50-350 ns
78
2E9 [A 2.01 E9 Hz Band Pass Filter oh Data2 VField
200
Time (ns)
79
Figure 5.11 is tiie raw waveform for tiie shot mentioned above. From Figure 5.11,
we can see tiiat most of tiie microwaves happen in the period of 50-250 ns. ft is not clear
why tiie microwave last about 200 ns, which is tiiple of the pulse lengtii of tiie applied
voltage, 70 ns. Figure 5.12 is tiie FFT spectia for tiie horizontal and vertical fields. We
can see that tiie dominant frequency peak is 1.97 GHz or 1.99 GHz, which falls in the
band of 1.95-2.05 GHz. So, we use tiie following narrower bands to filter the horizontal
and vertical waveforms within 50-350 ns. 1.95-1.96 GHz, 1.96-1.97 GHz, 1.97-1.98
GHz, 1.98-1.99 GHz, 1.99-2.00 GHz, 2.00-2.01 GHz, 2.01-2.02 GHz, 2.02-2.03 GHz,
2.03-2.04 GHz and 2.04-2.05 GHz. We get Figure 5.13 and Figure 5.14 from these
filtered data.
From Figures 5.13 and 5.14, we can see that tiie waveforms for different bands
are not as neat as the above example. It seems that each band waveform is a multi-
frequency signal combmation. We do not do any fiirther analysis for this group of data.
5.6 Discussions on Propagating Modes
For 3.16 GHz microwaves, the possible modes are TMoi, TMn, TM21, TM02,
TM31, TEii, TE21, TE31, TE41, and TE12, whose cutoff frequencies are less than 3.16 GHz.
For 2.0 GHz microwaves, the possible modes are TMoi, TMn, TEn, TE21 and TE3),
whose cutoff frequencies are less than 2.0 GHz. Here we consider the measurement
errors, calculation errors and analysis errors to include TE31 as a possible mode for a 2.0
GHz microwave.
80
Let us examine tiie first example. The raw waveforms can be separated into
several single frequency signals witii narrow band filters. From Figure 5.7, we can see
tiiat most of tiie suigle frequency signals have a 180° phase difference or a 0° phase
difference between tiie horizontal and vertical fields, which is consistent with the analysis
above. Altiiough tiie amplitiide ratio of the horizontal and vertical fields is ahnost
constant for one single frequency, these ratios for different bands are not same. If we
assume all these smgle frequency signals share the same propagating mode, we can easily
find that those propagating modes with n=0, n=2 or n=4 are not possible modes since the
amplitude ratio for two E-probes with an azimuth difference of 90° should be equal to 1
for those modes. This statement can be justified from Figure 5.1. The possible modes
for a 3.16 GHz microwave are TMn, TM31, TEn, TE31, or TE12. For tiie 2.0 GHz
microwave, we have no further smgle frequency information, so we can not do further
mode exclusions here. We believe that the 2.0 GHz microwave could be filtered as
single frequency signals if we could record those microwave signals with a higher speed
oscilloscope.
The possible modes for the 3.16 GHz microwaves could be TMn, TM31, TEn,
TE31, or TE12. To get the exact mode information, we have to improve the experimental
configuration. Let us rewrite and examine Eq. (5.16)
^ = ^°^("ô^ , (5.16)
cos(nô + n;r/2) '
81
This equation is an amplitiide ratio of detected microwaves from two E-probes with nl2
azimuth difference, sharing the same z position. If we have another E-probe between tiie
two E-probes in the azimuth dkection, we can get another amplitude ratio equation
cos{n^,)
^'^ r A J.V (5.17)
cos(nô+«^J ^ '
where (p^ is the new angle direction. Once we get a smgle frequency using digital filters,
we can get two amplitude ratio equations for three E-probes along the azimuth direction.
The two equations have two unknown variables, ^, and n, which means that we can get
these two variables easily. This way, we can get an important clue for the single
frequency microwave. Once we get the half-wave number along the azimuth direction, n,
we need to know the half-wave number along the radial dfrection, m. To determine m,
we can install three E-probes along the radial direction at the waveguide open end or we
can take a picture of a gas lamp array at the waveguide open end. However, these
methods are complicated and can not solve the problem of the identification of TE or TM
modes. Let us examine the following expressions,
E = Acos{n(t>)e~'ê''^ = Acos(n(/>)cos((Ot-pz), (5.18)
2
^« >
A=f,(.- (5.19)
Eq. (5.18) is a general expression for E^ which occurs in TE or TM modes. Eq. (5.19) is
a phase factor for a specific cutoff frequency, which is connected with tiie propagating
mode. This means, we can determine the mode directly if we can measure tiie phase
factor, ;9, for a certain frequency microwave. Eq. (5.18) offers us a way to measure the
82
phase factor. If we set two E-probes on the waveguide along the z direction, sharing the
same azimuth position, tiie phase difference between tiie two detected microwaves comes
from Pz in Eq. (5.18). Once we get the phase difference, we can get the cutoff circular
frequency, (o^, which is unique for any mode m a fixed size waveguide. Occasionally,
two cutoff frequencies for two different modes coincide. To solve this problem, we need
to discover the n value to determine the mode.
The above analysis offers a shnple method to determine the microwave
propagating mode with two E-probes along the z dhection. To avoid the cutoff frequency
coincidence for different modes, two extia E-probes are necessary to determine the n
value at the same time.
83
CHAPTER 6
CONCLUSIONS AND SUGGESTIONS
This dissertation can be divided into two parts: tiie theoretical part and tiie
experimental part. The common goal is to investigate the microwave frequency and
propagating mode determination mechanisms. The other purpose of this project is to
improve the power efficiency.
In this dissertation, a physical approximation method is intioduced to analyze the
nonlinear differential equation. An approximate analytical expression for the space-
charge limited current in a 1-D cylindrical diode is proposed. This approximation
method is used to analyze a relativistic corrected nonlinear differential equation for the
same physical situation and we get a relativistic corrected space-charge limited current.
Understanding the behavior of space charges in cylindrical diodes, we extend our
analytical expressions from 1-D to 2-D approximately.
Some possible applications of these formulas are proposed. The above
development has given some guidance on the diode design, especially in the V-I
characteristics of the diode and the virtual cathode oscillation frequency.
A set of experiments have been conducted and they prove an important h5ôthesis.
a virtual cavity forms in the cylindrical diode at TTU. The detailed physical mechaiusm
of the virtual cavity is still not clear because of the lack of the propagating mode
information. The propagating mode determination becomes very important for it is the
basis of the cavity resonance frequencies and the microwave power calculation.
84
Some attempts have been made in identifying the mode in previous research.
Typically, taking the pictures of a gas bulb array at the open end of tiie waveguide and
detectmg tiie E fields at tiie wall of tiie waveguide are primary metiiods. It is established
that the TEn mode is the dominant propagating mode. Based on the theoretical analysis,
one thing, the n value, is left undetermined assuming the m value is right from the picture
of tiie discharged gas bulb array at the end of the waveguide. We believe that the TEn
mode is possible but can not exclude the TMn or otiier modes. A fast way to determine n
and m values is to mstall the tiurd E-probes and take a picture of the gas lamp panel at tiie
open end of the waveguide. The most efficient way to determine the propagating modes
is to install four E-probes on the waveguide wall and observe the microwave properties
from the four E-probes.
Once the microwave propagatmg mode can be determined, our research on the
virtual cavity can be developed further. We can do further analysis on the virtual cavity
physical mechanism. Actually, the region called a virtual cavity is very hard to analyze
because of its complex structure.
Theoretically, we intioduced a physical approximation to analyze nonlinear
differential equations and got approxunate analysis expressions for one-dunensional and
two-dimensional cylindrical diodes. We intioduced some possible applications of these
results. Experimentally, we found that the microwave frequency of the vircator at TTU
stiongly depends on the velvet emitter position and a concept of virtual cavity is
intioduced. To do the fiirther research in physical mechanisms of a coaxial vircator, an
85
idea to determine the microwave propagating modes is intioduced based on the
microwave functions in a circular waveguide.
86
REFERENCES
1. K. S. Woolverton, "High-Power, Coaxial Vircator Geometiies," Ph.D.

Dissertation, Texas Tech University, 1998.
2. W. Jiang, K. Woolverton, J. Dickens, M. Kristiansen, "High Power Microwave

Generation by a Coaxial Virtual Catiiode Oscillator," IEEE Trans. Plasma Sci.,
Vol. 27, No. 5,1999.
3. J. Benford and L. Swegle, High-Power Microwaves, Boston: Artech House, Inc.

1992.
4. H. A. Davis, R. D. Fulton, E. G. Sherwood and T. J. T. Kwan, "Enhanced-

Efficiency, Narrow-Band Gigawatt Microwave Output of the Redition
Oscillator," IEEE Trans. Plasma Sci., Vol. 18, No. 3, 1999.
5. C D . Child, "Discharge from Hot Cao," Phys. Rev., Vol. 32, 1911.
6. I. Langmuir and K. Blodgett, "Currents limited by space charge between coaxial

cylmders," Phys. Rev., Vol. 22, 1925.
7. J. Benford, D. Price, H. Sze and D. Bromley, "Interaction of a Vircator

Microwave Generator With an Enclosing Resonant Cavity," J. Appl. Phys., Vol.
61,No. 5, 1987.
8. W. Jiang, J. Dickens, M. Kristiansen, "Efficiency Enhancement of a Coaxial

Vutual Cathode Oscillator," IEEE Trans. Plasma Sci., Vol. 27, No.5, 1999.
9. H. R. Jory and A. W. Trivelpiece, "Exact Relativistic Solution for the One-

Dimensional Diode," J. Appl. Phys., Vol. 40, No. 10,1969.
10. R. B. Miller, An Introduction to the Physics of Intense Charged Particle Beams,

New York: Plenum Press, 1982.
11. H. W. Chan, R. C. Davidson, K. T. Nguyen and H. S. Uhm, "Relativistic non-

neutial election flow in a planar tiiode," J. Appl. Phys., Vol. 61, No. 6,1987.
12. Y. Y. Lau, "Simple Theory for the Two-Dunensional Child-Langmuir Law," Phys.
Rev.Zett.'vol. 87, No. 27, 2001.
87
13. C. S. Hwang, M. W. Wu, P. S. Song and W. S. Hou, "High Power Microwave
Generation from a Tunable Radially Extiacted Vircator," /. Appl. Phys., Vol. 69,
No. 3, 1991
14. A. Kadish, R. Gaehl and C. Snell, "Analysis and Simulation of Virtual Cathode
Oscillations," Phys. Fluids, Vol. 29, No.2,1986.
15. J. D. Jackson, Classical Electrodynamics, New York: John Wiley & Sons, INC,
1998.
16. K. Zhang and D. Li, Electromagnetic Theory for Microwaves and Optoelectronics,
Berim: Springer-Veriag, 1998.
17. A. F. Harvey, Microwave Engineering, New York: Academic Press, 1963
18. D. K. Cheng, Field and Wave Electromagnetics, New York: Pearson Education,
INC, 1998.
88
APPENDIX
DATA ANALYSIS STATISTICS FOR CAVITY EFFECT
RESEARCH
89
Table A. 1 Frequency Record 1 in Cavity Effect Research at TTU
Freq H E-Pro 36 Peaks V E-Probe Peaks Norm. V Norm. I
Band Freq. (GHz) Amp. (A.U.) Freq. (GHz) Amp. (A.U) (A. U.) (A. U.)
Shot 60 2.90 4.05 3.18 4.46 2.47 1.40
Shot 64 3.20 4.15 3.19 2.80 2.45 1.45
Shot 68 1.90/3.17 1.30/2.12 1.90/3.17 2.10/2.90 2.62 1.55
Shot 70 1.91/3.19 1.70/1.85 1.90/3.17 2.30/3.30 2.63 1.52
Shot 74 1.91/3.23 2.40/2.0 1.91/3.17 3.95/4.45 2.68 1.64
Shot 78 2.02 11.0 2.0 13.5 2.82 1.56
Shot 80 1.91 5.25 2.0 13.0 2.83 1.46
Shot 82 2.04 15.6 2.0 13.0 2.83 1.62
Note.
Experiment Date. Feb. 26,2003
Physical Configuration:
Emitter Width. 1.375", Emitter Position. Di=4 cm. No Reflector
Table A.2 Frequency Record 2 in Cavity Effect Research at TTU

Freq H E-Probe Peaks V E-Probe Peaks Norm. V Norm. I
Shot 60 3.06/3.16 1.99/1.54 3.06/3.16 6.06/1.84 2.34 1.18
Shot 64 3.21 1.49 3.06 3.07 2.49 1.24
Shot 68 3.14 1.42 3.12 4.10 2.65 1.45
Shot 70 3.14 1.72 3.12 4.59 2.67 1.47
Shot 74 3.22 2.16 3.13 3.51 2.65 1.34
Shot 78 3.13 1.73 3.12 3.97 2.72 1.31
Shot 80 2.07 3.6 2.07/3.18 1.79/2.48 2.84 1.38
Shot 82 2.08 1.59 1.94/3.18 1.34/3.74 2.89 1.24
Note:
Experiment Date. Feb. 25, 2003
Emitter Width. 1.375", Emitter Position. Di=4 cm, Donut Reflector Position. D2=3.08 cm
90
Table A.3. Frequency Record 3 in Cavity Effect Research at TTU
Freq HE-Prol)e Peaks V E-Probe Peaks Norm. V Norm. I
Shot 60 1.81/3.19/3.78 2.86/1.09/4.86 1.94/3.19/3.86 11.8/2.22/5.38 2.39 1.26
Shot 64 1.99/3.19 7.61/1.82 1.99/3.17 6.52/2.78 2.56 1.29
Shot 68 1.99/3.18 10.3/2.90 1.99/3.16 7.79/2.96 2.61 1.39
Shot 70 2.0/3.17 12.6/2.06 2.0/3.16 8.51/3.10 2.68 1.48
Shot 74 2.0/3.19 11.6/2.36 1.99/3.16 8.43/2.99 2.69 1.60
Shot 78 2.01/3.17 17.3/2.10 2.0/3.19/3.98 9.74/2.27/4.04 2.80 1.57
Shot 80 2.03/3.19 1.20/2.68 2.04/3.19 1.92/4.30 2.59 1.71
Shot 82 2.03/3.16 7.79/3.18 2.0/3.19 7.79/3.10 2.62 1.66
Note:
Experiment Date. March 03,2003
Emitter Widtii. 1.375", Emitter Position. Di=4 cm, Stiip Refiector Position. D2=3.08 cm

Shot 60 1.99/3.16 2.89/1.32 1.91/3.17 2.57/2.72 2.34 1.24
Shot 64 3.16 2.46 3.16 3.49 2.37 1.45
Shot 68 2.0/3.22 1.85/2.73 1.99/3.19 2.32/3.83 2.59 1.43
Shot 70 3.08 2.55 3.14 2.23 2.35 1.59
Shot 74 3.22 2.64 3.19 2.69 2.52 1.50
Shot 78 2.08/3.36 1.04/2.32 3.26 4.72 2.76 1.26
Shot 80 1.99/3.16 3.32/1.49 1.99/3.19 6.68/3.70 2.79 1.27
Shot 82 2.03 7.69 2.01/3.27 4.99/3.35 2.85 1.36
Note:
Experiment Date. March 03, 2003
Physical Configuration
91
Freq H E-Prol3e Peaks V E-Probe Peaks Norm. V Norm. I
Shot 60 1.92/2.69/2.90/ 3.28/0.57/1.12/ 1.93/3.14/3.86 12.9/2.86/2.65 2.34 1.38
3.16/3.88 1.66/2.05
Shot 64 1.93/3.17/3.86 3.93/0.70/1.90 1.94/3.14/3.86 14.2/1.91/2.04 2.45 1.13
Shot 68 1.96/3.16/3.90 4.81/1.96/3.63 1.95/3.15/3.90 11.0/3.41/4.37 2.45 1.51
Shot 70 1.96/3.19/3.91 4.33/1.79/3.06 1.95/3.17/3.90 12.7/2.82/4.38 2.56 1.40
Shot 74 2.01/3.16/3.91 5.42/1.60/3.23 1.99/3.17/4.0 18.4/2.38/3.70 2.71 1.66
Shot 78 1.99/3.19/3.99 9.91/2.52/4.78 2.0/3.19/3.95 20.8/2.546/3.5 2.85 1.63
Shot 80 2.01/3.18/3.97 7.32/2.39/3.69 2.0/3.16/3.96 12.0/2.54/3.82 2.63 1.66
Shot 82 2.0/3.19/3.91 5.08/1.74/3.84 1.99/3.17/3.91 16.9/3.43/4.05 2.73 1.56
Note:
Emitter Widtii. 1.375", Emitter Position. Di=4 cm, Stiip Refiector Position. D2= 4.08 cm

Band Freq. (GHz) Amp. (A.U.) Freq. (GHz) Amp. (A.U) (A.U.) (A.U.)
Shot 60 1.90/3.0 3.27/1.20 1.89/3.17 3.35/3.09 2.36 1.27
Shot 64 1.99/3.15 10.9/1.1 1.99/3.16 7.0/2.32 2.56 1.28
Shot 68 2.02/3.16 1.66/2.04 2.01/3.16 1.92/3.48 2.62 1.35
Shot 70 3.08 1.71 3.08 2.91 2.32 1.56
Shot 74 3.18 3.17 3.19 5.37 2.40 1.72
Shot 78 2.02/3.19 1.48/3.34 2.01/3.16 1.47/3.27 2.57 1.72
Shot 80 3.0 1.71 3.09 5.71 2.39 1.70
Shot 82 2.02/3.19 1.1/1.98 3.19 4.2 2.62 1.88
Note:
92
Shot 60 1.89/3.76 2.13/3.50 1.89/3.76 1.54/3.23 2.35 1.24
Shot 64 1.91/3.84 3.87/4.20 1.87/3.82 16.5/3.88 2.49 1.26
Shot 68 1.91/3.81 3.59/3.76 1.89/3.87 11.0/3.96 2.63 1.44
Shot 70 1.93/3.87 3.72/5.74 1.92/3.87 11.4/5.54 2.72 1.38
Shot 74 1.90/3.81 1.0/4.87 1.92/3.19/3.86 11.0/2.22/4.71 2.82 1.53
Shot 78 2.0/3.88 2.82/3.96 1.94/3.17/3.87 11.9/2.19/4.54 2.79 1.37
Shot 80 1.94/3.88 3.32/5.01 1.93/3.19/3.86 10.4/1.6/3.21 2.72 1.62
Shot 82 1.94/3.82 2.72/3.45 1.94/3.19/3.82 10.9/1.87/4.81 2.72 1.55
Note:
Emitter Width. 1.375", Emitter Position. Di=4 cm, Stiip Reflector Position. D2=5.08 cm

Band Freq. (GHz) Amp. (A.U.) Freq. (GHz) Amp. (A.U) (A. U.) (A.U.)
Shot 60 3.15 4.61 3.14 7.29 2.41 1.24
Shot 64 3.14 2.96 3.14 6.81 2.38 1.28
Shot 68 3.12 3.1 3.14 4.14 2.61 1.45
Shot 70 3.13 2.89 3.14 7.27 2.56 1.45
Shot 74 3.13 4.07 3.13 4.65 2.64 1.41
Shot 78 1.92/3.13/3.23 2.32/1.73/1.98 2.0/3.16 3.82/3.78 2.71 1.62
Shot 80 3.12 6.02 3.13 5.11 2.77 1.42
Shot 82 2.02/3.25 2.74/2.95 3.25 3.69 2.80 1.55
Note:
93
Shot 60 1.86/3.72 3.93/8.67 1.85/3.71 19.3/5.35 2.27 1.26
Shot 64 1.90/3.82 3.50/3.92 1.85/3.82 10.1/6.12 2.58 1.30
Shot 68 1.89/3.85 4.09/4.89 1.92/3.19/3.86 12.8/2.23/3.70 2.70 1.45
Shot 70 2.01/3.78 4.53/4.80 1.92/3.19/3.85 13.6/2.02/3.66 2.72 1.24
Shot 74 2.01/3.84 9.57/4.21 1.92/3.16/3.86 13.1/1.95/3.94 2.66 1.38
Shot 78 2.01/3.16/3.85 4.19/1.29/4.13 1.92/3.19/3.86 12.3/2.35/6.25 2.80 1.37
Shot 80 1.93/3.84 2.43/4.27 1.92/3.19/3.86 12.6/2.39/4.86 2.82 1.54
Shot 82 1.81/3.19/3.78 2.86/1.09/4.86 1.84/3.19/3.86 11.8/2.22/5.38 2.88 1.50
Note:
Emitter Widtii. 1.375", Emitter Position. Di=4 cm, Stiip Reflector Position. D2=6.08 cm
Table A. 10 Frequency Record 10 in Cavity Effect Research at TTU

Shot 60 2.92 19.0 2.92 21.5 2.56 1.32
Shot 62 3.14 22.0 ?? ??0 2.55 1.33
Shot 64 3.15 32.0 3.15 17.2 2.62 1.40
Shot 66 2.90 15.8 2.91 18.0 2.57 1.24
Shot 68 2.94/3.10 6.70/8.0 2.93/3.07/3.17 12.5/13.5/11.0 2.59 1.26
Shot 70 3.16 19.0 ?? ?? 2.73 1.72
Shot 72 3.17 8.0 3.17 19.5 2.80 1.87
Shot 74 3.15 6.30 3.15 15.0 2.73 1.41
Shot 76 3.16 18.0 3.19 12.8 2.82 1.89
Shot 78 3.19 10.5 3.16 12.0 2.87 2.11
Shot 80 3.25 21.3 3.16 17.0 2.95 1.96
Shot 82 3.24 9.1 ?? ?? 2.94 1.83
Note:
Emitter Width. 1.375", Emitter Position. Di=2", No Reflector
94
Table A. 11. Frequency Record 11 in Cavity Effect Research at TTU
Shot 60 3.06 3.74 3.06 12.8 2.55 1.27
Shot 62 3.08 3.17 3.07 6.60 2.56 1.26
Shot 64 3.09 2.47 3.09 5.95 2.47 1.36
Shot 66 3.07 2.50 3.06 8.75 2.52 1.30
Shot 68 3.14 2.33 3.14 7.30 2.58 1.48
Shot 70 3.16 4.49 3.17 6.81 2.62 1.53
Shot 72 3.16 4.19 3.16 6.04 2.61 1.75
Shot 74 3.18 4.53 3.17 6.88 2.66 1.61
Shot 76 3.16 5.28 3.17 4.36 2.67 1.95
Shot 78 3.16 8.12 3.17 5.06 2.75 1.67
Shot 80 3.06 3.25 3.09 3.47 2.52 1.51
Shot 82 3.16 1.83/3.86 3.15 3.22 2.77 1.68
Note:
Expermient Date. Feb. 25,2003
Emitter Widtii. 1.375", Emitter Position. Di=2", Donut Reflector Position. D2=3 cm

Shot 60 1.93/2.89 2.10/2.03 1.92/2.89 9.95/4.48 2.5 1.29
Shot 62 1.91/2.89 2.49/2.14 1.92/2.89 11.9/4.48 2.44 1.33
Shot 64 2.0/2.91 4.8/1.76 1.96/2.91 9.5/2.3 2.60 1.38
Shot 66 2.0/2.91 5.72/1.95 2.0/3.06 10.7/1.80 2.65 -
Shot 68 2.0/2.92 2.47/1.97 2.0/2.90 10.0/2.76 2.58 1.20
Shot 70 2.01/3.06 3.77/1.88 2.0/3.06 11.8/3.2 2.59 1.37
Shot 72 2.03/3.07 5.79/1.11 2.0/2.91 13.76/2.16 2.63 1.52
Shot 74 2.03 7.09 - 2.65 1.49
Shot 76 2.03 8.77 2.03 11.9 2.80 1.35
Shot 78 2.03 7.83 2.03 12.1 2.81 1.53
Shot 80 2.05 8.03 2.07 15.6 2.91 1.71
Shot 82 2.04 7.29 2.03 14.4 2.83 1.79
Note:
Experiment Date. Feb. 24,2003;
Emitter Widtii. 1.375", Emitter Position. Di=2", Stiip Reflector Position. D2=3 cm.
95
Shot 60 2.99 1.85 2.97 3.60 2.21 1.34
Shot 62 3.03 3.10 3.04 6.40 2.36 1.23
Shot 64 3.08 3.13 2.53/3.05 1.61/3.89 2.49 1.37
Shot 66 3.08 3.46 2.53/3.05 1.67/3.42 2.53 1.31
Shot 68 3.07 3.79 3.06 6.52 2.43 1.53
Shot 70 3.13 3.50 3.14 2.40 2.54 1.60
Shot 72 3.16 3.90 3.16 6.75 2.67 1.52
Shot 74 3.14 4.06 3.16 5.23 2.64 1.65
Shot 76 3.09 4.14 3.09 2.93 2.46 1.87
Shot 78 3.08 2.77 3.04 2.88 2.52 1.85
Shot 80 2.53/3.15 0.86/3.25 2.53/3.16 1.57/5.97 2.70 1.87
Shot 82 3.06 2.10 3.08 2.85 2.68 1.80
Note:

Shot 60 1.91/-/3.84 15.3/-/13.4 1.91/2.07/3.86 16.5/18.1/10.5 2.35 1.12
Shot 62 1.94/-/3.87 15.7/-/17.2 1.92/2.06/3.88 15.5/18.0/19.0 2.49 1.14
Shot 64 1.93/-/3.89 14.7/-/22.7 1.93/2.05/3.88 8.05/9.5/33.1 2.52 1.11
Shot 66 1.95/-/3.89 14.5/-/14.3 1.95/2.05/3.88 11.1/7.5/20.9 2.57 1.35
Shot 68 1.95/-/3.89 14.1/-/12.3 1.95/-/3.89 10.5/-/18.0 2.75 1.33
Shot 70 1.95/-/3.93 20.0/-/7.96 1.95/2.04/3.93 12.4/10.9/9.3 2.75 1.49
Shot 72 1.96/-/3.91 17.0/-/14.8 1.96/2.04/3.91 15.5/11.0/10.1 2.78 1.29
Shot 74 1.99/3.93 16.4/16.9 1.98/3.98 14.1/12.2 2.82 1.51
Shot 76 1.95/3.91 11.7/14.3 1.96/3.94 11.6/11.5 2.71 1.21
Shot 78 1.99/3.93 20.1/13.3 1.97/3.94 11.9/7.7 2.69 1.58
Shot 80 1.97/3.97 13.4/15.9 2.01/3.93 12.5/17.8 2.85 1.44
Shot 82 1.99/3.97 14.3/9.6 1.99/3.98 12.2/15.5 2.85 1.60
Note:
Experiment Date. Feb. 18, 2003;
Physical Configuration.
Emitter Width. 1.375", Emitter Position. Di=2", Stiip Reflector Position. D2=4 cm.
96
Shot 60 1.94/3.08 1.13/1.65 3.04 3.86 2.36 1.23
Shot 64 1.99/3.07 8.03/1.76 1.99/3.07 5.67/7.15 2.46 1.34
Shot 68 2.0/3.14 1.8/2.06 3.07 4.22 2.58 1.39
Shot 70 2.0/3.14 3.79/3.16 2.0/3.14 1.55/4.44 2.56 1.37
Shot 74 3.16 2.89 3.14 6.43 2.66 1.54
Shot 78 2.02/3.14 1.11/1.73 2.02/3.16 1.45/4.56 2.77 1.71
Shot 80 2.0/3.16 0.95/3.34 3.16 7.22 2.72 1.76
Shot 82 3.14 5.08 3.15 5.66 2.63 1.81
Note:
Emitter Widtii. 1.375", Emitter Position. Di=2", Stiip Reflector Position. D2=4 cm
Table A.16. Frequency Record 16 m Cavity Effect Research at TTU

Band Freq. (GHz) Amp. (A.U.) Freq. (GHz) Amp. (A.U) (A.U.) (A U.)
Shot 60 1.86/2.13/3.74 7.79/5.81/10.9 1.86/2.14/3.75 9.44/9.91/9.40 2.26 1.19
Shot 62 1.87/2.13/3.80 7.40/9.40/7.38 1.87/2.11/3.75 9.74/6.63/7.74 2.46 1.31
Shot 64 1.88/2.17/3.80 19.7/10.5/7.60 1.92/2.10/3.77 9.60/13.6/3.81 2.53 1.16
Shot 66 1.90/2.10/3.78 8.1/18.1/7.9 1.88/2.10/3.84 12.1/7.5/11.2 2.53 1.33
Shot 68 1.89/2.10/3.84 20.1/11.1/6.76 1.89/2.11/3.76 12.5/12.9/6.14 2.61 1.29
Shot 70 1.93/2.07/3.86 8.29/5.50/7.12 1.93/2.07/3.86 10.5/8.26/20.0 2.60 1.42
Shot 72 1.93/2.07/3.86 14.5/7.63/9.70 1.93/2.07/3.86 14.6/7.99/16.6 2.69 1.25
Shot 74 1.95/2.06/3.90 6.35/4.1/9.5 2.69 1.59
Shot 76 1.93/2.07/3.88 12.6/6.45/5.63 1.93/2.05/3.88 13.9/5.68/11.0 2.77 1.57
Shot 78 1.93/2.09/3.90 6.40/4.51/9.5 1.93/2.05/3.90 6.20/7.0/8.71 2.83 1.70
Shot 80 1.93/2.10/3.9 7.18/4.37/8.40 1.93/2.05/3.89 9.02/5.47/10.5 2.79 1.52
Shot 82 1.98/2.06/3.90 6.11/4.80/5.63 1.93/2.07/3.92 5.22/3.97/5.45 2.87 1.74
Note:
97
Shot 60 1.91/3.01 2.17/1.51 1.92/2.97 1.44/2.78 2.15 1.21
Shot 64 2.01/3.13 8.72/2.64 2.0/3.13 4.13/3.25 2.47 1.35
Shot 68 3.14 3.89 3.07 3.65 2.56 1.52
Shot 70 3.14 4.09 3.16 3.67 2.62 1.44
Shot 74 3.14 2.20 3.14 3.11 2.69 1.56
Shot 78 3.16 3.22 2.37/3.16 1.27/2.46 2.69 1.53
Shot 80 2.03/3.25 1.95/3.17 3.25 4.47 2.71 1.54
Shot 82 2.03/3.19 1.54/2.21 2.48/3.19 1.08/3.89 2.76 1.46
Note:

Shot 60 1.92/2.18/3.76 5.2/5.38/5.26 1.92/2.17/3.75 5.20/6.13/4.63 2.27 1.30
Shot 62 1.93/2.07/3.86 6.40/3.49/5.33 1.93/2.06/3.86 5.22/4.44/4.69 2.37 1.34
Shot 64 0.86/3.10 2.44/2.10 3.11 2.88 2.52 1.33
Shot 66 2.0/3.16 2.56/2.22 1.73 3.0 2.57 1.35
Shot 68 1.93/2.15/3.88 3.84/3.13/4.76 1.94/2.06/3.77 3.07/6.17/2.30 2.50 1.26
Shot 70 1.94/3.86 1.9/1.56 2.0/3.94 1.25/1.2 2.64 1.49
Shot 72 1.94/3.98 1.87/1.93 2.01/4.0 1.5/1.83 2.72 1.52
Shot 74 1.97/3.98 1.63/1.42 2.0/3.98 1.85/1.12 2.66 1.47
Shot 76 1.97 2.54 - 2.56 1.61
Shot 78 1.99 2.35 2.02 1.57 2.61 1.65
Shot 80 0.82/2.0/3.17 2.15/1.42/1.32 0.82/2.03/3.16 1.24/1.17/1.31 2.85 1.84
Shot 82 2.0 3.22 1.98/3.98 2.22/2.19 2.79 1.65
Note:
98
Shot 60 2.7 4.40 2.71 6.05 2.19 1.52
Shot 64 2.69 3.10 2.71 3.65 2.25 1.49
Shot 68 2.69 4.40 2.69 2.95 2.39 1.60
Shot 70 2.69 3.85 2.69 2.80 2.48 1.71
Shot 74 2.69 3.70 2.69 2.50 2.65 1.80
Shot 78 2.02/2.69 7.05/3.0 1.97/2.71 2.95/3.05 2.74 1.67
Shot 80 2.0/2.69 5.60/4.0 2.0/2.69 2.20/3.60 2.75 1.86
Shot 82 2.70 2.90 2.69 2.80 2.80 1.82
Note:

Shot 60 2.67 2.83 2.69 2.75 2.32 1.33
Shot 64 2.69 5.49 2.69 6.06 2.39 1.46
Shot 68 2.69 7.83 2.71 2.88 2.55 1.56
Shot 70 2.70 3.35 2.76 2.78 2.59 1.47
Shot 74 2.69 3.27 2.73 3.18 2.55 1.62
Shot 78 2.0/2.70 3.6/3.05 2.01/2.64 1.09/2.64 2.54 1.71
Shot 80 2.71 4.11 2.70 4.73 2.41 1.80
Shot 82 2.01/2.70 2.65/2.85 1.94/2.69 1.96/2.23 2.60 1.84
Note:
99
Shot 60 1.25/2.71 2.95/4.27 2.68 4.44 2.30 1.32
Shot 64 2.75 8.0 2.70 3.82 2.49 1.34
Shot 68 2.76 4.02 2.69 3.54 2.61 1.43
Shot 70 1.22/2.75 6.84/6.36 1.35/2.69 3.50/5.10 2.67 1.52
Shot 74 1.23/2.67 2.49/2.51 2.75 1.55
Shot 78 1.16/2.73/3.12 4.92/3.48/3.44 1.16/2.69/3.16 3.01/4.02/- 2.83 1.64
Shot 80 2.70 3.02 2.68 3.48 2.79 1.55
Shot 82 2.69 2.39 2.69 2.96 2.91 1.63
Note:
Experiment Date. Feb. 18, 2003

Shot 60 2.64 1.51 2.57 3.52 2.04 1.59
Shot 64 2.71 7.68 2.71 6.43 2.26 1.41
Shot 68 2.69 4.24 2.71 4.92 2.49 1.53
Shot 70 2.69 5.13 2.70 5.83 2.55 1.46
Shot 74 2.69 4.04 2.73 3.62 2.67 1.61
Shot 78 2.69 3.09 2.71 3.57 2.72 1.61
Shot 80 2.0/2.69 1.57/3.15 2.01/2.72 1.71/3.94 2.56 1.80
Shot 82 2.69 3.50 2.70 4.47 2.68 1.70
Note:
100
Shot 60 2.71 5.33 2.69 4.54 2.32 1.34
Shot 64 2.67 5.08 2.68 4.63 2.38 1.21
Shot 68 2.67 2.83 2.67 2.42 2.65 1.51
Shot 70 2.68 3.74 2.68 3.74 2.65 1.43
Shot 74 2.05/2.69 2.87/2.53 2.06/2.84 2.17/2.44 2.75 1.55
Shot 78 2.66 3.30 2.67 3.50 2.77 1.50
Shot 80 2.06/2.69 3.32/2.84 1.92/2.68 1.96/2.68 2.83 1.58
Shot 82 2.68 2.87 2.69 2.81 2.87 1.59
Note:

Shot 60 2.69 4.70 2.70 6.24 2.27 1.36
Shot 64 2.69 4.60 2.76 7.16 2.46 1.31
Shot 68 2.69 5.80 2.78 6.38 2.57 1.43
Shot 70 2.69/2.83 5.08/3.52 2.70/2.83 4.97/5.06 2.69 1.52
Shot 74 2.69/2.83 6.16/2.41 2.70/2.83 4.68/4.94 2.73 1.55
Shot 78 2.0/2.69/2.84 2.83/4.04/4.0 1.90/1.99/2.76 3.16/2.78/4.67 2.72 1.47
Shot 80 2.69/2.84 5.79/4.24 2.70/2.84 4.53/5.0 2.79 1.64
Shot 82 2.0/2.69/2.79 2.48/5.69/2.68 2.0/2.71/2.81 1.06/5.62/5.18 2.68 1.69
Note:
Emitter Width. 1.375", Emitter Position. Di=2", Stiip Reflector Position. D2=4 cm
101
Shot 60 2.75 6.11 2.75 2.89 2.29 1.28
Shot 64 2.70 4.99 2.68 2.67 2.53 1.35
Shot 68 2.69 4.33 2.69 2.81 2.56 1.43
Shot 70 2.03/2.69 2.56/2.85 1.91/2.84 1.56/2.22 2.67 1.47
Shot 74 2.05/2.69 6.15/3.37 1.93/2.68 2.62/2.40 2.76 1.56
Shot 78 2.07/2.76 10.2/3.07 1.99/- 4.63/- 2.88 1.61
Shot 80 2.05 11.5 2.06 3.84 2.84 1.55
Shot 82 2.07 13.2 2.07 6.27 2.89 1.56
Note:

Shot 60 2.69 7.86 2.70 10.4 2.18 1.28
Shot 64 2.69/2.76 2.76/3.37 2.76 8.23 2.45 1.35
Shot 68 2.76 4.20 2.76 8.92 2.49 1.29
Shot 70 2.69 2.90 2.70 5.47 2.61 1.42
Shot 74 2.69 4.47 2.69 7.36 2.67 1.50
Shot 78 2.0/2.84 4.03/3.47 2.0/2.69 4.82/6.06 2.81 1.55
Shot 80 2.0/2.69/2.84 1.23/3.87/3.57 2.69 7.16 2.79 1.60
Shot 82 2.0/2.69/2.84 3.03/2.62/4.19 2.02/2.69/2.84 2.07/5.04/3.78 2.79 1.70
Note:
Emitter Width. 1.375", Emitter Position. Di=6 cm, Stiip Reflector Position. D2=5.08 cm.
102
Table A.27 . Frequency Record 27 in Cavity Effect Research at TTU
Shot 60 2.76 8.15 2.76 4.13 2.36 1.24
Shot 64 2.77 5.64 2.69 3.38 2.47 1.26
Shot 68 2.69 4.28 2.68 3.46 2.59 1.35
Shot 70 1.90/2.76 3.86/3.60 2.0/- 3.86/- 2.76 1.44
Shot 74 2.03/2.76 5.42/2.82 1.99/- 5.94/- 2.79 1.52
Shot 78 2.07 13.1 2.05 7.61 2.84 1.56
Shot 80 2.03 9.34 1.98 4.96 2.83 1.56
Shot 82 2.01 12.4 2.01 7.61 2.75 1.61
Note:
Emitter Widtii. 1.375", Emitter Position. Di=6 cm, Donut Reflector Position. D2=6.08 cm.

Shot 60 2.76 3.94 2.76 8.92 2.42 1.23
Shot 64 2.69 3.73 2.73 5.11 2.49 1.28
Shot 68 2.75 12.1 2.69 7.11 2.55 1.33
Shot 70 2.69 3.58 2.69 5.90 2.61 1.41
Shot 74 2.69 4.17 2.69 4.67 2.71 1.47
Shot 78 2.0/2.84 2.74/3.44 2.69 4.84 2.75 1.61
Shot 80 2.0/2.69 3.53/2.86 1.95/2.69 2.66/5.73 2.84 1.50
Shot 82 2.84 3.93 2.69 6.32 2.79 1.53
Note:
Emitter Width. 1.375", Emitter Position. Di=6 cm, Stiip Reflector Position. D2=6.08 cm.
103
Shot 60 2.65 2.90 2.61 4.30 2.34 0.88
Shot 64 2.65 2.70 2.62 3.30 2.74 2.19
Shot 68 1.93/2.67 4.90/3.70 1.93/2.67 3.60/5.30 2.89 3.60
Shot 70 1.92/2.68 2.25/4.95 2.69 7.20 3.00 3.56
Shot 74 2.66 4.80 2.71 4.05 3.08 3.15
Shot 78 1.99/2.65 3.80/3.90 2.69 4.20 3.18 4.18
Shot 80 2.65 4.05 2.71 4.40 3.21 4.44
Shot 82 2.01/2.68 4.60/3.96 2.69 5.03 3.26 4.45
Note:
Emitter Widtii. 1.375", Emitter Position. Di=6 cm, Stiip Reflector Position. D2=6.08 cm.

Shot 60 2.57 3.39 2.57 4.20 2.22 1.48
Shot 64 2.66 10.9 2.64 4.41 2.39 1.32
Shot 68 2.68 8.66 2.69 5.61 2.52 1.57
Shot 70 2.69 8.35 2.70 11.2 2.56 1.54
Shot 74 2.69 4.70 2.70 5.73 2.58 1.63
Shot 78 2.69 6.54 2.70 9.40 2.75 1.71
Shot 80 2.69 5.85 2.63 4.87 2.61 1.35
Shot 82 2.57 5.07 2.63 5.03 2.30 1.57
Note:
104
Shot 60 2.61 3.58 2.61 3.94 2.28 1.33
Shot 64 1.92/2.64 2.12/4.87 2.63 5.88 2.37 1.22
Shot 68 2.67 6.16 2.63 4.77 2.55 1.46
Shot 70 2.69 7.73 2.69 7.73 2.67 1.42
Shot 74 2.69 6.95 2.69 7.79 2.68 1.48
Shot 78 2.68 6.50 2.70 7.70 2.76 1.64
Shot 80 2.69 7.66 2.71 8.06 2.75 1.67
Shot 82 2.68 6.63 2.70 5.56 2.79 1.58
Note:
Emitter Widtii. 1.375", Emitter Position. Di=7 cm, Stiip Reflector Position. D2=4.08 cm
Table A.32. Frequency Record 32 in Cavit>' Effect Research at TTU

Shot 60 2.87 0.58 2.92 0.84 2.41 1.22
Shot 64 2.96 0.96 2.96 1.09 2.60 1.33
Shot 68 2.93 0.80 2.94/3.10 0.66/0.72 2.76 1.36
Shot 70 3.08 1.04 3.07 1.02 2.82 1.44
Shot 74 3.05 0.53 3.05 1.70 2.77 1.47
Shot 78 2.24 1.80 2.18 2.30 2.68 1.73
Shot 80 2.07/3.04 0.42/0.38 2.18/3.08 0.52/0.70 2.87 1.82
Shot 82 3.08 0.66 3.07 1.30 2.96 1.54
Note:
Experimental Date. March 11,2003
Emitter Widtii. 1 cm. Emitter Position. Di=4 cm. No Reflector.
105
Freq H E-Probe Peaks V E-Probe Peaks Norm. Norm. I
Band V
Freq. (GHz) Amp. (A.U.) Freq. (GHz) Amp. (A.U) (A. U.) (A. U.)
Shot 60 - - . . 2.83 0.86
Shot 64 - - . _ 2.96 0.94
Shot 68 - - - _ 3.0 1.12
Shot 70 - - - •
3.10 1.19
Shot 74 1.36/2.90 0.56/0.48 1.36/2.86 0.32/0.32 2.71 1.64
Shot 78 1.39/2.09/2.92 0.76/1.04/0.57 1.46/2.11/2.86 0.5/0.84/0.54 2.59 1.85
Shot 80 1.41/2.86 0.72/0.63 1.46/2.27/2.96 0.54/0.59/0.58 2.60 1.75
Shot 82 1.39/1.67/2.91 0.64/0.77/0.63 1.36/1.64/2.90 0.55/0.59/0.44 2.71 1.77
Note:
Emitter Width. 1 cm. Emitter Position. Di=4 cm, Donut Reflector Position. D2=3 cm

Shot 60 - - - 2.66 0.90
Shot 64 - - - - 2.77 0.97
Shot 68 2.76 0.1 2.71 0.07 2.48 1.39
Shot 70 - - - - 2.64 1.33
Shot 74 - - - 2.79 1.39
Shot 78 2.95 0.08 3.02 0.07 2.86 1.50
Shot 80 - - 2.97 1.60
Shot 82 - - - - 3.10 1.14
Note:
Emitter Width. 1 cm. Emitter Position. Di=4 cm, Stiip Reflector Position. D2=3 cm.
106
Shot 60 2.59 1.08
Shot 64 2.67 1.11
Shot 68 1.51 0.36 1.31 0.28 2.71 1.30
Shot 70 1.36 0.36 2.80 1.29
Shot 74 2.95 0.31 1.45/2.95 0.24/0.36 2.84 1.43
Shot 78 1.49 0.60 1.46 0.76 2.93 1.40
Shot 80 2.90 0.37 2.95 0.31 3.02 1.46
Shot 82 1.37/2.93 -/0.87 1.37/2.96 0.23/0.63 2.79 1.70
Note:
Emitter Width. 1 cm, Emitter Position. Di=4 cm, Donut Reflector Position. D2=4 cm.

Band Freq. (GHz) Amp. (A.U.) Freq. (GHz) Amp. (A.U) (A.U.) (A. U.)
Shot 60 - - - 2.43 1.14
Shot 64 - - 2.83 0.99
Shot 68 - - - 2.85 1.02
Shot 70 2.69 0.09 2.94 0.13 2.69 1.52
Shot 74 - - - - 2.91 1.20
Shot 78 1.67/3.04 0.33/0.08 1.64/3.08 0.68/0.11 3.06 1.38
Shot 80 - - - - 3.08 1.43
Shot 82 1.54/3.03 0.42/0.15 1.58/3.07 0.36/0.22 2.78 1.78
Note:
Emitter Width. 1 cm. Emitter Position. Di=4 cm. Strip Reflector Position. D2=4 cm
107
Shot 60 2.69/2.82 0.23/0.22 2.69 0.20 2.28 1.32
Shot 64 - - 2.69/2.88 0.21/0.39 2.60 1.30
Shot 68 2.92 0.23 2.91 0.36 2.71 1.14
Shot 70 2.91 0.32 2.69/2.97 0.30/0.37 2.97 1.15
Shot 74 1.47/2.99 0.49/0.36 1.58/3.01 0.37/0.32 2.67 1.59
Shot 78 1.67/2.09/2.95 0.78/1.38/- 1.67/2.07/2.93 0.80/0.96/0.40 2.52 1.91
Shot 80 1.67/2.07/2.93 0.64/1.31/0.62 1.67/2.07/2.92 0.55/1.04/0.93 2.62 1.77
Shot 82 2.07/2.93 0.30/0.47 2.05/2.94 0.55/0.78 2.73 1.77
Note:
Emitter Width. 1 cm. Emitter Position. Di=4 cm, Donut Reflector Position. D2=5 cm.

Band Freq. (GHz) Amp. (A.U.) Freq. (GHz) Amp. (A.U) (A.U.) (A. U.)
Shot 60 1.36/2.70 0.72/0.14 2.70 0.1 2.37 1.27
Shot 64 - 2.72 1.13
Shot 68 - - - - 2.97 0.98
Shot 70 - ~ 2.98 1.05
Shot 74 1.59/3.13 0.38/0.1 3.06 1.31
Shot 78 1.67/3.21 0.67/0.11 1.64/3.16 0.65/0.1 2.93 1.50
Shot 80 - - - 3.0 1.23
Shot 82 1.63 1.5 1.64/3.3 0.92/0.22 3.08 1.55
Note:
Emitter Width. 1 cm. Emitter Position. Di=4 cm, Stiip Reflector Position. D2=5 cm.
108
Shot 60 1.92/2.78 0.23/0.40 -/2.76 -/0.48 2.26 1.40
Shot 64 2.67/2.95 0.29/0.51 2.69/2.94 0.40/0.31 2.55 1.32
Shot 68 2.69/2.95 0.27/0.34 2.89 0.52 2.72 1.34
Shot 70 2.85 0.27 2.81 0.24 2.71 1.17
Shot 74 2.69 0.37 2.71 0.34 2.95 1.36
Shot 78 2.68/2.96 0.23/0.72 3.02 0.48 2.91 1.65
Shot 80 2.71/2.98 0.26/0.31 2.69/3.02 0.33/0.48 2.93 1.29
Shot 82 2.70/3.0 0.23/0.62 2.69/3.0 0.23/0.38 2.99 1.60
Note:

Shot 60 2.70 0.09 2.74 0.07 2.71 0.93
Shot 64 - - - 2.80 1.04
Shot 68 2.98 0.13 2.90 0.14 2.99 1.16
Shot 70 3.02 0.17 3.02 0.16 2.92 1.29
Shot 74 2.89 0.25 2.89 0.22 2.98 1.09
Shot 78 - - 1.59/2.99 0.19/0.11 3.03 1.28
Shot 80 2.91 0.1 2.98 0.2 3.17 1.17
Shot 82 1.63 0.67 1.59 0.44 3.02 1.39
Note:
Emitter Width. 1 cm. Emitter Position. Di=4 cm, Stiip Reflector Position. D2=6 cm
109
Shot 60 2.91 1.45 2.91 2.20 2.42 1.13
Shot 64 2.98 2.98 2.97 4.50 2.68 1.24
Shot 68 3.05 2.45 3.06 4.90 2.84 1.32
Shot 70 3.13 2.65 3.12 3.95 2.91 1.40
Shot 74 1.91/3.09 1.75/2.65 -/3.08 5.50 2.94 1.41
Shot 78 1.82/3.13 2.0/1.85 3.13 4.40 3.0 1.87
Shot 80 3.14 3.75 3.16 7.20 3.14 1.96
Shot 82 3.09 5.40 3.07 3.10 3.14 1.71
Note:
Emitter Width. 1 cm. Emitter Position. Di=5 cm. No Reflector

Shot 60 2.86 0.65 2.86 1.79
Shot 64 2.69/2.98 0.73/1.77 2.69/3.04 0.54/1.85 2.95 1.03
Shot 68 2.69/3.08 -/1.95 2.69/3.08 0.35/1.28 3.12 1.07
Shot 70 1.81/3.07 2.14/1.52 1.79/3.07 0.53/2.24 3.02 1.06
Shot 74 1.82/3.07 0.47/1.83 3.04 1.59 3.26 1.13
Shot 78 1.90/3.08 1.24/2.35 3.07 2.40 3.45 1.12
Shot 80 1.90/3.06 3.24/2.23 1.90/3.04 1.44/3.34 3.40 1.21
Shot 82 2.93 1.24 2.93 1.15 2.90 1.59
Note:
110
Shot 60 2.89 1.49 2.87 1.64 2.52 1.11
Shot 64 3.05 1.38 3.06 1.15 2.88 1.06
Shot 68 1.83/3.04 0.85/1.67 1.83/3.07 0.44/1.28 2.94 1.10
Shot 70 3.07 1.29 3.07 2.24 2.95 1.20
Shot 74 1.90/3.09 4.64/1.28 1.91/3.08 1.56/2.38 3.08 1.13
Shot 78 1.82/3.13 1.89/1.14 1.84/3.13 1.15/2.04 3.17 1.38
Shot 80 1.90/3.13 3.92/1.45 1.87/3.13 1.18/2.91 3.28 1.37
Shot 82 1.90/3.82 5.78/1.14 1.91/3.81 2.23/1.23 3.37 1.17
Note:
Emitter Widtii. 1 cm. Emitter Position. Di=5 cm, Stiip Reflector Position. D2=3 cm.

Shot 60 2.83 0.77 2.83 0.81 2.51 1.17
Shot 64 1.74/2.96 1.08/2.29 1.73/2.96 0.85/1.30 2.78 1.04
Shot 68 3.06 3.34 3.06 1.78 3.10 1.18
Shot 70 2.99 2.08 2.97 2.31 3.11 1.03
Shot 74 1.82/3.05 1.89/2.78 3.06 2.97 3.22 1.23
Shot 78 3.06 0.70 3.01 0.78 3.18 0.97
Shot 80 1.91/3.04 2.06/1.38 1.89/3.07 1.61/1.45 3.21 1.09
Shot 82 3.0 3.38 3.01 1.85 3.10 1.53
Note:
Ill
Shot 60 2.86 0.52 2.86 0.65 2.59 0.98
Shot 64 3.06 1.03 2.98 1.71 2.86 1.07
Shot 68 1.82/2.99 0.94/2.07 1.85/2.98 0.43/1.86 2.93 1.12
Shot 70 1.84/3.07 1.51/1.36 1.87/3.06 1.78/1.69 3.06 1.12
Shot 74 1.91/2.98 2.09/0.87 1.91/3.16 4.0/1.72 3.12 1.06
Shot 78 1.92/3.08 4.34/2.12 1.92/3.07 5.13/5.74 3.22 1.37
Shot 80 1.90/3.06 3.09/2.48 1.93/3.07 2.67/3.09 3.18 1.21
Shot 82 1.92/3.08 5.81/4.84 1.91/3.07 3.10/3.26 3.27 1.36
Note:
Emitter Width. 1 cm. Emitter Position. Di=5 cm. Strip Reflector Position. D2=4 cm.

Shot 60 2.89 2.02 2.89 1.04 2.66 1.14
Shot 64 2.98 4.24 2.97 2.19 2.87 1.02
Shot 68 2.93 1.64 2.95 2.12 2.72 1.30
Shot 70 2.94 2.27 2.96 1.49 2.78 1.49
Shot 74 2.99 2.71 2.95/2.99 1.42 2.75 1.65
Shot 78 2.94/2.99 1.24/1.19 3.01 0.87 2.70 1.66
Shot 80 2.95 2.41 2.97 2.14 2.81 1.60
Shot 82 2.95/2.99 1.92/2.28 2.99 1.56 2.88 1.63
Note:
Emitter Widtii. 1 cm. Emitter Position. Di=5 cm, Donut Reflector Position. D2=5 cm
112
Shot 60 2.91 1.14 2.90 1.69 2.66 1.02
Shot 64 1.82/2.99 6.06/1.67 1.85/2.98 3.72/2.38 2.99 1.11
Shot 68 1.83/2.93/3.70 6.24/1.11/2.32 1.84/2.94/3.67 11.3/1.02/1.50 2.90 0.89
Shot 70 1.92/2.98 2.06/0.84 1.91/2.97 1.94/2.25 2.87 1.45
Shot 74 1.91/2.99/3.80 3.63/1.72/1.78 1.91/2.99/3.78 6.62/3.75/1.45 3.13 1.38
Shot 78 1.92/2.99/3.80 5.08/1.09/2.51 1.83/3.02/3.78 9.03/1.82/2.96 3.26 1.25
Shot 80 1.84/2.99/3.69 4.44/1.83/0.82 1.85/3.06/3.78 4.11/2.42/0.58 3.19 1.56
Shot 82 1.87/3.02/3.76 6.84/1.37/1.55 1.87/3.02/3.78 8.12/2.70/1.22 3.22 1.48
Note:
Emitter Width. 1 cm. Emitter Position. Di=5 cm. Strip Reflector Position. D2=5 cm

Shot 60 2.84 1.24 2.81 0.97 2.45 1.06
Shot 64 2.97 2.62 2.96 2.82 2.87 1.07
Shot 68 2.96 3.02 2.89/2.96 2.08/1.59 2.83 1.02
Shot 70 2.98 3.11 2.99 3.24 3.06 1.28
Shot 74 2.99 3.19 3.08 5.96 3.19 1.30
Shot 78 2.99 3.15 2.98/3.07 3.87/4.46 3.22 1.33
Shot 80 2.99/3.08 3.28/2.05 3.07 5.83 3.21 1.39
Shot 82 2.98 2.94 2.96 5.13 3.10 1.14
Note:
113
Shot 60 2.97 1.32 2.89 1.42 2.72 1.09
Shot 64 1.82/2.95/3.66 7.13/1.50/5.04 1.82/2.95/3.63 13.9/2.52/4.87 2.97 0.92
Shot 68 1.83/2.97 2.89/4.26 1.84/2.96 2.78/2.37 2.99 1.11
Shot 70 1.83/2.96/3.67 7.13/1.43/2.34 1.83/2.93/3.68 14.9/1.76/1.64 3.16 1.09
Shot 74 1.83/2.96/3.68 8.75/1.25/2.07 1.83/2.93/3.68 11.8/1.66/2.43 2.99 1.17
Shot 78 1.83/3.70 11.5/1.80 1.84/3.68 15.8/2.78 3.16 1.25
Shot 80 1.84/2.97/3.70 10.3/1.60/3.28 1.84/3.69 14.2/2.21 3.13 1.34
Shot 82 1.84/2.98/3.70 8.95/1.90/3.70 1.85/3.72 12.8/2.56 3.26 1.42
Note:
Emitter Widtii. 1 cm. Emitter Position. Di=5 cm. Strip Reflector Position. D2=6 cm

Shot 60 - 2.80 2.05 2.4 1.27
Shot 64 2.89 2.80 2.80 2.05 2.8 1.13
Shot 68 2.96 3.20 2.94 2.45 3.02 1.2
Shot 70 1.67/2.73 2.90/1.95 1.67/2.86 3.15/1.75 2.78 1.42
Shot 74 1.82/2.17/2.93 6.50/2.70/2.50 1.82/2.18/2.93 3.90/3.10/2.70 3.0 1.33
Shot 78 1.82/2.94 6.05/- 1.84/2.92 4.0/2.80 3.07 1.53
Shot 80 1.82/2.92 1.55/3.20 1.73/2.91 1.80/4.05 3.0 1.54
Shot 82 1.82/2.92 6.70/2.20 1.82/2.92 6.0/3.40 3.12 1.59
Note:
Emitter Width. 1 cm. Emitter Position. Di=6 cm, No Reflector
114
Shot 60 2.41/3.96 -/2.80 2.38/2.41/3.97 -/-/4.05 2.54 1.22
Shot 64 2.69/2.88 0.65/0.78 2.72/2.89 0.55/0.76 2.87 1.14
Shot 68 2.69/3.74 -/1.42 2.69/3.74 -/1.64 2.91 1.05
Shot 70 2.93/3.64 0.92/1.06 2.87/3.66 0.88/1.15 3.08 1.26
Shot 74 3.56/3.61 5.09/5.47 3.57/3.61 6.40/5.68 3.19 1.30
Shot 78 3.56 6.11 3.55/3.61 5.84/6.35 3.30 1.24
Shot 80 3.58 6.58 3.59 6.72 3.28 1.37
Shot 82 3.62 3.25 3.58 2.76 3.29 1.39
Note:

Freq H E-Probe Peaks V. E-Probe Peaks Norm. V Norm. I
Band Freq. (GHz) Amp. (A.U.) Freq. (GHz) Amp. (A.U) (A.U.) (AU.)
Shot 60 2.72 1.31 2.73 0.98 2.53 1.14
Shot 64 2.86 1.03 2.86 1.62 2.85 1.15
Shot 68 2.89 1.87 2.89 1.01 3.03 1.18
Shot 70 2.84 1.07 2.86 0.72 2.65 1.20
Shot 74 2.69 0.97 2.89 2.15 2.95 1.33
Shot 78 2.96 2.17 2.90 2.25 3.08 1.28
Shot 80 2.99 1.10 2.90 1.12 3.23 1.28
Shot 82 1.83/2.86 1.33/1.31 2.91 1.24 2.91 1.64
Note:
115
Band Freq. (GHz) Amp. (A.U.) Freq. (GHz) (A. U.)
Amp. (A.U) (A. U.)
Shot 60 2.79 1.80 2.79 1.65 2.61 1.07
Shot 64 2.89 1.19 2.89 1.46 2.91 0.97
Shot 68 2.95 1.65 2.92 2.34 3.04 1.13
Shot 70 2.96/3.61/3.68 2.93/-/- 2.96/3.68 2.71/1.15 3.15 1.01
Shot 74 2.93 1.97 2.93 1.46 3.13 1.14
Shot 78 2.96 2.29 2.97 3.04 3.28 1.24
Shot 80 2.98/3.64 2.06/0.90 2.98/3.66 2.60/0.74 3.32 1.31
Shot 82 2.95/3.62 1.71/- 2.96/3.62 1.72/- 3.25 1.12
Note:

Shot 60 2.63 1.06 2.61 1.20 2.54 1.23
Shot 64 1.8/2.89/3.69 1.89/1.09/0.73 1.83/2.87/3.67 8.0/1.39/0.96 2.92 1.04
Shot 68 1.82/2.96/3.68 2.50/0.81/2.08 1.82/-/3.67 10.8/-/2.01 3.06 1.07
Shot 70 1.9/2.89/3.80 6.56/1.37/4.31 1.82/2.88/3.67 10.4/1.19/3.66 3.04 1.16
Shot 74 1.89/2.92/3.72 4.27/1.0/3.30 1.82/2.90/3.67 10.3/0.92/3.72 3.13 1.15
Shot 78 1.89/2.99/3.80 5.25/0.80/4.59 1.83/3.67 17.2/4.66 3.20 1.14
Shot 80 1.92/2.91/3.85 5.51/1.32/6.09 1.92/2.91/3.85 7.86/1.42/6.12 3.17 1.37
Shot 82 1.82/2.83/3.68 3.77/1.02/1.46 1.84/2.91/3.76 5.95/1.52/1.18 2.95 1.67
Note:
116
Shot 60 2.79 3.15 2.79 1.63 2.64 1.09
Shot 64 2.83 2.09 2.86 2.46 2.94 1.02
Shot 68 2.89 2.02 2.90 2.67 3.07 1.07
Shot 70 2.93 2.22 2.93 2.52 3.25 1.14
Shot 74 2.97 1.57 2.88 2.02 3.30 1.19
Shot 78 2.97 2.96 2.97 4.73 3.46 1.22
Shot 80 2.96 1.54 2.93 2.50 3.36 1.26
Shot 82 1.90/2.93 1.21/1.90 2.93 2.29 3.40 1.33
Note

Shot 60 2.72/3.54 1.73/0.84 1.73/2.59/3.53 1.46/1.31/0.97 2.51 1.19
Shot 64 1.82/3.64 14.0/16.7 1.82/3.64 15.2/19.5 2.78 0.99
Shot 68 1.82/3.64 8.83/16.2 1.82/3.64 18.8/20.9 3.0 1.03
Shot 70 1.81/3.68 5.46/17.1 1.83/3.62 17.9/12.5 3.0 1.09
Shot 74 1.83/3.69 6.66/14.5 1.84/3.68 16.2/9.68 2.92 1.43
Shot 78 1.82/3.68 11.6/14.2 1.84/3.68 22.4/8.61 3.27 1.31
Shot 80 1.86/3.70 6.71/15.4 1.85/3.70 16.4/6.88 3.98 1.56
Shot 82 1.82/3.69 8.75/19.4 1.85/3.70 17.6/9.39 3.12 1.50
Note:
Expermient Date. March 19,2003
Emitter Widtii. 1 cm. Emitter Position. Di==6, Stiip Reflector Position. D2=5 cm.
117
Shot 60 2.77 1.65 2.79 1.97 2.75 1.0
Shot 64 2.79 2.48 2.79 1.24 2.84 1.04
Shot 68 2.82 2.92 2.81 3.85 2.98 1.22
Shot 70 - - - _ 2.94 1.32
Shot 74 2.82 2.50 2.81 5.50 2.96 1.32
Shot 78 2.84 1.90 2.81 2.58 2.98 1.51
Shot 80 2.82 3.28 . _ 2.99 1.41
Shot 82 2.85 1.50 2.83 2.51 3.10 1.56
Note

Band Freq. (GHz) Amp. (A.U.) Freq. (GHz) Amp. (A.U) (A. U.) (AU.)
Shot 60 1.67/2.70 2.08/1.19 1.67/2.69 3.83/1.21 2.37 1.40
Shot 64 1.75/2.79/3.52 3.89/1.50/3.05 1.76/2.53 6.70/3.23 2.15 1.19
Shot 68 1.78/3.55 3.23/3.23 1.78/3.56 11.7/1.62 2.86 1.19
Shot 70 1.77/3.55 3.53/2.55 1.79/3.57 14.4/2.43 2.89 1.31
Shot 74 1.78/3.55 6.35/5.30 1.78/3.55 13.3/3.11 2.98 1.24
Shot 78 1.84/3.64 5.70/8.09 1.79/3.64 12.1/7.98 3.06 1.36
Shot 80 1.83/3.67 7.36/3.23 1.83/3.62 14.4/4.01 2.98 1.47
Shot 82 1.83/3.64 6.45/3.55 1.83/3.62 15.8/3.75 3.05 1.52
Note:
Emitter Width. 1 cm, Emitter Position. Di=6 cm, Stiip Reflector Position. D2=6 cm
118
Table A.59. Frequency Record 59 ui Cavity Effect Research at TTU
Shot 60 2.45 2.70 2.53 1.98 2.38 0.88
Shot 64 2.57 7.50 2.54 6.40 2.76 2.20
Shot 68 2.56 3.70 2.60 4.50 2.91 3.60
Shot 70 2.54/2.62 4.55/4.80 2.54/2.62 4.80/5.55 3.00 3.60
Shot 74 2.58/2.63 6.50/7.0 2.54 5.50 3.10 3.17
Shot 78 2.54 4.70 2.59 6.60 3.20 4.40
Shot 80 1.90/2.60 3.20/4.30 2.59 6.05 3.23 4.45
Shot 82 1.92/2.70 3.30/6.40 2.62 5.50 3.26 4.44
Note:
Emitter Width. 1 cm. Emitter Position. Di=7 cm. No Reflector
119

Investigation of A High-Power Virtual Cathode Oscillator

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Investigation of A High-Power Virtual Cathode Oscillator

Uploaded by

Copyright:

Available Formats

INVESTIGATION OF A HIGH-POWER

VIRTUAL CATHODE OSCILLATOR

XUPENG CHEN, B.E., M.S.

Submitted to the Graduate Faculty

Dean of the Graduate School

I would like to express my deepest appreciation to Dr. J. Dickens, Dr. L. L.

repairing the coaxial VIRCATOR.

I am gratefiil to Dr. W. Jiang for his introduction to this project in my start-up

had troubles during this project.

James Chen, for his bringing a lot of enjoyment to my life.

This work was supported by AFOSR and FMV, Sweden.

2. SPACE-CHARGE LIMITED CURRENT FOR COAXL\L DIODES 4

2.2 Poisson's Equation for 1-D Cylindrical Diodes 5

2.3 Poisson's Equation for 1-D Cylindrical Diodes Simplification and

2.4 Comparison of Langmuir-Blodgett Law with Our Approximation

2.5 Relativity Corrected Space-Charge Limited Current for the

2.6 Space-Charge Limited Current for 2-D Coaxial Diodes 24

3. APPLICATION OF ANALYTICAL CURRENT EXPRESSION IN

3.1 Current Evaluation of Coaxial Diodes 30

3.2 Virtual Cathode Position Determination and Virtual Cathode

4. EXPERIMENTAL CONFIGURATIONS AND VIRTUAL CAVITY 36

4.1 Experimental Configurations 38

4.3 Experimental Results 43

4.4 Cavity Resonance Effect and Virtual Cavity 46

4.5 The Function of the Reflectors 48

5. PROPAGATING MODE DETERMINATION 54

5.1 Cutoff Frequencies of the Waveguide 54

5.2 Field Distribution for Different Modes in a Circular Waveguide 56

5.3 Wave Combinations 56

5.4 Theoretical Analysis for Microwave Modes 60

5.5 Analysis of Experimental Data 62

5.6 Discussions on Propagatmg Modes 80

6. CONCLUSIONS AND SUGGESTIONS 84

In this dissertation, three topics are investigated in different degrees: space-charged

diodes and microwave propagating mode identification.

By introducing a physics approximation method into analyzing the nonlinear

Poisson's equation, an analytical expression for the space-charge-limited current density

1-D cylindrical diodes. This 1-D result is extended to 2-D.

formation. Some possible explanations are proposed.

suggestions are proposed to identify microwave propagating modes.

3.1. Virtual Cathode Position Predictionfi-omDifferent Models 32

4.1. Examples of Frequency Record in Cavity Effect Research of a Coaxial

4.2. Typical Experimental Records for a Certain Diode Geometiy 45

4.3. An Example Experunental Records for a Certain Diode Geometiy 45

4.4. Frequency List for Geometries without Reflectors 46

4.5. Frequency List for Geometries with a Donut Reflector 49

4.6. Frequency List for Geometries with a Strip Reflector 50

5.1. x„„ Array 55

5.2. x'„m Array 55

5.3. Cutoff Frequencies (GHz) for TMnm Mode 55

5.4. Cutoff Frequencies (GHz) for TEnm Mode 56

A.l. Frequency Record 1 in Cavity Effect Research at TTU 90

A.2. Frequency Record 2 in Cavity Effect Research at TTU 90

A.3. Frequency Record 3 in Cavity Effect Research at TTU 91

A.4. Frequency Record 4 in Cavity Effect Research at TTU 91

A.5. Frequency Record 5 in Cavity Effect Research at TTU 92

A.6. Frequency Record 6 in Cavity Effect Research at TTU 92

A.7. Frequency Record 7 in Cavity Effect Research at TTU 93

A.8. Frequency Record 8 in Cavity Effect Research at TTU 93

A.9. Frequency Record 9 in Cavity Effect Research at TTU 94

A.l 1. Frequency Record 11 in Cavity Effect Research at TTU 95

A.12. Frequency Record 12 in Cavity Effect Research at TTU 95

A.13. Frequency Record 13 in Cavity Effect Research at TTU 96