close

Вход

Забыли?

вход по аккаунту

?

High-impedance high-speed microwave detector

код для вставкиСкачать
INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI
films the text directly from the original or copy submitted. Thus, some
thesis and dissertation copies are in typewriter face, while others may be
from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and photographs, print bleedthrough, substandard margins,
and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note will indicate
the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand comer and
continuing from left to right in equal sections with small overlaps. Each
original is also photographed in one exposure and is included in reduced
form at the back of the book.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6” x 9” black and white
photographic prints are available for any photographs or illustrations
appearing in this copy for an additional charge. Contact UMI directly to
order.
UMI
A Bell & Howell Information Company
300 North Zeeb Road, Ann Arbor MI 48106-1346 USA
313/761-4700 800/521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
HIGH IMPEDANCE HIGH SPEED
MICROWAVE DETECTOR
by
Juan R. Luglio B.S., M.S.
A dissertation submitted to the faculty of the
Graduate School, Marquette University, in
Partial Fulfillment of the Requirements for
the Degree of Doctor of Philosophy
Milwaukee, Wisconsin
December 1997
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 9825394
Copyright 1998 by
Luglio, Juan Raimundo
All rights reserved.
UMI Microform 9825394
Copyright 1998, by UMI Company. All rights reserved.
This microform edition is protected against unauthorized
copying under Title 17, United States Code.
UMI
300 North Zeeb Road
Ann Arbor, MI 48103
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
To my parents, Angel and Lois Luglio.
They are my first teachers. Their encouragement support and love are what
made the completion of this dissertation possible.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Preface
Commercially available detectors are matched to the characteristic impedance of
the system in which they are intended to be used and thus are low impedance. The work in
this dissertation was motivated by a need for a high impedance detector to measure
microwave phase velocities as part of research being conducted by Dr. Thomas K. Ishii.
The high impedance detector that resulted from this work has many possible
applications that go beyond that of the original motivating need. A high impedance
detector is desirable any time a microwave or RF signal needs to be monitored and there is
a desire not to load the signal being monitored. This can be the case with automatic gain
control (AGC) circuits, signal strength monitoring circuits, standing wave ratio (SWR)
measuring circuits and any other microwave or RF monitoring circuit.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Acknowledgments
I would like to thank the members o f my dissertation committee, Dr. Ishii, Dr.
Richie, Dr. Jeutter, Dr. Seitz and Dr. Schlager for all of their time and effort in reviewing
this dissertation.
I would also like to thank all o f the students and faculty that have attended the
microwave engineering seminars on Friday afternoons at Marquette. These seminars are a
great place to discuss varied microwave related engineering topics and helped greatly in
developing this dissertation.
Recognition is necessary for Theodore Camenisch, Joseph Ratke, and Dr. Robert
Strangeway, engineers at the National Biomedical ESR Center, my place of employment.
Their support and being available to discuss engineering concepts was o f great help. The
Instruction and guidance of Richard Tonkyn and Richard Scherr, machinists at the ESR
Center, in the machining of parts necessary to complete this dissertation was also
appreciated.
The donation of sample FETs from California Eastern Laboratories (CEL) and the
help of Susumu Kodani, an applications engineer at CEL, also needs to be recognized.
Mike Larson at Hewlett Packard was instrumental in arranging the loan o f the test
equipment necessary to do the DC curve trace measurements o f the FETs.
The mechanical drawings of the TRL calibration kit shown in the appendix were
drawn by my father Angel Luglio. His drafting expertise was invaluable.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Abstract
The majority of today's microwave detectors are based on microwave diodes.
These detectors generally present a low impedance to the microwaves they are used to
detect. Input matching circuits and filtering, as well as parasitics can slow the detectors
response and require larger microwave power consumption. The research presented in this
dissertation develops the theory and optimized design procedure of a fast, sensitive, high
impedance detector based on a negatively biased microwave Field Effect Transistor (FET)
at 9GHz for the first time.
A microwave FET was chosen because, as an amplifier, it’s bandwidth is large
enough to handle the modulated microwave carrier frequency. The gate will be able to
follow the modulated microwaves and the drain will be able to follow the lower frequency
video output. The circuit is negative output detector, which means that the video output
voltage increases negatively as the microwave input increases.
The impedance of the FET gate circuit is low at the desired carrier frequency, but
the impedance o f the detector can be made large by the use of a transmission line of the
proper length. The detector developed has an input impedance of up to 2.9KQ.
It was found that the FET’s internal inter-electrode capacitances and electron
transit time through the channel limit the bandwidth and the output rise and fall times of
the detector as a result are limited. This problem is addressed by choosing a FET with
sufficient bandwidth.
R-C time constants of the external circuits of the detector will also limit the rise
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and fall time. Typically rise and fall times are thought to be equivalent, hi this detector
circuit the FET drain impedance changes with the application of microwaves to the gate
and it was found that the rise and fall times can be slightly different. The fall time o f the
circuit was measured and determined to be 2.053ns and the rise time 1.87ns on the
average. It was also found that these times varied with the microwave carrier frequency
used.
Low noise design o f this detector was investigated. As a result an equivalent TSS
5.9dB below that of a Schottky diode was measured.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table of Contents
P reface................................................................................................................. ii
Abstract ..................................................................................................................iv
Table of Contents................................................................................................... vi
List of Figures and T a b le s.....................................................................................ix
Chapter I - INTRODUCTION................................................................................ 1
Chapter II -THEORY DEVELOPMENT ..............................................................3
A. General Detector Theory........................................................................3
B. High Impedance Detector Theory..........................................................5
C. General Time Response Study ..............................................................S
D. Response to Pulse Modulated Microwaves........................................14
Chapter III - CIRCUIT DESIGN OPTIMIZATION AND FABRICATION . . . 18
A. Bias Point Determ ination.................................................................... 18
B. Microwave Circuit Design and Optimization..................................... 24
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vii
C. Predicted R esults...................................................................................50
D. Circuit Fabrication .............................................................................. 57
Chapter IV - EXPERIMENTAL MEASUREMENTS ........................................ 63
A. Detector efficiency .............................................................................. 63
B. Conversion Loss ...................................................................................76
C. Detection Speed - Rise and Fall T im e.................................................. 78
D. Sensitivity.............................................................................................83
E. Noise Figure .........................................................................................87
F. Bandw idth.............................................................................................94
G. Comparative response tests.................................................................. 98
Chapter V. COMPARISON OF THEORY AND RESULTS............................ 106
Chapter VI. CONCLUSIONS AND SUGGESTIONS FOR FURTHER
RESEARCH............................................................................................. 107
BIBLIOGRAPHY............................................................................................... 109
A ppendix............................................................................................................. I l l
TRL Calibration Kit Mechanical Drawings............................................ I l l
CNC Circuit Board D raw ings.................................................................124
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
viii
CNC Board Program ............................................................................... 126
CNC Cover Drawings ............................................................................. 134
CNC Cover Program ............................................................................... 136
FET Rise and Fall Time Data .................................................................145
Schottky Diode Rise and Fall Time Data .............................................. 157
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ix
List of Figures and Tables
Figure 1 50Q System with detector in parallel to the load......................................6
Figure 2 Diode detector c irc u it.............................................................................. 6
Figure 3 FET Detector .......................................................................................... 7
Figure 4 Common source FET amplifier equivalent circuit..................................9
Figure 5 Circuit for finding Z ................................................................................ 9
Figure 6 Circuit for finding short circuit current ................................................ 10
Figure 7 Periodic gate function............................................................................ 15
Figure 8 Spectrum of periodic gate function........................................................ 15
Figure 9 MDS simulation o f Pulsed puled microwave spectrum ........................17
Figure 10 Test fixture built for curve trace measurements.................................. 19
Figure 11 Measured IDvs. VGS at VDS= 2.02 V ..................................................20
Figure 12 CRV vs. VGS data ................................................................................ 20
Figure 13 Smoothed CRV vs. VGS data................................................................ 21
Figure 14 Self bias circuit.................................................................................... 22
Table 1 Bias points for maximum curvature........................................................23
Figure 15 TRL Calibration kit ............................................................................ 24
Figure 16 NE42484A S-parameters from 2.1 to 15.5 G H z..................................26
Figure 17 NE42484A S-parameters from 8.6 to 9.6 G H z....................................28
Figure 18 FET D SM .............................................................................................29
Figure 19 MDS circuit used to simulate the FET input impedance rotation . . . 30
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
X
Figure 20 Rotated impedance.............................................................................. 31
Figure 21 FET detector with radial stub shorts....................................................32
Figure 22 MDS circuit diagram for FET detector with radial stub shorts ..........33
Figure 23 FET stability circuit without radial stubs and transmission line ____39
Figure 24 Stability circles for FET A.................................................................... 40
Figure 25 FET stability circuit with radial stubs and transmission lin e ..............41
Figure 26 stability circles for FET with radial stubs and transmission line . . . . 42
Figure 27 Stability circles for FET without radial stubs and transmission line . 43
Figure 28 Load impedance circuit........................................................................ 44
Figure 29 Plot of the load impedance.................................................................. 46
Figure 30 MDS source impedance circuit............................................................. 48
Figure 31 Simulated source impedance ..............................................................49
Figure 32 MDS circuit used to simulate the ratio of input to gate voltage..........50
Figure 33 Simulated ratio of input to gate v o ltag e..............................................51
Figure 34 simulated loss of transmission line and detecto r................................ 54
Figure 35 Simulated loss o f Transmission lin e ....................................................55
Figure 36 MDS circuit diagram of the final detecto r..........................................58
Figure 37 Photograph of final detector and cover................................................59
Figure 38 Close up of circuit................................................................................ 60
Figure 39 Test setup for measuring efficiency....................................................64
Table 2 Data needed to calculate efficiency at 9.5GHz, for Pin= 12dBm
Table 3 Data needed to calculate efficiency at 11.1GHz, for Pin= OdBm
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
66
67
xi
Figure 40 Measured insertion loss o f 50Q microstrip l i n e ..................................70
Figure 41 Measured insertion loss of 50fl microstrip line and FET c irc u it
71
Figure 42 Comparison of 50Q line with and without FET circuit ..................... 72
Figure 43 MDS circuit used to determine detector input im pedance................. 73
Figure 44 MDS simulation of detector input impedance ....................................73
Figure 45 Efficiency at 9.5 G H z .......................................................................... 75
Figure 46 Efficency at 11.1 G H z .......................................................................... 76
Figure 47 Conversion loss at 9.5G H z.................................................................. 77
Figure 48 Conversion loss at 11.1GHz................................................................ 78
Table 4 Rise and fall times of FET detector ........................................................ 80
Figure 49 Rise and fall time plot for a carrier of 9.5GHz....................................81
Figure 50 Rise and fall time data showing ringing..............................................82
Figure 51 Test setup for measuring T S S .............................................................. 84
Table 5 Tangential signal sensitivity.................................................................... 84
Figure 52 Cascaded two stage network................................................................ 93
Figure 53 Plot of measured t i l t ............................................................................ 97
Figure 54 Efficiency o f Schottky diode at 9.5GH z..............................................99
Figure 55 Efficiency of Schottky Diode at 11.1G H z........................................ 100
Figure 56 Convesion loss of Schottky diode at 9.5GHz.................................... 100
Figure 57 Schottky diode conversion loss at 11.1GHz...................................... 101
Table 6 Rise and fall times of Schottky d io d e .................................................... 102
Figure 58Schottky rise and fall times at 9.5GHz .............................................. 103
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 7 Schottky TSS ..............
Table 8 Schottky equivalent TSS
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
Chapter I - INTRODUCTION
A high impedance microwave detector allows the monitoring of microwave signals
without loading or drawing significant power out of the signal. The majority o f today's
microwave detector circuits are based on microwave diodes. These circuits extract power
from the incoming microwaves due to their low impedance. A more sensitive detector
circuit can be built by using a negatively biased Field Effect Transistor (FET)1,2. In this
circuit the necessary power will be supplied by a power supply at the receiving point and
the detector itself will not extract power from the incoming microwaves3,4.
The theory for the detection using a negatively biased FET was previously
reported by the author in a master's thesis5. The theory for detection and mixing of
microwaves using a negatively biased FET are similar and this earlier work paralleled the
mixer work done by Oxner6.
LIshii, T.K., "Theory of Delay Time Reduction for Deep Space Communications", IEEE
MTT/ED Societies, Milwaukee Chapter Meetings, November 18,1987, Milwaukee,
Wisconsin.
2Ishii, T.K., "Wavefront Detector NSF Proposal", Marquette University, Milwaukee
Wisconsin, December 1988.
3Ishii, T.K., "Theory of Delay Time Reduction for Deep Space Communications", IEEE
MTT/ED Societies, Milwaukee Chapter Meetings, November 18,1987, Milwaukee,
Wisconsin.
4Ishii, T.K., "Wavefront Detector NSF Proposal", Marquette University, Milwaukee
Wisconsin, December 1988.
sLuglio, Juan R., "Negatively Biased Field Effect Transistor Used As A Microwave
Detector", (Master's thesis, Marquette University, Milwaukee, Wisconsin, 1989).
6Oxner, Ed, "FET's in Balanced Mixers", (Siliconix Application Note, Small-Signal FET
(continued...)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
The objective of this research is to further develop a sensitive microwave detector
based on a low noise negative gate FET. Since the power is supplied by the bias power
supply instead of the microwave signal, it possible to make the detector very sensitive. The
gate of a FET, when properly biased and phased, has a high impedance which will result in
the detector having a high impedance. The upper frequency limit of the FET chosen for
this design, when used as an amplifier, is 18 GHz. As a result, if care is taken in the input
and output circuit design, high speed can be obtained.
%.. continued)
DataBook, Santa Clara, CA, 1986) ch7 pp55-64.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
Chapter H -THEORY DEVELOPMENT
A. General Detector Theory
According to The New IEEE Standard Dictionary o f Electrical and Electronics
Terms detection is defined as “(A)Determination of the presence o f a signal. (B)
Demodulation. The process bywhich a wave corresponding to the modulating wave is
obtained in response to a modulated wave.”7 A detector is defined as "A device for the
indication of the presence of electromagnetic fields."8 In this dissertation detection will
be considered to be the demodulation of an amplitude modulated signal and a detector
will be considered to be a device to demodulate amplitude modulated signals.
For most detectors, detection is possible due to the nonlinearity of a currentvoltage (I-V) characteristic. Whether the device used for detection is a diode, or a field
effect transistor (FET) or something else, the current in the detector is given by
I=f(V)
(1)
where I is the device current in the output circuit and V is the device voltage at the
input circuit.
If V =v0+v, where v0 is a D.C. bias and v is a small signal
I=f(v0+v)
(2)
1The New IEEE Standard Dictionary o f Electrical and Electronics Terms, fifth edition,
chair Gediminas P. Kurpis, ed. Christopher J. Booth, (New York, 1993) s.v. detection.
8Ibid, s.v. detector.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
Assuming small signal variation this can be approximated by a Taylor series about the
bias point
(3 )
If v=ViSin(o)t)
I = f(v0) +vif /(v0)sin((ot) + ^ - f //(v0)sin2(<ot) +...
(4)
using the identity
1-cosu
(5)
and ignoring the higher order terms
I = f(v0) +vif /(v0)sin(cot) - - J - f //(v0)cos(2&)t)+-J-f//(v0)
(6)
The first term, f(Vq), is a D.C. term and is due to the bias. The second term,
v i f ( v o )sin ((o t),
is a current at the carrier frequency. The third term,
v 2/4 f '(vo)cos(2ci>t), is a current at the second harmonic of the carrier frequency. The
fourth term, v 2/4 f
'(V q) ,
is the term o f interest in this dissertation. It is a D .C. current
proportional to the square of the magnitude of the carrier and the curvature, f"(vo), of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
the I-V characteristic at the bias point and allows the demodulation or detection of an
amplitude modulated signal.
B. High Impedance Detector Theory
High impedance is a relative term. It must be used in comparison to the system
being probed. Consider the 50Q system with a detector in parallel with the load as
shown in figure 1. For convenience assume the microwave input impedance of the
detector and the load impedance are real. Also assume that the parallel combination of
the load and detector impedances matches the system impedance. The higher the
impedance of the detector, the less microwave current it draws from the source. If the
detector input impedance is equal to the load it draws half of the available microwave
current, and hence half of the available power. However if the impedance of the
detector is ten times the impedance of the load it will draw one tenth of the available
current and hence only ten percent of the available power.
Measuring the parameters of a system presents a problem. How does the
instrumentation used to make the measurement affect the measurement? The objective
of a high impedance detector is to minimally affect the system being measured.
As stated in the introduction, most of today's microwave detectors are based on
microwave diodes. A diode has a nonlinear I-V characteristic. An example of a diode
detector circuit is shown in figure 2. The inductance L acts as a high pass filter which
shorts the video components of the detected signal at the input. At the microwave
frequencies the impedance o f the inductance is high. At video frequencies it’s
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
impedance is low. The capacitance, C, acts as a low pass filter and blocks the
microwave components of the detected signal from the load Zj.. The impedance of the
capacitance is low at microwave frequencies and high at video frequencies. It is
important to note that the detected current delivered to the load was sourced by the
microwave signal being detected.
gen
□
M Detector
Figure 1 50Q System with detector in parallel to the load.
V,gen
Figure 2 Diode detector circuit.
The input impedance of the diode circuit can be defined as the ratio of the
microwave voltage applied across the input of the diode circuit to microwave current
sourced by the microwave signal. If the video current delivered to the load is to be high
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the video input impedance of the circuit has to be low since all of the detected video
current goes through the input circuit of the detector. This also means that the diode
video impedance must be low since the video impedance o f the inductance is low.
Typically a detector diode has an impedance of 200Q.
LPF
Deblock.
gen
Figure 3 FET Detector
A microwave FET has a nonlinear I-V characteristic that can also be used for
microwave detection. The drain characteristic, or IDvs. Vqs characteristic, is nonlinear.
A circuit that takes advantage of this characteristic is shown in figure 3. The circuit has
a D.C. source labeled VDD. The resistors R,, Rd, and R, are present for biasing. The
capacitor 0*.,,,<** blocks the D.C Bias voltage from the output. The capacitance
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
is used to bypass R, at the microwave and video frequencies. A low pass filter keeps
the microwave frequencies from getting to the output. The load in this circuit is 7^.
The current in the drain circuit goes through an impedance equivalent to Rd in parallel
with ZL.
It is important to note that the microwave source supplies voltage to the gate of
the FET. If the gate resistor R, is large, and the input impedance of the FET is large,
the circuit input impedance will be large and will draw little current from the
microwave source. If the input impedance of the FET is not high but the magnitude of
the reflection coefficient is large it can be transformed to a high impedance through a
transmission line of the appropriate length. The current delivered to the load is supplied
by a D.C. source. The input impedance of the circuit can be high and draw little
current from the microwave signal while still delivering large current to the load.
In summary, to detect an amplitude modulated signal with minimal disturbance
to the system a high impedance detector is highly desirable. Such a detector can be
designed using a microwave FET.
C. General Time Response Study
The speed aspect of a detector must now be considered. What is the highest
carrier frequency that can be detected and what modulation rate can be detected? These
questions can be answered by examining the detector circuit and the device used.
First let us look at a packaged FET. The parasitic capacitances and inductances
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
Figure 4 Common source FET amplifier equivalent circuit
Figure 5 Circuit for finding Z
associated with a packaged FET as well as the transit time across the channel of the
FET affect the upper frequency limit that the device can be used. Modulation rate will
be examined in the next section
The equivalent circuit of a common source FET amplifier is shown in figure 49.
This is a simplified equivalent circuit but is sufficient to show the limiting effect of
interelectrode capacitances. The gate to source capacitance is given by Cp . The gate to
9Millman, Jacob, Microelectronics, Digital and Analog Circuits and Systems (McGrawHill, New York, 1979), p 480.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
il
S
S
Figure 6 Circuit for finding short circuit current
drain capacitance is given by Cgdand the drain to source capacitance is given by C*.
The output impedance can be found using a theorem from Millman10.
vo=iz
(7)
Where V0 is the output voltage, / is the output terminal short circuit current and Z is the
impedance between the output terminals. To find Z let the input voltage, V;, be short
circuited (Vj= 0) then gn, Vf =0. The circuit reduces to that shown in figure 5, and Z is the
parallel combination of Z^, r* C& and Cgd- Hence
1
(8)
where YL is the admittance corresponding to
gd = l/rd is the conductance
I0Ibid, p 479 and pp 714-718.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
corresponding to r* Ydl=jcoCdl is the admittance corresponding to C*, and Ygd= jo) C#
which is the admittance corresponding to Cgd.
The circuit for finding the short circuit current is shown in figure 6. The current
from drain to source in a zero resistance wire connecting them would be
I - S mV
^ gd
(9)
The voltage gain \ is
_ Vo_IZ_ I
V ^
Vi V?
Using equations (8) and (9)
V
-
YL+gd+Yd +Ygd
(11)
It can be seen in figure 4 that the gate circuit is not isolated from the drain circuit.
They are connected by Cgd. From Miller’s theorem11this admittance can be replaced by
Ygd(l-K) between the gate and source and Y ^ l-l/K ) between the drain and source,
where K = A*. Therefore the input admittance is given by
YrYgs+{\-A v)Ygd
(12)
“Millman, Jacob, Microelectronics, Digital and Analog Circuits and Systems (McGrawHill, New York, 1979), p 481 and p728
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
Substituting in Ay results in
Y=Y
I gs + 1-
1-
-gmV<*Ced
JaC,*<*
(13)
=/co £ +q
SmCgd+jf^Cgd
gs sd rL-gd- M C ^ c gd)
The quotient in the brackets can be separated into real and imaginary parts using
Ai _ ai+Jbi _ (f*iV \Xq2~ A ) _
A2
a 2+J b 2
(fh + jb jfo t-jh j
. .bia2~aib2
a l + bl
1
(15)
letting a! = gmC^, bL= coC^2, %= YL+g,,, and b2= w ^ + C g j) the quotient becomes
(rL+gf+vXPt+cJ*
(rL+gd)2+*\cds+cgf
Then
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(16)
13
I
o
c
V-
li
2
gd°>
+<"
\-t.< l r 2 / Y
+tr 'k
(17)
/
This can be simplified by considering the order of magnitude o f the admittance values and
the frequency. If we assume that the frequency is on the order of 1010Hertz, and the
capacitances are on the order of picofarads (10‘12farads) and the YLand g,, are on the
order o f milli-Siemens (10'3 Siemens) then equation (17) becomes
+/G) C
(18)
This shows that the input admittance is frequency dependant and that as the
frequency increases eventually the input signal is shorted out. It should also be noted that
the FET internal capacitances and resistances in figure 4 are all frequency dependant
terms.
The time it takes for an electron to travel through the channel of the FET is given
by
(19)
where v is the electron drift velocity in the channel and L is the channel length. For GaAs
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
v is 2.0X 107cm/s.12 The electron transit angle is defined
(20)
where co is the operating angular frequency. For a NE42484A FET, the device to be used
later in this dissertation, the gate length is 0.35 pm, which is approximately the channel
length. The transit time is 1.75 X 10'l2s or a transit angle of 5.67 degrees at 9 GHz. The
manufacturers data sheet for an NE42484A includes data for use through 20 GHz.
D. Response to Pulse Modulated Microwaves
To determine if a detector’s output can accurately reproduce the envelope of a
pulse modulated microwave signal the bandwidth of the pulse needs to be determined.
Consider the periodic gate function shown in figure 7. The periodic gate function is a
generalized case of the square-wave in which the width and height o f the rectangular
pulses are variable and one level is at zero amplitude. The Fourier coefficients are given
by
(21)
where
I2Liao, Samuel Y., Microwave Devices and Circuits,2nA edition, (Prentice Hall, Inc.,
Englewood Cliffs, NJ, 1985), p 331
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
As seen in figure 8, aside from a scaling constant A, the spectrum is dependant
only on r and T. As T is increased the amplitude of the spectrum decreases as 1/T and the
spacing of the line decreases as 2rr/T. As t increases the amplitude increases proportional
to t and the frequency content is compressed into a narrower range o f frequencies.13
-T/2
0
T/2
Figure 7 Periodic gate function.
The use o f the frequency translation property of the Fourier transform makes
finding the amplitude spectrum of an ideally pulsed microwave signal fairly simple. It is the
amplitude spectrum of the periodic gate function centered at the carrier frequency. Ideal
I3Stremler, Ferrel G., Introduction to Communication Systems, (Addison-Wesley
Publishing Company, Reading, MA, 1977) pp 41-45.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
. £s
4n
2s
6^k • • •
Figure 8 Spectrum of periodic gate function,
pulsed microwaves however are impossible to generate because o f the need to have a
periodic gating function with zero rise and fall times. Close approximation can be made
however. Microwave PIN diode switches with manufacturer specified rise times on the
order of 2ns are available. A spectrum simulated on Hewlett-Packard Microwave Design
System (MDS) of a 9.1GHz carrier pulse modulated with a pulse 40ns wide with 2ns rise
and fall times and a repetition rate of 10kHz is shown in figure 9. As can be seen from the
markers the 30dB bandwidth of the pulse is 374MHz.
The output circuit o f the detector also needs to be considered. If the output circuit
has a capacitance of 10 pF and a resistance of SO Q the resulting R-C time constant will be
S00 ps. Assuming three time constants gives a reasonable estimation of the rise time, the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
rise time is limited to l.S nS. Thus, for the detector to accurately follow the envelope of a
pulse modulated microwave source the microwave band width of the detector has to be
large enough to pass the modulated microwave spectrum and the output time constant
must be small enough to track the envelope of the signal.
D a t a s e t * p u ls e
cn
0 .6
GHz
M2-M1— 3 0 . 922E+00
I 1-M1= 187. 00E+06
Ml
f r e «1
Ml=-3 7 . 965E+00
I l “ 9 . 1000E+09
9 .6
GHz ft
T ra c e l= d B (o u tp u t)
Figure 9 MDS simulation o f Pulsed puled microwave spectrum
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
Chapter HI - CIRCUIT DESIGN OPTIMIZATION AND FABRICATION
A. Bias Point Determination
In chapter II it was shown that detection is possible due to a nonlinearity in a
device characteristic. To design a detector based on a negatively biased FET the D.C.
characteristics of the FET must be determined and a bias point must be picked. The
device chosen for the design was an NE42484A. This device meets the criterion stated in
chapter II section C and was available as a samlpe from California Eastern Laboratories,
the device manufacturer. The D.C. characteristics of five NE42484A FETs were measured
using an HP4155A semiconductor parameter analyzer and an HP16442A test fixture. This
is essentially an automatic curve tracer that can be set up to measure the IDvs. VGS family
of curves for a FET. Drain current is measured while holding VDS constant and stepping
VGS. The next curve in the family is then measured by changing the value o f VDS and
stepping VGS again. The analyzer also calculates ^ and the curvature (CRV), by taking
the derivatives of the data (CRV is the second derivative of the IDvs. Vos curve) using
the following algorithm
y= ~ ~
*a-*i
when w=l
=yr r y "-' When lO K A T
Xn . r Xn-l
_yN-yN.
when n=N
XN~XN-l
where,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
yn’:
the differential coefficient for measurement index number n
ya:
value of 1“ expression for measurement index number n
x„:
value of 2nd expression for measurement index number n
N:
number of steps or number of samples
The data measured, and that calculated, can then be stored on a floppy disk and
analyzed on a computer.
The test fixture shown in figure 10 was constructed in order to connect the
NE42484A FETs to the HP16442A test fixture clip leads. The ferrite beads and 50Q
loading were necessary to keep the FET from oscillating during the measurements.
Drain
Gate
Ferrite
Beads
IOOOpF
SOQ
IOOOpF
DUT
SOQ
Ferrite
,
Beads \
I
Source
Figure 10 Test fixture built for curve trace measurements
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Drain Currant vs. Gate Voltage
3
3.006-02
FETA
FETB
200E-Q2 -
FETC
FETD
FETE
0.006*00 lttHltllltllllllHlllllilHIIHIttHllllllllltllllllllllHlllllllllllllllllllllllllllllHHimilllllllllllllllllllHlHSwiiml^liilHimitllltlM
HW
mHHHIIIIIIHIW
IHIIHtHtHIHHII
^ a oi cq n s (D io t
v pi n n ^ 7 7 o» a s s id in m. v n n n »• 9
Figure 11 Measured IDvs. VGS at VDS = 2.02 V
CURVATURE • Raw Data
3.006-01
2006-01 •
1.006-01
FET D
|
0.006*00
-1.006-01
-2.006-01
-3.00E-01
Vo
Figure 12 CRV vs. VGS data
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
The data was collected on the five FETs swept VGSfrom -2 to 0 volts in .01 volt
steps and VDS was stepped from 1.5 to 2.46 volts in .04 volt steps. As can be expected
there was some variation in the curves from device to device. There was little variation
within each device as a function of VDS. A plot of IDvs. Vos with VDS = 2.02 volts for
each of the FETs is shown in figure 11 and a plot of the curvature is shown in figure 12.
Each of the FETs was given a label, A thru E. Noise in the measured data is accentuated
by the differentiation used to calculate the curvature and is evident in figure 11. The data
was filtered using Matlab Software by taking the Fourier transform o f the data, finding the
CURVATURE-SMOOTHED DATA
0.25
FETA
0.15
FET B
FET C
>
ai
fO
l£
FETD
FETE
0.05-
-aos - 0.1
Vfl
Figure 13 Smoothed CRV vs. VGS data
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 14 Self bias circuit
frequency of the noise, applying a filter to eliminate it and again taking a Fourier
transform. A plot of the filtered data is found in figure 13. As can be seen in figures 12 and
13 CRV has a peak. In chapter II equation (6) showed that the detected signal was
proportional to the second derivative (in this case CRV) of the characteristic being used
for detection. The optimal bias point for detection is at the peak of CRV. The maximum
values of the CRV and the corresponding values of VGS and IDfor the filtered data at VDS
= 2.02 volts for each of the five FETs is shown in table 1.
Using the self bias circuit shown in figure 14 values can be calculated for the bias
-V GS
resistors Rq and Rs.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Using VDD IS volts and the data for FET D
R
s
( 0-49*0 = 72 0Sq
0.0068/4
R = l P2f ~ .1?.r -72.0SQ= 1.837A&
D -0.0068(4
The nearest standard 5% values of 75Q and 1.8kQ will be used.
MAX-FILTERED
FET
(Vg, In, CRV)
A
-0.58V, 6.3mA, 0.2054
B
-0.58V, 7.0mA, 0.211
C
-0.55V, 6.5mA, 0.1833
D
-0.49V, 6.8mA, 0.2438
E
-0.46V, 6.7mA, 0.2453
Table 1 Bias points for maximum curvature
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
B. Microwave Circuit Design and Optimization
After the bias conditions have been determined the microwave circuit of the
detector has to be designed. In this design the D.C. bias circuit as well as the video output
circuit must be considered which can be quite challenging. In chapter II, figure 3 showed a
FET detector circuit however the implementation o f the bypass capacitor and the low pass
filter were not addressed. Also the fact was mentioned that the FET input impedance
might not be large. For the device chosen for this design, the NE42484A this is exactly the
case. Since the goal of the design is to obtain a high microwave impedance at the detector
input, a method o f dealing with this problem was found and will be presented.
Figure 15 TRL Calibration kit
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
As was discovered with the initial design, when dealing with a device with a large
gain bandwidth (the NE42484A data sheet shows more than 11.4 dB of gain in the range
o f 0.1 to 20GHz), careful attention must be paid to design stability. Some basic guidelines
for detector stability will also be presented.
The first step in designing the microwave circuit is determining the S-parameters
o f the FET at the bias conditions of operation. The S-parameters were measured on an
HP8510C network analyzer calibrated with a Thru Reflect Line (TRL) calibration kit
designed following an application note from Hewlett Packard14and implemented in 5880
Duroid microstrip. A photograph of the calibration kit is shown in figure 15 and the
mechanical drawings are included in the appendix. The mechanical alignment of the
microstrip standards and the end launches are critical to the calibration so dowel pins were
used. The calibration kit can, by design, be used to make calibrations o f the network
analyzer from 2.1 to 15.5 Ghz.
The repeatability o f connections is of concern in this type of calibration kit. It is the
largest single factor that limits the effectiveness of the calibration15. The calibration
depends on the measurement of the S-parameters of a zero length thru (the T in TRL), a
large reflection (the R, the angle o f which only needs to be known to ±90 degrees, an open
or a short) and a section o f transmission line (the L) of known length and the impedance
to which the calibration is to be normalized, usually 50Q. Slight missalignment of the line
l4Hewlett Packard Product Note 8510-8A, Network Analysis, Applying the HP8510 TRL
calibration for non-coaxial masurements, (Hewlett Packard, Palo Alto, CA 1992)
l5Hewlett Packard Product Note 8510-8A, Network Analysis, Applying the HP8510 TRL
calibrationfor non-coaxial masurements, (Hewlett Packard, Palo Alto, CA 1992) p 21
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T rm l-fitia f..ltl,l]
T r « * w i t a f .. i a . i ]
T r n W itta f .ja .il
T m W * tta f ..S a a|]
TmHiW.iKliU
T rM W *tM ..K l,ll
sss
M Ok
2.1 Ok
2.1 Ok
T rn7>f*taf..K ll21
T rn W attaf..111.21
TracaWattaf..KI,21
na
no
15.5 Ok*
15.5 H I
15.5 QkC
2.1 Ok
2.1 Ok
2.1 Ok
15.1 Ok«
IS.S flk l
15.1 OkC
T rn lW « tarf..S a ,2 1
TrnlI«fattaf..SO,2]
T r n lW itta f ..ia .2 1
Figure 16 NE42484A S-parameters from 2.1 to 15.5 GHz
standard with the microstrip on the launcher or the two launchers can cause calibration
problems due to the reflection at the interface. As stated earlier dowel pins were used to
obtain good alignment, however if connection of the launchers is not accurately repeatable
errors will occur. The application note suggested a method of checking the repeatability of
the connection. This method consists of calibrating the system then connecting the
launchers to the line standard. The reflection at each port is measured and this data is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
stored in the network analyzer memory. The launchers are disconnected and then
reconnected and the connections are measured again. The network analyzer math function
is then used to subtract the stored data from the newly measured data and display it in dB.
The connection repeatability of the end launchers were measured to be better than -40dB.
This is about the same repeatability obtained in the example in the application note.The Sparameters of the three of the FETs were measured over two different frequency ranges.
FETs C and E were destroyed by instability problems that will be discussed later. The first
measurement was a wide band measurement over the frequency range of 2.1 to lS.SGHz
with 801 points in the frequency span. A plot the four S-parameters of each FET is shown
in figurel6. The second was a narrower span of 8.6 to 9.6GHz. A plot of this data is
shown in figurel7. The HP8S10C allows bias to be applied to the ports through a bias tee
internal to the analyzer. The bias conditions used for each FET is that presented in table 1.
A closer look at Su for one of the FETs is shown in figurel8. The marker shows
that the magnitude of the impedance of the FET is 19.46Q at 9.1GHz. This is quite low
however the reflection coefficient is large. A transmission line can be used to rotate this
low impedance around the Smith chart to a high impedance. A circuit for simulating this in
Hewlett Packard’s Microwave Design System software (MDS) is shown in figurel9. In
the circuit CMP50 is twoport model that contains the measured two port data for FET D
from 8.6 to 9.6 Ghz. The output port, port 2, is loaded in 50 ohms, CMP51. The input
port, port 1, is connected to a microstrip line, CMP36, that is 5mil (0.126mm) wide and
762mil (19.255mm) long. This is the narrowest line manufactureable with the available
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
1.1 Ok*
Tml*f*tM..Stla I]
T n a W ttn . .HZ, 11
TranM atk*..in(U
T rw fr4at*..lt!.l]
Tnu*4itka..KI,U
TriuW «tia..iri.l]
11s
II!
i.s
1.1
1.1
Tr>ca7*fatiii..ftlaU
TraeO-r«tkai..Itl,i:
rrm S-r«t*i..K l,U
ma
I .( Ok
1.1 OkA
i.a m i
1.1 QkC
1.1 M
1.1 Ok
TraM lt-tatai. .10,11
T raH lW «tka..ia,n
TrwalMM*. .10,11
Figure 17 NE42484A S-parameters from 8.6 to 9.6 GHz
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1.1 0 k «
l.a O k l
1.1 OkC
29
Trace 1
<C
9.6GHz
Y-FS=
1.0
8.6GHz
8 .6
GHz
FREQ
9 .6
GHz A
Ml
M l= Z 0 # (3 1 6 .4 3 E -0 3 -j2 2 6 .6 8 E -0 3 )
1 1 = 9 . 1000E+09
T ra c e l= S [l,l]
Figure 18 FET D Su
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
OFSO
TWPORT
one
NSTL
-gate
SUBST=s3lmtl
L=7G2 m il
W=5 m ii
CMP34
PORT SPAR
PORTNUM=I
R = 5 0 .0 OH
J X = 0 .0 OH
DATA=fetdn. FORMATTEDQfSFEtM req
R
-A A A r
M
AGROUND
R=50 OH
^7
AGROUND
oris
AGROUND
NSSUB5TRATE
-y — — — -
SUBST=s3lmiI
HU=t9 mm
ER=2 - 2
1 ^T ftP um MUR=I
f * - ------- C0ND=5.49e?
H=31 mi I
R0UGH=0 urn
- J -------------TAND=0.0009
S7
AGROUND
Figure 19 MDS circuit used to simulate the FET input impedance rotation
facilities. The label AGROUND indicates analog ground. The values of the substrate,
CMP 16, correspond to 5880 Duroid 31 mill thick with a Vioz. copper which is equivalent
to a thickness of 17pm. The model for the substrate includes loss and a cover height. The
other end of the microstrip is connected to a 50Q S-parameter port, CMP34, to allow Su
to be measured. The results of the simulation are plotted in figure20. The resulting
impedance is 1.72kQ at 9.1 GHz which is more than an order of magnitude larger 50Q.
FET D was chosen for this example since it is the FET that will be used to build the
detector, however as seen in figures 16 and 17 all o f the devices are similar.
In the previous example the bias circuit, the lowpass filter and the bypass
capacitors were not considered. These will affect the impedance that needs to be rotated.
To include them the method for implementing must first be determined. The
bypass capacitor must operate as a good short at both video and microwave
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Y-FS=
1.0
T racel
31
8.6
GHz
f req
9.6
GHz A
Ml
M1=Z0*(34. 377E+00+J1 . 1472E+00)
1 1 = 9 . 1 0 5 0E+09
T r a c e l = r o t a t i o n . . S C I ,1]
Figure 20 Rotated impedance
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 21 FET detector with radial stub shorts
frequencies to obtain good gain. A first design attempt tried to implement the bypass
capacitor as a microstrip double radial line short at a half wavelength from the one of the
source leads of the FET with a 15pF tantalum capacitor and a .OlpF capacitor connected
between the short and ground as shown in figure 21. The half wavelength transmission line
was used to make physical space to layout the double radial stub. A half wavelength away
from a short is a short. The other source lead was connected to a single microstrip radial
line short. The idea was that the microstrip circuits would provide the bypassing at
microwave frequencies and the two capacitors would provide the bypassing at video
frequencies. Another double radial line short half a wavelength away from the drain would
short the microwaves at the output, acting as the low pass filter. The MDS circuit diagram
o f this is shown in figure 21 and a photograph of the circuit that was constructed as shown
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
la y o u t.« n jta f lc « » c h ip l206
0*44
C
ONI
*€Tl
SU 8 S T * s 3 I a i I
H
C *l
/ ^ v
L -IS O
6 7 .5
\
SU8 S T * * 3 I a i I
pF
\
a iI
8*15
deg
a iI
\H * tO o i l
OH?
H * I 0 M tl
re tL
am
N$n.
5 U B S T * i3 I a i I
oni
-
-■
"
8*15 - i P I ”
\\w -io
R *50.0 OH
JX *0 .0
H* 5 0
SUB5 T * * 3 l m l
PO R TK U K - 3
Wi SMI
L*500 a t I
la y o u t.tn s tj r»ce*chip!206*
o iI
> -
OH
V 0C *I5.0 V
AGROUND
' l
AGROUND
P O R T N ltf- 2
PORT SPflfl
8 -5 0 .0
J X -0 .0
—
OH
OH
AGROUND
0»?t
1*yout. m ttance*nec64
I«l.n*424H
SUB5 T « s 3 I a . I
5 UB5 T * * 3 I a i I
L -2 0 3 .6 0 8
L -5 3 7 .8 7
K* 5
a il
*>l
am
H- 5
ASPIl
~*L
a iI
81*10Ail
la y o u t. m stance*chipI206
^7
_
80*143
5
ANC' 1 3 5
A il
deg
<3 j —SUB5T - s 3 I a « I
AGROUND
PORTNUM*I
poorsp*c
2 *5 0 .0
OH
JX -0 .0
OH
«O
B1VLI2
AGROUND
!? r
*
i
lay o u t.in ttan ce * ch tp l2 0 6
la y o u t. m stance*cH ip
V
AGROUND
lay o u t. instance*ehtpI206
Figure 22 MDS circuit diagram for FET detector with radial stub shorts
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
in figure 22.
The circuit was built and when power was first applied the detector detected a
signal, but then quickly stopped working. The measured bias conditions o f the FET
indicated it was burned out. The initial thought was that static discharge had destroyed the
FET. A second detector was built and this one was observed to go into oscillation at
about 2GHz just before it burned itself up. This forced the consideration of something that
is not generally considered with most typical detectors, namely stability.
The self bias detector circuit being used for this detector can be analyzed as if it
were an amplifier. Gonzales 16gives the conditions for stability in, terms o f reflection
coefficients, of a two port device as
ir s |< 1
(2 8 )
|r j < i
(29)
<1
irj=
"
l ' S2ir L
16Gonzalez, Guillermo, Microwave Transistor Amplifiers, Analysis and Design, 2nd
Edition (Prentice Hall, Upper Saddle River, NJ, 1997) pp 217 -219
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(30)
35
and
lr <
OUT 1
iSM+
<
22 1-^11r ,
i
(31)
where all coefficients are normalized to the same characteristic impedance Z0 and where
r s is the reflection coefficient of the source impedance connected to the two port, TLis
the reflection coefficint of the load impedance connected to the two port, Z ^ is the input
impedance of the two port, ZOUTis the ouput impedance of the two port, r * is the
reflection coefficient of the input of the two port and r*ouT is the reflection coefficient of
the two port. When the two port is potentially unstable there are values of r s and TL for
which the real parts of Z^, and ZOUTare positive thus making the two port stable.
To determine what source and load impedances produce a stable system the
regions where TL and Ts produce II^I = 1 and |r oLrr| = 1 respectively are found. This is
done by setting the magnitude of equations 30 and 31 equal to 1 and solving for the values
of Tl and r s. These values fall on circles (stability circles) given by the following
equations17
"Ibid pp 217-219
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
r
G V ^ u )*
L |S22I2- |A |2
s iA i
(32)
| £ J 2- |A |2
and
p
O V A .S g* l^ul2- |A |2
^12^21
(33)
l-S'nl2- |A |2
Where
A =SxlS22- Sl2S2l
(34)
In the Tl plane, the center and radius of the circle of values of r Lwhere r m| = 1
(output stability circle) is given b y 18
rL=
^12^21
„
(S22-bSu)'
Cr= --------------
(radius)
(35)
(center)
(36)
|S22|2-|A |2
"Ibid pp 217-219
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
In the r s plane, the center and radius of the circle of values of r s where r oux| = 1
(input stability circle) is given by
*^12^21
|S nl2- |A |:
(radius)
(37)
„
( s.'-A S ^y
Cs= ---------------
{center)
(38)
rs=
|SU|2- | A |2
To determine which region is stable, where values of r L (where r L < 1) produce
| r j <1 and where values of r , (where T, < 1) produce |r oUT| < 1 a point that does not lie
on the circle is examined. The simplest point to choose lies at the center o f the smith chart.
For example if
= Z„, then TL= 0 and [1^1 = Su. If |SU| < 1 when 7^ = Z0, then |r w| <
1 when r L = 0 and the center of the Smith chart is stable. If |SU| > 1 when
= Z0 , then
| r j > 1 when r L= 0 and the center o f the Smith chart is unstable.
Similarly, if Zs = Z0, then Ts = 0 and [roux| = Sa . If>|S22| < 1 when Zs = Z0 , then
\Tom\ < 1 when Ts = 0 and the center of the Smith chart is stable. If ISd > 1 when Zs =
Z0, then \rom\> 1 when Ts = 0 and the center of the Smith chart is unstable19.
It is important to note that a stability circle is valid at only one frequency. To
19
Ibid pp 217-219
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
determine stability using this method manually is very tedious since circles must be
plotted for each frequency of interest. Hewlett Packard’s MDS makes plotting these
circles relatively simple. It is an automatic in the presentation portion of the program.
Insertion of a stability circle equation calculates the requested circle at all frequencies for
which the S-parameters o f the two port are available. The software also determines if the
inside or outside of the circle is the stable region.
The input and output stability circles for FET A is shown in figure 23. These are
the stability circles of a common source configuration of the FET however. The circuit
used in the design of this dissertation is a self bias circuit that has a source resistor and
bypassing of that resistor. The stability circuit for determining the stability of the FET with
the resistor and bypassing used in the first design is shown in figure 24. The stability
circles are plotted in figure 25. This plot shows that the circuit had potential for instability
and as stated earlier instability was observed.
The second design removed the microwave radial stubs and transmission lines from
the circuit. It was decided that the variation in impedance as a function of frequency of this
configuration was most likely the problem with the first design. The MDS circuit used to
determine the stability of a FET with a bypassed source resistor and no microstrips in the
source circuit is shown in figure 26. The Stability circles are shown in figure 27. The
improvement is dramatic
It was also decided to remove the radial stubs and transmission line that was being
used as a lowpass filter at the out drain of the FET. The microwaves still have to be
blocked from the output. To accomplish this a 10 pF capacitor from the drain to ground
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
□VII
IMPORT
OATA=fetaw.FORMATTED.,FREO=freq
OV5
RORTSMR
P0RTNUM=l
R=50.0 OH
JX=0.0 OH O f*
QV9
m
AGROUND
AGROUND
AGROUND
on
PORTSPflR
AGROUND
Figure 23 FET stability circuit without radial stubs and transmission line
was used. At 9GHz 10 pF has an impedance of 1.77Q and will short out the microwaves.
At a video frequency of 30kHz the capacitor has an impedance of 530kfl. As discussed in
chapter II section D the R-C time constant of lOpF and 50Q, SOOpS, will allow for fast
rising and falling edges of a pulse modulated wave form.
The output circuit also needs a DC block to separate the bias from the detected
signal. A large capacitor in parallel with a smaller value high frequency capacitor is used
for this. The MDS circuit used for evaluating the impedance that the microwave shorting
capacitor, bias resistor RD, DC blocking capacitors and load, decoupling capacitor and
supply presents to the FET is shown in figure 28. A plot of the microwave impedance is
shown in figure 29. It can be seen that the impedance falls within the stable region
indicated in figure 27.
As discussed earlier the input impedance of the FET is quite low at 9.1GHz. It was
shown that the impedance of the FET could be made to look large thru the use of a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
m
freq
2.100E+09
3.577E+09
5.055E+09
S.533E+09
9.011E+09
9.480E+O9
10.9SE+09
12.44E+09
13.92E+09
15.40E+09
StabRegion_
o u tsid e
o u tsid e
o u tsid e
o u tsid e
o u tsid e
o u tsid e
o u tsid e
o u tsid e
o u tsid e
o u tsid e
0.0E400
0
fre q
2.100E+09
3.577E+09
5 . 055E+09
6.533E+09
B.011E+09
S.400E+O9
10.9BE+09
12.44E+09
13.92E+09
15.4 0E+09
indep(5tab_S,2) B.2B31E+00B
StabRegion.
o u tsid e
o u tsid e
o u tsid e
o u tsid e
o u tsid e
o u tsid e
o u tsid e
o u tsid e
o u tsid e
o u tsid e
id
♦»
tn
■
in
O.OE+OO indep(Stab_L,2) B.2B31E+OOB
Figure 24 Stability circles for FET A.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
an
KSSU8STRA1C
HU*19 ««
1 a T - 17u»»
+ f “ - -------
5UBST*s30m.I
ER-2.2
MUR-1
C0N0-5.49e?
H*31 mil
ROUGH-O un
1
-1---------------
TflND*0.0009
EQUATION 1=102 * < I
E0UAT10N th e ta l* 135 deg
3
,
.... ,
mm
OATA=feta««.FORMATTED.,FREO*freq
S7
AGROUND
AGROUND
e
.n
e
■*
e
CK*?I
n
«5»Hg
EOUATION th e ta - 115 deg
EQUATION 12=246 n i l
AGROUND
Figure 25 FET stability circuit with radial stubs and transmission line
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
E2
frag
2.500E+09
3.000E+OS
3.500E+09
4.000E409
4.500E+09
5.000E409
5.500E+09
G.OOOE+09
6.500E+09
7.00*408
7.500E+Q9
8.000E408
0.500E+09
9.000E40S
9.500E+09
I0.08E409
I0.50E408
I1.00E40S
U.50E40S
I2.00E4C9
I2.50E409
13.00E409
!3.50E+09
14.00E40S
0
atabrogion
inaida
inaida
inaida
inaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
inaida
inaida
inaida
56
*
0.CE400
indap (St ab_S,2) 6.2931E4001
St ib_S-SrcStibCi reI■(S)
StibflogionJXrcStabRagionfS)
E2
frag
2.500E409
3.000E40S
3.500E409
4.000E409
4.500E409
5.000E40S
5.500E409
6.000E408
6.500E409
7.000E409
7.S0CE40S
B.OOOE4Q9
I.500E40S
9.000E408
9.500E40S
IO.O0E4O9
I0.50E40S
1I.00E408
11.S0E40S
12.00E408
I2.50E409
13.00E40S
I3.SOE409
I4.00E40S
Q
atabragian
inaida
inaida
inaida
inaida
inaida
inaida
inaida
inaida
inaida
inaida
inaida
outaida
autaida
outaida
autaida
autaida
outaida
outaida
outaida
autaida
outaida
autaida
inaida
inaida
i
O.flE+OO
f ndap(St«b_L,2) G.2831E+OOB
Stab_L<oidStabCirclo(S, 1000)
StibRogionJAoadStibRagiantS)
Figure 26 stability circles for FET with radial stubs and transmission line
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
ea
ea
FREQ
S tib b g io n .
2.5OOE+Q0
3.000E+09
3.500E+Q9
4.00CE+8S
4.500E+Q9
5.000E+09
5.50CE+Q9
G.OOflE+09
6.300E+09
7.000E+09
7.500E+09
B.M0E+Q9
B.500E+09
9.000E+09
9.500E+Q9
10.00E+Q9
10.5CE+09
11.00E+09
11.30E+Q9
12.00E+Q9
12.50E+Q9
I3.00E+09
13.50E+Q9
14.00E+09
□
FREQ
2.SOOE+09
3.000E+Q9
3.500E+09
4.0ME+C3
4.500E+09
5.0ME+QS
5.50QE+09
6.000E+Q9
6.5ME+09
7.000E+09
7.500E+Q3
B.OOOE+Q9
8.500E+09
9.000E+Q9
9.500E+Q9
I0.ME+Q9
10.50E+Q9
II.0QE+Q9
11.50C+Q9
I2.00E+Q9
I2.50E+09
13.00E+Q9
13.50E+08
J4.00E+09
outaida
outaida
outaida
outaida
outaida
outaida
outaida
outaida
autaida
outaida
outaida
autaida
outaida
outaida
autaida
outaida
outaida
autaida
outaida
outaida
outaida
outaida
outaida
outaida
4
2S
*
O.flE+M
indop(Stib_Sa2) 6.283 IE+OOB
Stab_S-SrcStabCircla(S>
StabRagian_S-SreStibbgion(S)
E2
StabRagion.
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
autaida
outaida
autaida
autaida
autaida
autaida
outaida
autaida
autaida
4
35
O.QE+QO
indap(St«hJ.,2) 6.2831E+MB
StabJXoadStabCirclaCS)
StabRogian_L4jiadStaMbgion(S)
Figure 27 Stability circles for FET without radial stubs and transmission line
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
transmission line of appropriate length. The width of the microstrip line used to do this
impedance translation was as narrow as possible, Smil, in order to minimize the
disturbance at the junction with the 50Q line being monitored. The length o f the line was
determined by simulating the impedance of the FET and microstrip combination and then
adjusting the length until the maximum impedance at the microstrip line input was reached.
The Duroid substrate to construct this circuit has a relative dielectric
R = 5 0 OH
■A A A /—
MMB
R
CHP19
c
aground
o
o
^3"
ru
ii
C -J
R = 1 8 0 0 OH
-ii- —w v —
ow
CMP20
PORT SPAR
P0RTNUM=l
R = 5 0 . 0 OH
J X = 0 . 0 OH
AGROUND
R
a.
o
•IoTT
rvj
OP12
CMPll
/T \o
C
Of>lG
VDC=1 5 . 0 V
VDC
ii
<_>
S?
\7
AG RO UN D
~
AGROUND
AGROUND
Figure 28 Load impedance circuit
constant of 2.2 and is 31 mil thick. MDS calculated the impedance of this microstrip to be
174Q. The length of the 50Q microstrip from each of the connector to the junction with
the high impedance line is 2.2inches. This length was chosen to minimize the effect of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
microstrip launching connectors20. At a low impedance point on the high impedance line
the bias resistor Rq is placed. This is a high impedance point a the design frequency of
9.1GHz. It was discovered after building the circuit that if a high quality resistor (one that
has low parasitic capacitance and is capable of working at high frequencies) is not used
stability problems can occur. From a design stand point it probably would be better to
place the resistance directly at the gate of the FET. This would avoid the possibility of the
non-ideal impedance of the resistor being transformed by the transmission line, at a
frequency other than the design frequency, to a source impedance that would make the
FET go unstable.
An MDS circuit for simulating the source impedance presented to the gate of the
FET is shown in figure30. It includes the high impedance microstrip,CMP 14 and 15, the
gate resistor Re, CMP13, and the 50Q microstrip, CMP 18 and 19, with matched loads,
CMP 21 and 22, at both ends. As can be seen in figure 31 the source impedance presented
to the gate falls within the part of the Smith chart that was indicated to be stable by figure
27.
Here are some guide lines for designing the microwave circuit of the detector.
They also hold for other microwave circuit designs. In the microwave circuit try to place
any resistance as close to the desired location as possible. Placing them at a distance from
the desired location (for example one wavelength) makes the impedance o f the resistor
very frequency dependant and runs the risk of placing an undesired impedance at the
Packard Product Note 8510-8A, Network Analysis, Applying the HP8510 TRL
calibrationfor non-coaxial masurements, (Hewlett Packard, Palo Alto, CA 1992) p 9
20 Hewlett
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
<E
nj
SMI
8
12GHz
Cn
500.0
MHz
freq
12.0
T r a c e 2 = l o a d i m p . . SC 1 , 1 ]
Ml
Ml=ZO#( 1 . 255BE—03—j 3 4 . 9 3 3 E - 0 3 )
1 1 = 9 . 1000E+09
Figure 29 Plot o f the load impedance
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
GHz A
47
desired location. Microwave FETs have a large gain bandwidth and an undesired
impedance can make the circuit unstable.
Know the operating limitations of resistors and capacitors used in the circuit. Most
people think that chip resistors have a very high upper frequency limit. In reality a
standard 0805 package lOki) chip resistor has a large parasitic capacitance between it’s
leads, in parallel with the resistance, and at 10MHz the AC to DC resistance ratio is about
0.8. The resistor looks like a short at 1GHz. The larger the value of the resistance the
worse this problem becomes because the ratio of the desired AC to DC resistance
becomes smaller21.
^International Manufacturing Services, Inc. Data sheet for Precision Thin Film Surface
Mount Chip Resistors in 0805 package, Portsmouth RI, 1996.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
^7
AGROUND
CW>I7
MSSUBSTRATE
£
m
tA
CHPI8
MSTL
ii
►
—
Ul
□3
3
tn
i
i
V
o
o
.
o
ru m
CXI CD
ii
II
_ i 2
in
5UBST=s31miI
ER=2.2
H=31 mi I
-J-------------
MUR=I
C0N0=5.43e7
R0UGH=0 um
TAND=0.0009
CMPI5
MSTL
IL
CD
_
- » ----------HU=19 mm
1 *1=17um
CMP20
MSTEE
CMPH
MSTL
SUBST=s3ImiI
L=538.85B mi 1
W=5 mi I
5UBST=s3lmiI
L=249.74 mil
W=5 mi I
P0RTNUM=l
R -5 0 .0 OH
JX=0.0 OH
CMP5
PORT SPAR
<7
6
r-i
CMPI3
MSTL
6
LTi o j i n
CD CM O ')
3
CMPI3
R
—
AGROUND
II II
<7
AGROUND
W7I
R
<7
AGROUND
Figure 30 MDS source impedance circuit.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
500.0
MHz
freq
12.0
T r a c e l = s o u r c e i m p . . S C I , 1]
M1=Z0#(8.3977E+00—j 8 . 8 142E+00)
1 1 = 9 . 1000E+09
Figure 31 Simulated source impedance
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
GHz A
50
C. Predicted Results
M« Oi
. iM ttlD C I 'C liif U K
A CSOHO
AGHXJNfl
Figure 32 MDS circuit used to simulate the ratio of input to gate voltage
The design described is expected to have a high impedance and have a high
detection speed as described in chapter IIB and chapter m B. These were major design
goals o f the design. Can the output level of the detector be predicted? What about the
microwave insertion loss and the conversion efficiency?
To make these predictions an MDS ac simulation can be used to find the ratio of
the voltage at the input o f the 50Q transmission line to the microwave voltage that appears
at the gate of the FET. The circuit used to do this simulation is shown in figure 33. At the
one end o f the 50Q microstrip line, CMP39 and 40, is an ac source, CMP52, with a series
50Q resistor,CMP54. At the other end is a 50Q load,CMP53. The high impedance
line,CMP36 and 37 leading the FET circuit is in the middle of the 50Q line, CMP39 and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
40. The circuit was simulated sweeping the frequency of the ac source from 8 .6 to
9.6GHz. The ratio of the voltage at the point labeled gate to the point labeled inputvolt in
figure 32 as a function of frequency is shown in figure 33. Using this data along with the
curvature
11=13
113=11 3.86E -0 3
5000E
in
8 .6
GHz
fre q
a tte n = (m a g (g a te )/m a g (in p u tv o lt))
Figure 33 Simulated ratio of input to gate voltage
data in table 1 a video output voltage across Zu CMP 18, for a given microwave
input power to the transmission line can be calculated. Assume the device used for the
detector is FET D. Also assume that a pulsed microwave of lOmWrms during the pulse on
time is input into the 50Q line at 9 .1GHz. The input voltage during the pulse is given by
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
v -ft/F R
(39)
Where P is the rms input power, in watts, during the pulse and R is the input impedance.
Then for P=0.01 W, and R = 50Q
=72/01 *50=IF
(40)
The ratio of input to gate voltage at 9.1GHz from figure 33 is 0.13761 so the voltage at
the gate is 137.61mV. The output current at the drain of the FET is given by
Vt
i =— CRV
(41)
where Vg is the voltage at the gate and CRV is the curvature for the FET characteristic
shown in table 1. This current will develop a voltage across the parallel combination of the
50Q output impedance and the drain resistor RDas shown in figure 3 and 32. Assume that
this drain resistor Rq has a value o f 1.8kQ. This combination resistance, R, has a value of
48.65Q. The output voltage is then
2
V
ni
V0= -L x C R V * R = l:*
4
4
01
x 0.2438 x 48.65 = 56.0mV
With this output voltage the instantaneous power delivered to
(42)
=50Q can be calculated
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
62.%2\xW=-\2.Q2dBm
(43)
At first this appears to be a 22.02 dB conversion loss since
Conversion loss=PjndB -PouflB =10-( -12.02)=-22.02
(44)
The ratio of the voltage at the gate of the FET to the voltage at the 50Q microstrip is
0.13761 and
20 log(0.13761) = -ll.22dB
(45)
Taking this into consideration the conversion loss due to the FET alone is actually 4.8dB.
This number however is input power dependant. Most of the microwave power is not
lost in the conversion process. It is delivered to the microwave load at the end of the 50Q
microstrip. The insertion loss of the 50Q microstrip with the detector can be simulated in
MDS. The results of this simulation are shown in figure 34. The vertical axis is scaled in
dB and represents the microwave insertion loss of the 50Q microstrip and the detector.
The horizontal axis is frequency. Marker M l is at the minimum on the plot and M2 is at
the maximum. The difference of the two is 0.0497dB. Also the insertion loss of a 50Q
microstrip line without the detector can be simulated.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
HI—351.1 0E-03
M2— I01.27E-03
,25001*09-
^
---------
cn
freq
9.6
T racel= d B (lin ed et..S C 2 ,13)
Figure 34 simulated loss of transmission line and detector
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
GHz A
55
I
1
C \J
1—
"
'
I
■
o
I
\
\
N
\
CM
\
<D
a
ia
\
/
V
\
\
\
\
\
\
X
co
co
M
CM
o
I
1
/
8.6
GHz
freq
9 . 6 GHz A
Trace2=dB(SC2,1])
Ml
Ml=—2 1 9 . 45E—03
I 1=8.6000E+09
M2
M2=—2 3 7 . 38E—03
I 1=9.6000E+09
Figure 35 Simulated loss o f Transmission line
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
The results of this simulation are shown in figure 35. Again the vertical axis is
scaled in dB. The difference between the maximum and minimum insertion loss is 0.01793
dB. The difference between the two plots at 9.1GHz is 0.073 dB.
At 9.1GHz if the insertion loss o f the 50Q line is -0.228dB and the input power is
lOdBm then
lOdBm - 22%dB=9.112dBm =9.4S8mW
(46)
and the insertion loss of the SOQline and detector is 0.302dB
\0dBm-0302dB =9.69SdBm =9.32SmW
(47)
The difference of 0.160mW was delivered to FET circuit. A conversion efficiency can be
defined as the ratio of the power delivered to the load to the power absorbed by the
detector. For this example
0.06282 n , n,
T)=----------=-----------=0.393
P
0 16
r absorbed
rag)
*
1
The efficiency is microwave input power dependant. This can be seen in equation(43) and
(48) as well as figures 43 and 44. As the microwave input power increases the efficiency
will become larger and since output power P0 comes from a DC power source efficiency
greater than one is possible.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
D. Circuit Fabrication
The final circuit was constructed on 0.031 inch 5880 Duroid with Vi oz. Copper on
one side and 0.25 inch aluminum plate on the other. A cover was machined for the circuit
out o f0.500 inch aluminum plate with GDS eccosorb rubber glued to the top inside the
cover. The machining of the cover and the removal o f the Duroid where the cover attaches
to the circuit board as well as the cover screw holes were done on a CNC mill and the
layout for this was done with EZ Feature Mill software. The CNC programs and
mechanical drawings for the cover and the board are included in the appendix.
An MDS circuit diagram of the final detector design is shown in figure36. A photo
graph of the finished circuit board is shown in figure 37. The layout feature o f MDS used
this diagram to generate gerber22 files for the layout o f the microstrip circuit traces. These
files were sent to a circuit board manufacturing house, Advanced Circuits23 in Denver,
Colorado, and precision laser film plot mask was made. Advanced Circuits claims better
than lmil accuracy on their laser plot films. This mask was used in a photo-etch process to
transfer the copper traces to the circuit board. After the board was etched the dimensions
of the traces was checked under a calibrated microscope. At 1:1 scale the microscope
calibrated graticule has .005 inch divisions and at 3:1 this is equivalent to .001666 inches
the 5 mil(0.005 inch) trace measured exactly to those lines.
The resistors used in the construction of the circuit are chip resistors in a special
“ The Institute for Interconnecting and Packaging Electronic Circuits (IPC)
“ Advanced Circuits Inc.6875 East 48th Street, Denver, CO 80216
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
C 't S
u
F
S U 8 S T « s 3 ta i I
A H C « ~ 6 7 .S d e g
C -2 4 0 0
pF
R »15 a i l
S U 0 S T -s 3 la tl
i ,t y e u t . m s ta < ic e s b y p a s s
l
H-10ail
* I50 a il
H-IOatl
cm ?
« IL
lh
B
a o
• • in
5 U 8 S T -t3 la il
o
m ut
B
Wtt
-51-
1*500 « tl
H *50 a i I
C
9m
W
P O R T N U H -J
5 U 8 S T -s 3 la iI
R * 5 0 . 0 OH
J X « 0 . 0 OH
A N C '6 7 . 5
R -IS
C '2 4 0 0
deg
\ \ H - I 0
pF
V
AGROUND
a il
l a y o u t . m s t a n c e 'b y p a s s
ail
AGROUND
— la y o u t. m s ta n c e * ia s re s
V O O I S .O v
: -x
AGROUND
5 U B S T -i3 la i I
k
O P4I
POST.SHfl
L*40
a il
MMO a i l
W ? T W * -J o r a
R - 5 0 . 0 OH
J X - 0 . 0 OH
a
m m
cm a i
c*iqpr
AGROUND
^7
AGROUND
Iay o u t . m s t an ee* ch i p 1206
ana
i f t f l i k « H 2 4 l< i
S U B S i* s 3 J a t I
5 U S 5 1 * s 3 la tI
1 * 2 0 2 .2 1 8
L * 5 3 8 .8 6 0
a il
a il
M «5 a i I
layout. iAstance*ctiipl206
AGROUND
ly O U t.tn s t m c e a b y p e s s
la y o u t.in s ta n c e « ia s r e s
\7
AGROUND
----------------------- S U B S T - s 3 l » i l
R * 5 0 .0 OH
J X - 0 .0 OH
■13 « a
D W .J
fclf|7ua **•!
C 0 N D * 5 .4 9 e 7
H »3I
a il
- 1 -------------------
R 0U G H *0
ua
r A N D * 0 .0 0 0 9
AGROUND
Figure 36 MDS circuit diagram of the final detector
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 37 Photograph o f final detector and cover
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission
60
Figure 38 Close up of circuit
high frequency package made by International Manufacturing Services, Inc.(IMS)24. They
call this package a partial wrap chip resistor because the lead goes only part of the way
around the end of the resistor, reducing the package capacitance. The part number for this
package is RCC-0502PW. They are specified for use up through 26.5GHz. The size of
this package is also very small, 0.050 inch by 0 .0 2 0 inch, making them a challenge to
solder in place. With patience however this can be done with a small soldering iron. A
close up photograph of the circuit is shown in figure 38 to illustrate the size of the parts
^International Manufacturing Services, Inc., 50 School House Lane, Portsmouth, RI
02871-2418
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
61
being used. The resistors are available in standard 5% values from 10Q to 2kQ with
special values available on request. At the time the resistors were ordered the company
had 4.7kQ available from a special run done for another company. This is the value that
was used for Rg. The measured value of the actual resistor used is 4.61kQ. The value used
for R,, is 1.8kQ and the measured value resistor used isl.77kfl. The value used for R, is
75Q and the measured value is 73.4Q. These resistances were measured using a METEX
M-4650 digital multi meter.
The 2400pF capacitors used for the bypassing and decoupling are from Dielectric
Laboratories, Inc.23 They are sold for use as a broad band DC block and are specified for
use through 20GHz. The part number for this is C08BL BB1X 5UX. The lOpF capacitor
is part number ATC100A100KW from American Technical Ceramics (ATC)26 and is a
high Q, 10% tolerance part and is specified for microwave use. The lSpF capacitors used
for bypassing at the video frequency are tantalum capacitors with 10 volt rating DigiKey27 part number P2027. The FET used is an NEC NE42484A. Several were supplied as
samples from California Eastern Laboratories. The connectors used for the microwave
input and output as well as the DC input and video output are from M/A Com and are
designed for microstrip interfacing. The male is part number MA2051-1618. The female is
part number MA2052-1618.
In order to obtain a good ground, screws were used. The screws are brass 0-80
“ Dielectric Laboratories Inc., 2777 Route 20 East, Cazenovia, NY 13035-9477
“ American Technical Ceramics Corp., 1 Norden Lane, Huntington Station, NY 11746
27Digi-Key Corporation,
701 Brooks Avenue South, Thief River Falls, MN 56701-0677
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
62
machine screws with the head turned down to reduced its size to a minimum and then
plated with 24K bright gold to increase their conductivity. Three of these screws were
placed close to the critical grounding points in the circuit and connect the copper on the
circuit side of the Duroid to the aluminum base plate on the other side. Also the ground on
the circuit side of the board was clamped by ledges on the cover and shorted to the
aluminum base plate along the cover and base plate interface.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
63
Chapter IV - EXPERIMENTAL MEASUREMENTS
After the circuit was built the first test was to check the bias. The desired bias
point was at a gate to source voltage -0.49V with a drain current of 6 .8 mA and a drain to
source voltage of 2.02V as indicated by table 1 and figure 13. Since industry available
common value resistors had to be used for Rq, R^, and Ro in figure 3, the bias point was
expected to be slightly off from optimum. Manipulating equation (25) a new value of VDD
can be calculated to get closer to the desired bias point. Using VDS= 2.02V, ID= 6 .8 mA,
Rd = 1.77kQ and Rs =73.4Q, the measured values for the resistors
VDD=VDS+{RD+R^ID=2.02V+(\.llk£l+13AQ)x6.%mA=\4.55V
(4 9 )
With 14.55V bias measured at the circuit board, VDS measured 2.02V and the voltage
across Rs, which is equal to the negative of VGS, measured 0.494V. This is very close the
optimum bias point o f VGS = -0.49V.
A. Detector efficiency.
To measure the efficiency of the detector the test circuit shown in figure 39 was
used. It consists of an HP8690B signal source with, an HPX382A precision attenuator, a
MICA isolator, a General Microwave F9114 PIN switch, a Stanford Research Systems
DG535 pulse delay generator, a lOdB directional coupler, a waveguide DC block, and
HP54522A digital oscilloscope, a coaxial inside DC block, an HP436A power meter, and
an HP8596E spectrum analyzer.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
PRECISION
ATTENUATOR
GENERAL
MICROWAVE
P 9 1 I4
10dB
JW W
H P 8690B n« it
SIG.GEN. u u '
H P8S96E
SA
STANDARD
RESEARCH
SYSTEMS
D G 535
PU LSE GEN
X
PET
DET
COAX INSIDE v
DC BLOCK
A
VG DC BLOCK
— DC
g
HP436A
POWER
METER
H P 54522A
DIGITAL SCOPE
Figure 39 Test setup for measuring efficiency
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
First the power at point A in figure 39 was measured, with the PIN Switch closed
and the attenuator set to OdB, using the power meter. Then the 50Q line of the FET
detector was connected to point A and the other end of the 50Q line was connected to the
power meter through a coaxial inside DC block. The pulse generator was used to turn the
PIN switch on and off at a 30kHz rate with a 50 percent duty cycle. The video output of
the detector has a DC blocking capacitor to separate the bias from the video signal. The
frequency of 30kHz was chosen for convenience and 50 percent duty cycle was for ease of
converting the pulsed microwave power measurement to peak power.
The switch has a
manufacturer’s specification of an 80dB on to off ratio. The peak to peak voltage out at
the video output o f the detector was measured on the oscilloscope and the power at the
output of the 50Q microstrip line of the detector was measured with the power meter for
varying attenuator settings. The data was collected at 9.5GHz and 11.1GHz and is shown
in tables 2 and 3. These frequencies for the carrier were chosen because of the rise and fall
time characteristics for the detector at 9.5GHz and the largest voltage output was at
11.1GHz. The video power delivered to the load during the on time is calculated as
n
Where
video
_ V2
7
^L
(50)
is the impedance of the video load, in this case the 50Q scope input, and V is
the measured video voltage output. The efficiency can be obtained by converting the
attenuator setting to an input power, subtracting the power delivered to the power meter
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
66
attenuator
microwave
Vvideo out mV
Spttinn
Pnut dR
-1 6 .9 2
-1 5 .9 2
-1 4 .9 2
-1 3 .9 2
-1 2 .8 9
-1 1 .9
-1 0 .9 2
-9 .9 3
-8 .9 6
-7 .9 7
-6 .9 7
-5 .9 8
-4 .9 3
-3 .9 5
-2 .9 4
-1 .9 4
-0 .9 2
0 .0 3
1 .0 4
2 .0 5
3 .0 5
4 .0 4
5 .0 3
6 .0 4
7 .0 3
8 .0 2
0.5 6 2 5
0.625
0.6875
0.7 8 1 2 5
0.9 8 6 7 5
1 .15625
1 .40625
1.75
2 .125
2 .7 5
3.4375
4 .3 1 2 5
5.625
6.8 1 2 5
8.5
1 0 .9375
13.5
1 7 .1875
2 1 .5 6 2 5
2 6 .5 6 2 5
33.125
41 .8 7 5
52.5
65
80
100
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
Table 2 Data needed to calculate efficiency at 9.5GHz, for P;,, = 12dBm
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
67
attenenuator
Sflttinn
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
microwave
W ideo out mV
Pnirt riRm
•2 8 .4 5
-2 7 .4 2
-2 6 .3 4
-25.5
-2 4 .4 2
-23.41
-2 2 .4 2
-2 1 .3 9
-2 0 .4 2
-19.41
-1 8 .3 9
-1 7 .3 7
-1 6 .3 6
-1 5 .3 6
-14.31
-13.31
-1 2 .2 7
-1 1 .2 5
-1 0 .1 9
-9 .1 7
-8.1
1.3125
1.5625
1.96875
2 .3 7 5
3.125
3.75
4 .6 8 7 5
5.9 3 7 5
7 .5
9 .2 1 8 7 8
11.7188
14.6875
18.125
2 2 .5
2 8 .1 2 5
34 .3 7 5
4 2 .5
51.875
6 3 .1 2 5
7 6 .2 5
9 1 .2 5
Table 3 Data needed to calculate efficiency at 11.1GHz, for Pfa= OdBm
to get the power absorbed by the detector, and dividing the video power by the power
absorbed.
Pin[dBm]=PaU[dBm]-Attenuator [dB]
(51)
PJdBn>\
PJmW\=\0
10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(52)
68
P ab,orbed\.m W \ = P J m W \ ~ 2 x P m i c r o s
(5 4 )
P
im W \
P absorb*k m " J
(
)
Where the factor of 2 in equation (S3) comes from the fact that the microwaves are pulsed
at 50 percent duty cycle.
This method does not account for the loss o f the 50Q microstrip transmission line
or the connector/coax-to-microstrip transition loss which is not include in the simulated
data of figures 34 and 35.
In order to account for this a 50Q microstrip line with the same connectors and
enclosure was constructed. The insertion loss was measured on aHP8510C network
analyzer. A plot of this insertion loss is shown in figure 40. The insertion loss of the
detectors 50Q line was also measured on the network analyzer. The measurement was
done with DC power applied to the detector. This data is shown in figure 41. The
calculated loss of the line, shown in figure 35, was approximately -0.23dB at 9.1 GHz. An
insertion loss of approximately 0.25dB per connector is believable. The difference
between the two measurements should be accounted for by the detector. An overlay of the
two sets of data, shown in figure 42, however, indicates that the detector has less loss at
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
69
some frequencies, about 0. ldB than the microstrip line alone has. This is believed to be
due to the variation in the soldering of the connectors to the boards and stray standing
waves. The largest difference between the two is 0.2dB which indicates that the goal of
high impedance of the detector has been accomplished.
Another method of determining the power absorbed by the detector is to convert
the attenuator setting to an input power then assume a 0.25dB connector insertion loss
and 0.1 ldB insertion loss for half of the microstrip line. This will give the power at the
center of the 50Q line where the high impedance line is connected.
P.athighimpedance line<
[dBm] =Pin[dBm] -0.25dB-0.1ldB
(55)
This power can then be converted to a voltage and the power absorbed is the square of
this voltage divided by the impedance of the detector that was determined using the
method in chapter HI section B but using the correct device loading, as seen in figure 43,
for the impedance before rotation. That rotated impedance is 2.9kQ and can be seen in
figure 44.
athigh impedanceline
athighimpedanceline'
athighimpedanceline
2.9kQ
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(56)
(57)
° ID
i
¥S «H
E"Js
im i
* -------
U 1^
oin
tnm1
cu^i
____I
mm
1—
tra.
n o
I—H
01 01
a• ^• in
id
QQ ■
Figure 40 Measured insertion loss of 50£} microstrip line
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71
ST?
•. H 1
B
!H "
;U IB
!°
m
■Hi
1____ I
S
r—
/
(V
(ft
.....-
1
!
!
i
i
1
‘
1
i
ii
—
r
I
i
if
i
______
□n
a
L ,
—
—
r
j
t
?i
:
!
—— ___ __j
<
CUOT
. • uj
S « ?.
q
ttt>, • S ' . . . - .
'-r . —
____
_ _
j
i
___ J
*u
Figure 41 Measured insertion loss of 50Q microstrip line and FET circuit
Reproduced witd p e r .is s .o n of tde copyripd, owner. Fodder reproduction p r o d d e d witdou, perm ission.
Figure 42 Comparison o f 50Q line with and without FET circuit
Reproduced w ith permission o. the copyright owner. Further reproduction prohibited w ithout permission.
73
B - 5 0 OH
0*11 AGROUND
R « l 7 7 0 OH
D A T A * fc td n .F O R W TTCD. , F R C O * f r t q
Wl 5M0
PORTNUM*I
R -S O .O OH
X Y * 0 .0 OH
S U B S T « s 3 J a t1
S U B S T - s 3 la .l
L - 2 4 9 .7 4
L - 5 3 8 .6 6 8
M »5 a t I
a t I
H*S
a t
V O C * I 5 .0
V
a il
I
^7
^7
AGROUND
AGROUND
AGROUND
AGROUND
01*17
AGROUND
^7
AGROUNO
C S W S T fltfC
AGR^r0UND
-------------------
hL s „
1
f
* T « l? u a
r -
H »3I
a il
-i---------------------
S U 8 S T -* 3 la .l
■**;*
HUM
C 0 N D « 5 .4 9 e 7
R0U G H *0 u a
r A N 0 * 0 .0 0 0 9
Figure 43 MDS circuit used to determine detector input impedance
<c
Trace1=SC1
L .
M1=Z0*(5B. 4 18E+00+J1 .0 3 16E+00)'
Il=8.9275E+09
I2 -H
"
in
8 .6 GHz
fre q
9 .6
Figure 44 MDS simulation of detector input impedance
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
GHz A
74
The efficiency o f the FET detector as a function o f input power at 9.5GHz is
plotted in figure 45 and at 1 1. 1GHz in figure 46 using the data from tables 2 and 3
respectively and equation (54). In figures 45 and 46 the horizontal axis is the microwave
power available at the high impedance line in dBm and the vertical axis is the efficiency. It
can be seen that efficiency increases with input power. An efficiency greater than 1 is
possible since the circuit has a DC source to supply power.
Though the signal output and efficiency for o f the FET are greater at 11. 1GHz
than at 9.5GHz The data shows that a price has been paid in insertion loss.
^d e te c to r- ^'atA
line&connectors
dutycycle ^D CBlock ^microwave out
(58)
At 11.1GHz assuming 0.6 dB loss for the 50Q line and connectors, 3dB loss because of
the 50 percent duty cycle and a measured 0.28dB loss due to the coaxial DC block the
insertion loss of detector was 4.22dB. At 9.5GHz the insertion loss was O.ldB This
increase in insertion loss is due to the frequency dependant nature of the impedance
rotation performed by the high impedance microstrip line and a resulting decrease in
impedance at frequencies other than the design frequency.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
75
E fficen cy a t 9.5G H z
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
-15
-10
0
5
-5
Microwav* Pow*r a t High Impadac* Unt(dBm)
10
Figure 45 Efficiency at 9.5 GHz
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
76
efficiency
-25
-20
-10
-15
0
-5
Power at Juiction dBm
Figure 46 Efficency at 11 .1GHz
B. Conversion Loss
Conversion loss can be defined as the difference between the microwave power
absorbed by the detector and the video power delivered to the video load.
Conversion bss=PabmM-PvUeo
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(59)
The data to calculate this quantity was collected when the efficiency measurements
were made. A plot of the conversion loss at 9.5GHz is shown in figure 47 and at
11.1GHz in figure 48. From these plots it can be seen that insertion loss is input power
dependant and that the conversion loss decreases as the microwave power is increased. It
can also be seen in figure 48 that a conversion gain is possible.
Conversion Loss at 9.50Hz
22
o
2 .
M
8
-15 -14 -13 - 1 2 - 1 1 - 1 0 - 9 - 8 - 7 - 6 - 5 - 4 - 3 - 2 - 1
0
1 2
pow tr at Junction [dBm]
3
4
5
6
7
8
9
10
11
Figure 47 Conversion loss at 9.5GHz
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
13
78
comenkm toss at 11.1GHE
-20
-15
-10
»
8
Powtr at Junetian(dBn|
Figure 48 Conversion loss at 11.1GHz
C. Detection Speed - Rise and Fall Time
The test setup used to measure the video rise and fall times of the detector is the
same as that used to measure the efficiency, as shown in figure 39, a with one minor
change. The difference is the power meter is not connected to the 50Q microstrip line,
instead there is a 50Q microwave load.
The Stanford Research Systems DGS3S pulse generator was used to drive the PIN
switch with a 40ns pulse at a 30kHz rate, thus modulating the microwave carrier. The PIN
switch, General Microwave model F9114, has a specified rise and fall time of 2ns.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
79
The HP54S22A digital oscilloscope can measure the 10 percent to 90 percent time
automatically. This feature was used to measure the rise and fall times of the detected
video output from the detector. The pulse width of 40ns was chosen to allow the time
base of the oscilloscope to be set narrow enough to get an accurate measurement o f the
rise and fall times and still get both the rise and fall on the display at the same time. The
rate of 30kHz was chosen for convenience. The measurements were taken at carrier
frequencies that were varied from 8 .6 to 9.6GHz in 0.1GHz steps. The precision
attenuator was set to OdB during the measurements. This gave a power of approximately 9
to lOdBm at point A in figure 39.
An example plot of the data collected at 9.5GHz is shown in figure 49 and the
results at all the carrier frequencies are tabulated in Table 4. Plots at all of the measured
frequencies are included in the appendix. The time scale in figure 49 is Sns/div and the
vertical scale is SmV/div. The data seen in figure 47 was averaged 64 times to reduce
errors introduced by noise. The detector is a negative detector so the output falls as the
microwaves are turned on.
As can be seen from this data the detector has a rise and fall time that is close to
that o f the rise and fall time specified for the PIN switch (which is 2ns) creating the
modulation. The fastest of the rise and fall times is in fact faster than the PIN specification.
Manufacturer specifications are sometimes conservative, but not always. It is hard to
determine if the rise and fall times are being limited by the modulation, however it can be
stated that these measured rise and fall times are an upper bound for the rise and fall time
o f the detector.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80
Some of the plots show ringing as the microwaves are turned on, as shown in
figure SO, and this affects the measured fall times. It can not be determined if this ringing is
in the PIN switch or the detector. The best behaved carrier frequency, meaning symmetric,
fast rise and fall times, is 9.5GHz .
Carrier Frequency
GHz
Rise Time
Fall Time
ns
ns
8 .6
1.311
3.373
8.7
1.354
1.339
8 .8
1.507
1.604
8.9
1.439
3.643
9.0
2 .2 1 0
4.100
9.1
2.544
0.906
9.2
2.923
0.855
9.3
2.306
0.968
9.4
1.872
1.235
9.5
1 .6 6 8
1.887
9.6
1.489
2.672
average
1.874
2.053
Table 4 Rise and fall times of FET detector
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
81
CCS
o
ffi in
cu
( D
O
- *
D O *
O
• • (*)
i«
w
•s i l l
9
L
m m mm
3 0 0 0
5 ~ ww
c e
mm mm
4•»—*» QI
-0 o a.
u^ >
t&l
uc s
*a9D£Q
O
LQ
c in
cn
c
a. a
u
CJ
a
m
c
CL
3
O
A
«
o •
L«
CL
g
U r
u
zJ
U
o
a
=£
9
41 O
a a
u• 9a
v •
a tn
i
x. >
I
♦» -•
oo
to
>X
«• >
♦» K
9 O
S«
cn in
a
oo c
a
T
O
'O
Ui bi
9 9
cn cn
A
1 0» «O
~ i aa
T3 -• 9 1 0
O *»
> ■*
x §><
JJ3 = •
o»- t e w
-itu *■• —
n1 ea1 1—
^
• 9 o cm
9 >
<§£
5
Figure 49 Rise and fall time plot for a carrier of 9.5GHz
Reproduced * *
permission * * . copyd8 n, owner. Fodder r e p r o d u c e prodiPiied w«dou, p e n s io n .
82
c a» «
aa
a
in
r-*
CD
0
a is —.
m
•O N
• r a a <~
«*
sm . .
■
#
«N
M
^
M
.
^
M
5 wa wo wo
f
O
e mcm
«•
4i• —
«l a .
•-• «•
i
o a.
U
V
>
to
K
W
u
T11
J0
0 .0
E in
-N
^ oO
O
c)
*s
3 *0
oo
os
as
9
•8L «-
<r=u
a:
ui
2
UI
CJ
CL O
CJ
o
cn
c
a.
o§
uT
u.
<L
Q
o0
>
C
*a»O
a
0
L
<•*. oo
Vo
>
*D
V.
c
§
in
o in•
i
x>
~ 5
>V
Z
«
mm
•c O
o
•
.
tn in
oa
o
01
ux
Q
0
•at at aa 0c
o
Ui Ui
10 «
• 0 *0■*CiM# OQ
a
xO 4*cn
O0
> 80
^
T3 *□
0
a ,2
cn
0
—x
6
>0
c •» g l 8
£“ I
Figure 50 Rise and fall time data showing ringing
83
D. Sensitivity
The tangential signal sensitivity, TSS, of a detector is defined as the input power at
which the output power is twice the noise28. This can be measured with the setup shown in
figure 51. The test setup is very similar to the previous test setup. However instead of the
oscilloscope the spectrum analyzer is attached to the video output at point B and the
coupler is terminated. The pulse generator is set for a repetition rate o f 30kHz and 50
percent duty cycle. The rate was chosen to move the video signal away from the zero beat
noise in the spectrum analyzer. The 50 percent duty cycle was chosen to give a reasonable
average video output power.
Measurements were done at carrier frequencies of 8 .6 to 9.6GHz in 0.1 GHz steps.
For each frequency the power at point A(PJtA[dBm]) in figure 51 with the PIN switch
closed and the precision attenuator set to OdB was measured to calibrate the input power.
The detector was then attached to point A and, with the attenuator set to maximum, a
measured value of at least 80dB, the noise level on the spectrum analyzer was measured.
This measurement was done with 100 video averages with a resolution bandwidth o f 10Hz
and a video band width of 10Hz. The center frequency was 30kHz and the frequency span
o f the spectrum analyzer was 1kHz. The noise was read from the noise marker on the
spectrum analyzer at a frequency of 30kHz. Next the attenuator was opened to the
point(attenuator [dB]) where the detected video signal at 30kHz was 3dB above the noise
floor, again 100 video averages were used. The TSS was then calculated as follows
“ Lance, Algie L. Lance, Introduction to Microwave Theory and Measurements, (McGraw
Hill, New York, 1964) pl25
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
84
PRECISION
ATTENUATOR
GENERAL
MICROWAVE
F91I4
A
H P8690B OUT ■
S1G.GEN.
lOdB
50
- A /W V 1------------------- VW
< :T A w n A e n
RESEARCH
SYSTEMS
DG535
PULSE GEN
<1 A
“ DC
COAX INSIDE
DC BLOCK
Figure 51 Test setup for measuring TSS
frequency
RN7
8.6
8.7
3.8
8.9
9.0
9.1
9.2
9.3
9.4
9.5
9.6
power at A
riRm
0.1619735
0.2366392
0.0603795
0.2201574
0.3502928
0.3742650
0.2448567
0.0945090
0.2489596
0.0646604
0.0
attenuator
settinn dB
29.5
28.0
28.2
30.0
29.5
25.0
23.5
27.0
30.0
32.5
35.0
T SS dBm
Table 5 Tangential signal sensitivity
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
•32.3380
-30.7634
-31.1396
-32.7798
-32.1497
-27.6257
-26.2551
-29.9055
-32.7510
-35.4353
-38.0
85
TSS =PalA[dBm]- 3dB-Attenuator[dB]
(60)
The extra 3dB loss was included to account for the 50 percent duty cycle. The power at A
, attenuator setting, and the calculated TSS for the different frequencies is shown in Table
5. The maximum sensitivity of -38dBm in this frequency range was obtained. It must also
be noted that the measurement of TSS is dependant on the noise level of the test
equipment being used. The detector and measurement system can be thought of as a
cascaded two stage network with stage one being the detector and stage two being the
measurement system. The noise out of a two stage cascaded network is29
(61)
Where Gt is the gain of stage one, G2 is the gain of stage two, k is Boltzman’s
constant,T0 is the standard temperature, 290K, Blz is the bandwidth of the cascaded
network, and F12 the noise factor of the cascaded network and is given by 30
F =F
12
X
Gt
Bn
(62)
29Mumford, W.W. and E.H. Scheibe, Noise Performance Factors in Communications
Systems(Honzon House-microwave Inc., Dedham, MA, 1968) p47
30
Ibid p48
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
86
Since TSS is defined as an input level that gives an output level equal to double the noise,
A new term, tangential signal sensitivity output (TSSO), can be defined as the signal
output level under TSS conditions, then
TSSO[dBm]=TSS[dBm) - CL[dB]
(6 3 )
where CL[dB] is the magnitude of conversion loss of the detector. Equivalently
TSSO[W\=TSS\W]*Gy
(6 4 )
where Gt is the conversion loss in ratio form. Then
TSSO[W\=Nl2[W\ =Fl2GlG2kT0Bn [W\
(
B (F2 - l ) G ,)
r ,---- +
G f i ^ n \W\
Bn
Bn
(6 5 )
=(FlGlG2kTQBl +(F2- lykT 'fifij [W]
Solving for F ^ B i
TSSO (T7
F,iG,B,
-------(F , -1)5,
l l =—
GJcT,
(66)
The terms G2, F2 and B2 are the terms that make TSS measurement system
dependent. The terms Ft, G, and Bx however are inherent detector parameters that are
important in determining the sensitivity of the detector.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
87
E. Noise Figure
To measure noise figure a small signal method was used. It is a modified version of
the CW signal generator method described by Montgomery31. The CW signal generator
method is used to measure the noise figure of a microwave receiver. A signal generator is
attached to the receiver input and the output is attached to a power measuring device.
With no signal from the signal generator the power at the output is measured. The power
from the signal generator is then turned on and the level adjusted to give twice the power
of the noise floor. This signal is designated s(f„). Then the noise factor, F, is
where k is Boltzman’s constant (IX lO'^J/K) and T0 the standard room
temperature(290K) and B is the bandwidth of the system being measured.
This method can also be used to measure the noise figure of a detector with the
help of the cascaded noise figure equation for two stage network given in equation (62).
Rearranging this equation
B ,, F~ B.
T ,
3lMontgomery, Carol G., Technique of microwave Measurements, Volume l(Dover
Publications, New York, 1966) p224
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
m
88
A problem seems to appear in equation (6 8 ) ifF 12B 12/B 1^F2B2/G 1B1. This can be
the case if Gt is a loss instead of a gain. These equations are generally accepted since Gt is
usually large and F2 is generally small. But what if this is not the case? Examination of the
noise in the two stage cascaded network shown in figure 52 gives some insight.
The noise at the input of stage one is
N, r kT<Px
(69)
where Bt is the bandwidth of stage one. The noise at the output o f stage one due to the
input N;i is
No r kT<PxGxFi
(70)
where Gt is the gain one of stage one and Fx is the noise factor of stage one. The noise
available at the input of stage two, N q, is
Ni2=kT<PxGXFX+kT(Pl
(71)
where B2 is the bandwidth of stage two. The total available noise output of the cascaded
network is then
+Arr0fi2)G2F 2 +kTQBn
(72)
Considering the whole network, N 0 can also be written
^0=kFOFX2GXG2F 12 +kFoPi2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(73)
89
where B 12 is the bandwidth of the total network and F 12 is the noise factor or the total
network. Then
k T ^ G fify F 2 +kT(f i 2G2F2 +kT0Bl2 = kT(f i x^}yGJ ' l2 +kTQBl2
,F2+B2G7F j +Bl2 =Bl2GlG2Fl2+Bl2
(7 4 )
(7 5 )
Then solving for Ft
P _ ByfijGiFn ^
Bn
l~ B fifi^ 2
B fifi^ 2
B2G f 2
Bn
B fifiA
B 12
B 12
B^GyG2F2 i?1G 1 BlGlG2F2
B fi
F
_ B
17F 12
B2
l ~ b xf 2 b xg x
(
)
(7 8 )
(7 7 )
These equations assume that all of the noise power available at the input of a stage
will make it through the system to the output. The bandwidths of the stages, however,
keep this from happening. If B2 > Bt all of the noise from the output of stage one will get
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
90
through stage two, however if B2 < Bx the noise from stage one is limited by stage two.
For multistage cascaded networks the bandwidth of the narrowest stage dominates.
With the detector connected to the spectrum analyzer this is a two stage cascaded
system. Stage one is the detector and stage two is the spectrum analyzer. The bandwidth
of the detector is 178MHz and the bandwidth of the spectrum analyzer is the resolution
bandwidth setting of 10 Hz so B2< Bt. The band width of stage two dominates and the
approximation B ^ B ^ B ^ can be used. Then equation (77) becomes
F _B12F12_ B2
B12B12 Bf i l
(79)
^__L
~F1 ~Gl
Considering stage two, the noise floor of the spectrum analyzer with the input
terminated in 50Q can be measured directly from the noise marker on the spectrum
analyzer. At the spectrum analyzer settings used in the previous section the noise floor at
the output of the spectrum analyzer is -131.8dBm/Hz. Since this is in dBm/Hz and the
bandwidth is 10Hz the noise power can be obtained by converting to watts and
multiplying by 10. As a result
iV2=660.69xl0-18[ in
The double noise power point, s^fo), is then
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(80)
91
s2(f0)=1.32138xlO-l5|7F]
(81)
and
=455.65 xlO2
(82)
The noise measured with the detector connected to the spectrum analyzer and no
power applied is -120.9dBm/Hz. Proceeding as above
512C/-0)=16.218xlO'l5[lP]
(83)
F 12=5.591x10s
(84)
The noise factor of the detector can then be found using equation (79) and assuming G t
is large
r _ F i2 _ 5.592x10s _ 100C*
--------------=18.856
F2 455.65x10
(85)
To justify the assumption that Gx is large it is important to recognize that the
detector also acts as an amplifier at the video frequencies and that thermal noise at the
input is broad band. When measuring the noise out o f the FET detector it must be
recognized that the noise output of the system not only comes from conversion of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
92
microwave noise to video noise, it also includes video noise that makes its way through
the FET as an amplifier. Since the FET has large gain at low frequencies, the data sheet
shows in excess of34.189dB at 100MHz32, the noise out of the detector acting as an
amplifier at the video frequencies must be considered and Gt can be considered the gain at
the video frequency.
“ California Eastern Laboratories, RF andMicrowave Semiconductors,(California Eastern
Laboratories, Los Angeles, CA, 1995-96) ppl-53 to 1-59
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
93
1
1
r-
e,
6
fk
“
Q,
1
f>p.
C, t$x
i
Figure 52 Cascaded two stage network
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
94
F. Bandwidth
The response to a square wave modulation can be used to determine the detector’s
video bandwidth. Using the rise times measured in section C the upper 3dB frequency of
the video response can be determined.33
,
0.35
fir —
(8 6 )
where t, is the rise time. If the mean value of the average rise and fall times from table 4,
1.9636ns, is used
/„= - — - S =\l%.2MHz
H 1.9636n5
(87)
K ’
The lower 3dB frequency of the video response can be determined from the
percent tilt in the response to square wave modulation34. The percent tilt can be measured
using the test setup in figure 39. To make this measurement the pulse generator is set to
30kHz convenience and 50 percent duty cycle in order to meet the condition that the
33MilIman, Jacob, Microelectronics,Digital and Analog Circuits and Systems (McGRawHill, New York, 1979), p452
MIbid p453
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
95
square wave be symmetric35. The video response with OdB set on the attenuator is
measured on the oscilloscope. The percent tilt is defined as
V-V'
Tilt=———x 100%
V
(88)
where V is the peak to peak voltage observed on the oscilloscope and V’ is the sag in the
voltage. A plot of the video response of the detector is shown in figure 53. In this plot The
oscilloscope measured the peak to peak voltage as 45.9375mV.This is V. The Y markers
are set to measure half of the sa g , on the top part of the square wave, this is shown in the
markers area of the plot as delta Y and is 5mV. The term V’ is then V-2 delta Y,
45.9375mV-10mV = 35.9375mV then
r//f_ 45.9375-35.9375
*100%=21.78%
7 ~
45.9375
(89)
The lower 3dB comer is given by
Tilt f
_ 21.768*30A£fe
=2.078A£fe
7tx100%
TtxlOO
(90)
where f is the modulation frequency. The video band width is then
BWvideo=fH-fL=m.244MHz-2.01ZkHz=m.242MHz
3SIbid p453
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(91)
96
The maximum video output voltage was obtained at a carrier frequency of
11.1GHz. The -6 dB voltage points were at 9.72GHz and 11.58GHz giving a 1.86GHz
microwave bandwidth. This bandwidth however is where the detector is loading the
microwave circuit. The traditional method o f defining the microwave bandwidth is not
useful when describing this high impedance circuit. What must be considered is how much
loading is tolerable. The impedance, as shown in figure 44, is what must be referred to.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
97
» > >
0• 9• 0aa
Air* •
im m in
23.0000
IM CMO
•
o
3
m
C 3
oa oo oa
01
m
o o a
0 * 0 IM
0• €
O3 0• m
M N
I
lit
X II*
X
0
• • •
uo
6 fc*
I0 I0 ~2
CM** 0
KX^
a
in
AVERAGE - 128
&
I0 h0 5*2
Al«
>■>• •
Maaau'*a*onta
ua
m0 m
O
C 0
*O
o £o
0.00000
■
i-s
•J X JJ/irtttsd L tsts!
?
i-s
5
ua
0
c
u»
3
8
RR
<0
O
c
CENTER
U»
a
*u»
«i a
ja
io
w O.
3.00
ua/dlv
a
ua
0
0
i i
fi ;
>
+x»•D
°
»^ mm
>^ ■oa itjm
io
io
-23.0000
01
*0*«
a .
a c
g
om
a
uj uj
.
i • o '|
■o s**- -a i a5
z a >
^
.
*• t «n««
0
m
w 01 0 m a o
0£ V
I X «*U"
^
i
e
l
O I-
o
X
Figure 53 Plot of measured tilt
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
98
G. Comparative response tests.
To show some of the advantages o f the FET detector the efficiency, conversion
loss, rise and fall time, and tangential signal sensitivity of a Schottky diode (Advanced
Control Components model ACSP2S04PZ, serial number 0126) were measured. The same
procedures described in the preceding sections were used.
The diode is matched to 50Q at the microwave input and video output. In
measuring the efficiency it will therefore be assumed that the power absorbed is
100
percent of the available power at the microwave input. A plot of the efficiency of the
Schottky diode at 9.5GHz and 11.1GHz are shown in figure 54 and figure 55 respectively.
A comparison with figures 45 and 46 for the FET detector show that the FET is several
orders o f magnitude better in efficiency than the Schottky diode.
The conversion loss of the Schottky diode at 9.5 and 11.1GHz is shown in figures
56 and 57 respectively. At OdBm input power the FET detector has 23 dB less conversion
loss at 9.5GHz and 43dB less at 11.5GHz. It should be noted that the conversion loss and
efficiency of the Schottky diode at 9.5 and 11.5GHz are the same at the power levels
measured.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
99
Schottky Efficiency at 9.5GHz
0.0007
0.0006
0.0005
0.0003
0.0002
0.0001
-25
-20
-15
-5
-10
Microwav* P o w ir a t lnput(dBin)
0
5
Figure 54 Efficiency of Schottky diode at 9.5GHz
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
100
Schottky Efficiency a t 11.1GHz
0.0003
0.00025
0.0002
0.00015
0.0001
0.00005
-20
-18
-16
-14
-a
-10
-12
-6
-4
-2
0
Microwav* Powir at InputtdBm)
Figure 55 Efficiency o f Schottky Diode at 11. 1GHz
Conversion Loss a t 9.5GHz
60 -
55 -
a
T
5e
o
e
s
u§
50 ■
45 ■
40 -
35 -
•25
•20
-15
-10
-5
0
5
Microwav* Powar at InputfdBm)
Figure 56 Convesion loss o f Schottky diode at 9.5GHz
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
101
Schottky Conversion Loss at 11.1GHz
-6855
50
45 -
S
c
3
40 -
35 -
-88-20
-18
-16
-14
-12
-10
-8
-6
-4
-2
0
Microwav* Powarat lnput(dBni)
Figure 57 Schottky diode conversion loss at 11.1GHz
The rise and fall times of the Schottky diode at carrier frequencies from 8 .6 to 9.6GHz are
shown in table 6 and a plot of the data collected at 9.5GHz is shown in figure 58. The
plots for the data at the other frequencies is included in the appendix. It should be noted
that the Schottky diode used is a positive detector meaning the output voltage increases
positively with increasing microwave input. Comparing the data in table 6 with the data in
table 4 for the FET shows that on average the FET has 0.856ns faster fall time (the time at
the turn on of the microwaves, remembering that the FET is a negative detector) than the
Schottky rise time (again the time at the turn on of the microwaves). Also on average the
FET has 0.627ns faster rise time (at the turn off of the microwaves) than the Schottky
diode has fall times (again at the turn o f the microwaves). These rise and fall times are the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
102
Carrier Frequency
GHz
Rise Time
Fall Time
ns
ns
8 .6
3.160
2.574
8.7
2.246
2.700
8 .8
1.238
2.741
8.9
2.870
2.591
9.0
3.162
2.338
9.1
2.978
2.511
9.2
3.394
2.355
9.3
3.601
2.438
9.4
3.256
2.607
9.5
2.413
2.588
9.6
3.686
2.069
average
2.909
2.501
Table 6 Rise and fall times of Schottky diode
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CCS
cn
cn oo e
in
0
• # If*
m
IN N ^
1.
..
0AAA
S
(
m
a
m
3 0 0 0
{
w
w
«•
«•
O
•
* •s .■
v
• - a.
'c. «D
*. »L
h
*ts o
0 .0
mm
cn
c
«•
m*
0 .0
8
U
^
u
a
cn
c
t
ac
t
&
>
s
£«* §o
a
.
». >
5
54»5«
A
Itfl
Figure 58Schottky rise and fall times at 9.5GHz
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission
104
10 to 90 percent rise and fall times. Close inspection of the plots shows that the time to
double the noise is on the order of 2ns for the Schottky diode and Ins for the FET
detector.
The tangential signal sensitivity for the Schottky diode at carrier frequencies o f 8 .6
to 9.6GHz is shown in table 7. Comparing this to TSS for the FET detector shown in table
S, the Schottky diode has a better TSS by between 13 and 24.5 dB. Schottky Tangential
Signal Sensitivity however is for a diode match to and loading the input source and if the
coupling from the input of the microstrip to the gate of the FET is taken into consideration
using figure 33 and
1055coupling=20log(atten)
(92)
Where atten comes from figure 33. The equivalent TSS is shown in table 8 . The FET has
5.9dB better TSS at 9.6GHz.
frequency
GHz
power at A
dBm
8.6
8.7
8.8
8.9
9
9.1
9.2
9.3
9.4
9.5
9.6
TSS dBm
attenuator
setting dB
0.161973535
0.236639182
0.06037955
0.220157398
0.350292822
0.374264979
0.244856677
0.094508958
0.248959601
0.064660422
0
48
48
48
48
47.5
48
48
48
48
48
48
Table 7 Schottky TSS
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
-50.838
-50.7634
•50.9396
-50.7798
-50.1497
-50.6257
-50.7551
-50.9055
-50.751
-50.9353
-51
105
Frequency GHz
Coupling Loss dB
Equivalent TSS dBm
8 .6
15.9
-34.9
8.7
16.1
-34.7
8 .8
16.4
-34.5
8.9
16.7
-34.1
9.0
17.0
-33.1
9.1
17.2
-33.4
9.2
17.5
-33.3
9.3
17.8
-33.1
9.4
18.1
-32.6
9.5
18.4
-32.5
9.6
18.7
-32.1
Table 8 Schottky equivalent TSS
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
106
Chapter V. COMPARISON OF THEORY AND RESULTS
The theory presented at beginning of this dissertation was that a negatively biased
FET could be used to make a high impedance, high speed microwave detector. A high
impedance detector does not load the circuit being monitored. The calculated impedance
of the detector developed is 2.9kQ at 9GHz as can be seen in figure 44. In comparison to
a standard impedance of SOQ for most microwave systems this is quite high. Compared to
the impedance of400Q for wave guide systems this is high. Compared to the intrinsic
impedance of377Q for free space this is still high. The measured difference between the
insertion loss of the FET detector with a SOQ microstrip line and a 50Q microstrip line
alone, shown in figure 42 was at the most 0.2dB which indicates that the FET detector
input truly is high impedance and does not load the circuit being monitored.
The FET detector developed also met the goal o f high speed. The output R-C time
constant indicates that the detector should be able to follow a pulse with a 500ps rise time.
The measured rise and fall time of the detector are on the order 2ns which is also the the
specified rise and fall time of the circuit being used to create the pulse for the
measurement. As stated in chapter IV section C the measured times are the upper bound
of the detectors rise and fall times. Since the measured rise and fall times are at the limits
of the measurement system capability the detector may be even faster than measured.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
107
Chapter VL CONCLUSIONS AND SUGGESTIONS FOR FURTHER
RESEARCH
This dissertation presented the development of a of high impedance, high speed
microwave detector based on a negatively biased field effect transistor. It covered the
development o f theory, circuit design, optimization, fabrication and experimental
measurements of the detector.
In the process of developing the detector a method o f converting the inherently
low microwave impedance of a FET to a high impedance was conceived and presented.
The circuit design had to accommodate microwave frequencies at the input of the circuit
and video frequencies at the output o f the circuit. The problem of instability also had to be
solved and as a result some guidelines for the design of the biasing circuitry were
developed and presented.
In the presentation of the performance of the detector a method o f removing the
test equipment dependancy of TSS was presented. This is an industry standard
specification for detectors that can vary depending on the equipment used to make the
measurement. A noise analysis of the detector was also performed. This analysis had to
deal with the fact that the detector not only converts noise from microwaves to the
frequency of measurement but also has output noise due to the video frequency noise at
the input o f the detector.
The goals of high speed and high impedance were met by the design developed for
this dissertation. The detector developed also has better conversion loss and conversion
efficiency than a Schottky diode.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
108
As a suggestion for further research it is possible to design a matched impedance
microwave detector based on a FET. It is expected that such a detector would have even
greater sensitivity than the detector developed in this dissertation. It is also suggested that
a wave guide mountable version of the detector be designed.
Another suggestion is that a more complete noise analysis be done taking into
consideration the multiple band nature of this detector.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
109
BIBLIOGRAPHY
Boylestad, Robert and Louis Nashelsky, Electronic Devices and Circuit Theory, Third
Edition, Prentice Hall Inc., Englewood Cliffs, NJ, 1982.
California Eastern Laboratories, RF and Microwave Semiconductors, California Eastern
Laboratories, Los Angeles, CA, 1995-96, ppl-53 to 1-59
Cheng, David K., Fields and Electromagnetic Waves, Addison Wesley Publishing
Company, Reading, MA., 1983.
Edwards, T. C ., Foundations For Microstrip Circuit Design, John Wiley and Sons,
New York, 1985.
Golio, J. Michael, Microwave MESFETs & HEMPTs, Artech House, Inc., Norwood
MA, 1991.
Gonzalez, Guillermo, Microwave Transistor Amplifiers, Analysis and Design, 2ndEdition
Prentice Hall, Upper Saddle River, NJ, 1997
Hewlett Packard Product Note 8510-8 A, Network Analysis, Applying the HP8510 TRL
calibrationfo r non-coaxial masurements, Hewlett Packard, Palo Alto, CA 1992
Ishii, T. K., Microwave Engineering, The Roland Press Company, New York, 1966.
Ishii, T. K., Microwave Engineering, Second Edition, Harcourt Brace Jananovich
Publishes Technology Publications, New York, 1989.
Ishii, T. K., "Theory of Delay Time Reduction for Deep Space Communications",
IEEE MTT/ED Societies, Milwaukee Chapter Meetings, November 18, 1987,
Milwaukee, Wisconsin.
Ishii, T. K., "Wavefront Detector NSF Proposal", Marquette University, Milwaukee,
Wisconsin, December 1988.
Kosow, Irving L., Microwave Theory and Measurement, Prentice Hall, Inc.,
Englewood Cliffs, NJ, 1963.
Lance, Algie L. Lance, Introduction to Microwave Theory andMeasurements, McGraw
Hill, New York, 1964
Liao, Samuel Y., Microwave Devices and Circuits,2nd edition, Prentice Hall, Inc.,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
110
Englewood Cliffs, NJ, 1985.
Luglio, Juan R., "Negatively Biased Field Effect Transistor Used As A Microwave
Detector", Master's thesis, Marquette University, Milwaukee, Wisconsin, 1989.
Miller, Edmond K., Time Domain Measurements in Electromagnetics, Van Norstrad
Reinhold Company Inc., New York, 1986.
Millman, Jacob, Microelectronics, Digital and Analog Circuits and Systems, McGrawHill Book Company, New York, 1985.
Montgomery, Carol G., Technique o f microwave Measurements, Volume 1,Dover
Publications, New York, 1966.
Mumford, W.W. and E.H. Scheibe, Noise Performance Factors in Communications
Systems, Horizon House-microwave Inc., Dedham, MA, 1968.
Muow, Robert B. and Fred Schumacher, "Tunnel Diode Detectors", Microwave
Journal, Vol. 9, January 1966.
The New IEEE Standard Dictionary o f Electrical and Electronics Terms, fifth edition,
chair Gediminas P. Kurpis, ed. Christopher J. Booth, New York, 1993
Oxner, Ed, "FET's in Balanced Mixers", Siliconix Applications Note AN72-1.
Stremler, Ferrel G., Introduction to Communication Systems, Addison-Wesley Publishing
Company, Reading, MA, 1977.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
111
Appendix
TRL Calibration Kit Mechanical Drawings
LU
_l
c
s:
UJ
C/0
UJ
li-
o
U i
CJ
X
_l
a
ID
Q
c
_
>
a
CK
UJ
CJ
CL
_l
UJ
a
UJ
UI
cn
m
( /)
C /0
O'
UJ
ui
_ j
<t
<
r
O'
a
i—
CJ
ui
□
CJ
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
112
co
UJ
CJ
c
UJ
Li­
ck
C
a
1
+1
L IN E
CDNTACT
1
Y
ai
a
t—
u
UJ
z
z
a
CJ
0 0
C/0
<r
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
113
ui
_i
<z
S a
CJ
CO
CO
< n
O'
a
i—
u
UJ
z
z
a
u
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
in
cu
114
vD
o
o
o
in
o
£
3
'c
£
D
_r
♦
DR. 8, TAP THRU
# 2 -5 6 SCREWS
FDR
o
in
cu
O
<Z
o
o
cu
o
L_
in
cu
cu
CK
o
UJ
to
_l
□
UI
oc
C3
in
o
o
Q J
CU
CU
CU
CJ
o
o
00
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
115
in
cu
vD
O
E
3
C
o
o
in
o’
E
3
C
t
o
THRU
FDR
in
cu
o'
cxi
a
o
o
in
cu
cu
o
Qi
Ld
cu
o
o
o
■sr
TAP
O'
CO
C^» LJ
_ _I
, 'G □
°®in x
< (X)
—• UJ
« -J
o
in
cy
a
in i
crR cu
CC
co
00
CO
UO '
in 2
o'
00
oo
CJ
co
oo
00
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
116
E
in
OJ
\D
3
‘c
o
’e
_3
CD
c
QL
UJ
o
OJ
o'
o
o
oj
in
OJ
CD
OJ
o
o
o
o
a
UJ
in
CJ
o
in
rj
a
CD
( Y
\D
on
.
oo
OJ
OJ
oo
00
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CJ
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DRILL THRU FOR
2 - 5 6 CLEARANCE
4 HOLES
2.200
DRILL & TAP . 2 5
FOR 2 - 5 6
2 HOLES
1 .6 3 7 —
0 ,3 8 3 —
0.200 —
0.1 3 9 —
0 .1 2 5
0 .3 4 0
0 .5 7 7
0 ,4 5 5 f
0 .2 3 8
DRILL & REAM .1 2 5 DEEP'
FOR S.F. 1 / 1 6 D1A DDWEL PIN2 HOLES
0 .0 9 5 0 4 4 C0PPER
TRACE
2 .0 7 5
2 .1 3 8
DIRLL THRU FOR
2 - 5 6 CLEARANCE
FEMALE LAUNCHER
— 0 .2 5 7
DEEP
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DRILL THRU FDR
2 - 5 6 CLEARANCE
4 HDLES
DRILL 8. TAP . 2 5
FDR 2 - 5 6
2 HDLES
2 .200 -
DEEP
-1 .6 3 7
0.383
r 0 ,0 9 5 0 4 4
COPPER
TRACE
0.031
0.139
0.200
0.125
0.340
1.25 0 1.12 5 -
1.000
0.5 7 7
0.063
0 .2 5 7
0 . 0 6 3 - 10 . 2 3 8
2.138
DIRLL 8. TAP . 2 5 DEEP
FDR 2 - 5 6 SCREW
L A U N C H E R MALE
DRILL 8. REAM .5 0 DEEP
FDR S . F .1 /1 6 DIA. DDWEL
2 HDLES
MATERIAL* D u roid w ith Aluminum P l a t e
oo
ZD
O'
X
o
o
OJ
CD
<c
OO
I—
LJ
Q
I
,063
O
00
CD
cu
00
CO
C\J
CD
CD
uo
cu
0 0
□
in
60
OJ
on
OJ
CD
oo
CD
CD
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
MATERIAL: ALUM.
DIA. THRU
119
X
ALUMINUM
CU
CK
CD
Ld
<n
i—i
Q
00
CD
CD
00
cu
CO
063
□
CD
CD
DN
ID
(¥.
u
u
<n
cu
j—|
^u
DURDID
DIA, THRU
PLATE
120
o
o
00
00
CD
in
cu
O'
LJ
CD
in
cu
O'
cu
CD
c5
CO
CD
CD
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
MATERIAL:
o
o’
121
1=1
I
CJ
f—
.063 DIA. THRU
2 HOLES
CD
CD
CU
T
o
<3
CD
,
CD
CO
CD
CU
CD
O
O
CU
cu
CD
CO
o
cd"
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
MATERIAL: ALUM.
5 5
DEVI CE
o
MDUNT
CD
ATTACH S I L V E R
FDILTD END
ATTACH S I L V E R
FDIL TD END
0.125 —■
0.125
.090
00.500
.090
0.112
00.500
DR. THRU FDR
2 - 5 6 CLEARANCE
2 HDLES
DR. THRU FDR
2 - 5 6 CLEARANCE
2 HDLES
T HRU C D N T A C T
LINE C D N T A C T e
M A T E R I A L 1- R e x o l i t e
MATERIAL: R e x o l i t e
RE Q 'D
123
X
Id
Id
c
□ (X
IX
Q
o
in
c
<£
t—i
_l ~
J
o
■sr
x
i ■“I
CD
SL
U
CJ
X
Qo
Ld
ZD _ J 00
X CJ LJ
X
*” kD □
. in x
(X I
O CU OJ
CJ
<r
f—
z
□
CJ
LO
CU
\D
CD
LJ
CJ
i— i
>
LJ
UJ
n
LO
CU
O
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
MATERIAL: R e x o l i t e
h"
124
CNC Circuit Board Drawings
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
125
-------------------------L-
3.269
4.400
-----------
2.200
0 . 1 2 3 — r_-
1 .1 3 1
0.123
1.239 n
0.593D.54B.550
1 .2 1 5
0.063
0.925
1.850
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
126
CNC Board Program
%BOARD CNC PROGRAM
N15G00Z1.0
N20G00G90X0Y0
N25G00Z0
N30T1M6
N35G43H1Z1.0
N40X0.0625Y0.925S2000M03
N45G00Z0.1
N50G81X0.0625Y0.925R0.1Z-0.0981F2.0
N55G80
N60G00Z1.0
N65X1.1312Y1.7875
N70G00Z0.1
N75G81X1.1312Y1.7875R0.1Z-0.0981F2.0
N80G80
N85G00Z1.0
N90X3.2688
N95G00Z0.1
N100G81X3.2688Y1.7875R0.1Z-0.0981F2.0
N105G80
N110G00Z1.0
N115X4.3375Y0.925
N120GO0Z0.1
N125G81X4.3375Y0.925R0.1Z-0.0981F2.0
N130G80
N135GO0Z1.0
N140Y0.0625
N145GO0Z0.1
N150G81X4.3375Y0.0625R0.1Z-0.0981F2.0
N155G80
N160GO0Z1.0
N165X3.2688
N170GO0Z0.1
N175G81X3.2688Y0.0625RO. 1Z-0.0981F2.0
N180G80
N185G00Z1.0
N190X1.1312
N195GO0Z0.1
N200G81X1.1312Y0.0625R0.1Z-0.0981F2.0
N205G80
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
127
N210G00Z1.0
N215X2.2Y1.7875
N220G00Z0.1
N22SG81X2.2Y1.7875R0.1Z-0.0981F2.0
N230G80
N235M09
N240G00Z1.0
N245G00X-8.0Y0.
N250G00Z0.0
(HL3)
(TOOL NUMBER:2)
(SPINDLE RPM:2000)
N270M06
N275S2000M03
N280G00Z1.0
N285X0.0625Y0.925
N290G00Z0.1
N295G83X0.0625Y0.925R0.1Z-0.4288Q0.0429F2.9
N300G80
N305G00Z1.0
N310X1.1312Y1.7875
N315G00Z0.1
N320G83X1.1312Y1.7875R0.1Z-0.4288Q0.0429F2.9
N325G80
N330G00Z1.0
N335X3.2688
N340G00Z0.1
N345G83X3.2688Y1.7875R0.1Z-0.4288Q0.0429F2.9
N350G80
N355G00Z1.0
N360X4.3375Y0.925
N365G00Z0.1
N370G83X4.3375Y0.925R0.1Z-0.4288Q0.0429F2.9
N375G80
N380GO0Z1.0
N385Y0.0625
N390GOOZ0.1
N395G83X4.3375Y0.0625R0.1Z-0.4288Q0.0429F2.9
N400G80
N405G00Z1.0
N410X3.2688
N415G00Z0.1
N420G83X3.2688Y0.0625R0.1Z-0.4288Q0.0429F2.9
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
128
N425G80
N430G00Z1.0
N435X1.1312
N440G00Z0.1
N445G83X1.1312Y0.0625R0.1Z-0.4288Q0.0429F2.9
N450G80
N455G00Z1.0
N460X2.2Y1.7875
N465G00Z0.1
N470G83X2.2Y1.7875R0.1Z-0.4288Q0.0429F2.9
N475G80
N480M09
N485G00Z1.0
N490G0OX-8.0YO.
N495G00Z0.0
(HL1)
(TOOL NUMBER:3)
(SPINDLE RPM:2000)
N515M06
N520S2000M03
N525G00Z1.0
N530X4.5392Y-0.0047
N535G00Z0.0625
N540G01Z-0.032F1.5
N545X4.4748Y-0.0288F3.0
N550G02X4.4287Y-0.0748I-0.2623J0.2166
N555G01X4.4045Y-0.1392
N560X4.3392Y-0.0825
N565G03X4.4825Y0.0605I-0.1267J0.2703
N570G00Z1.0
N575Y1.1955
N580G00Z0.0625
N585G01Z-0.032F1.5
N590X4.4725Y1.183 5F3.0
N595G02X4.4694Y1.17991-0.1975JO. 1643
N600G01Y0.1878
N605G02X4.2125Y-0.0691I-0.2569J0.
N610GO 1X2.5222
N615G02X2.5053Y-0.0825I-0.1679J0.1944
N620G01X2.43 3
N625X2.4781Y-0.0507
N630G03X2.5059Y-0.0274I-0.1239J0.176
N635G01X4.2125
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
129
N640GO3X4.4277Y0.187810.J0.2152
N645G01Y1.1961
N650G03X4.4405Y1.2101I-0.1527J0.1517
N655G01X4.4825Y1.2606
N660Y1.3257
N665X4.4085Y1.2368
N670G02X4.3861Y1.2144I-0.1335J0.111
N675G01Y0.1878
N680G02X4.2125Y0.0142I-0.1736J0.
N685G01X2.4877
N690G02X2.4542Y-0.0167I-0.1334J0.1111
N695G01X2.3607Y-0.0825
N700X2.2884
N705X2.4302Y0.0174
N710G03X2.4665Y0.0558I-0.076J0.1079
N715G01X4.2125
N720GO3X4.3445Y0.187810.JO. 132
N725G01Y1.2356
N730G03X4.3765Y1.2634I-0.0695JO. 1122
N73 5G01X4.4825Y1.3908
N740G00Z1.0
N745X4.4727Y1.8864
N750GO0ZO.0625
N755G01Z-0.032F1.5
N760G03X4.4067Y1.93071-0.1977J-0.2236F3.0
N765G00Z1.0
N770X-0.0067
N775GO0Z0.0625
N780G01Z-0.032F1.5
N785G03X-0.0727Y1.886410.1317J-0.2679F3.0
N790G01X-0.0825Y1.8142
N795G02X0.125Y1.919710.2075J-0.1514
N800GO 1X4.275
N805G02X4.4825Y1.814210.J-0.2569
N810G01Y1.7336
N815X4.4618Y1.7698
N820G03X4.275Y1.87811-0.1868J-0.107
N825G01X0.125
N830G03X-0.0618Y1.769710.J-0.2153
N835G01X-0.0825Y1.7336
N840Y1.6498
N845X-0.0257Y1.749
N850G02X0.125Y l.836410.1507J-0.0862
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
130
N855GO 1X4.275
N860G02X4.4257Y1.749110J-0 .1736
N865G01X4.4825Y l.6499
N870Y1.5661
N875X4.3895Y1.7284
N880G03X4.275Y1.7948I-0.1145J-0.0656
N885G01X0.125
N890G03X0.0104Y1.7284I0.J-0.132
N895G01X-0.0825Yl .566
N900G00Z1.0
N905Y1.1956
N910GOOZ0.0625
N915G01Z-0.032F1.5
N920X-0.0728Y1.1839F3.0
N925G03X-0.0694Y1.179910.1978JO. 1639
N930G01Y0.6475
N935G03X-0.0825Y0.63110.1944J-0.1679
N940G01Y0.5586
N945X-0.052Y0.602
N950G02X-0.0277Y0.631210.177J-0.1225
N955G01Y1.1961
N960G02X-0.0408Yl .210510.1527J0.1517
N965G01X-0.0825Yl .2608
N970Y1.3261
N975X-0.0087Y1.237
N980G03X0.0139Y1.2144I0.1337J0.1108
N985G01Y0.613
N990G03X-0.0179Y0.5782I0.1111J-0.1334
N995G01X-0.0825Y0.4862
N1000Y0.4138
N1005X0.0163Y0.5544
N1010G02X0.0555Y0.5918I0.1087J-0.0749
N1015G01Y1.2356
N1020G02X0.0234Y1.2636I0.0695JO. 1 122
N1025G01X-0.0825Y1.3913
N1030GOOZ1.0
N1035X-0.0729Y-0.0356
N1040GOOZ0.0625
N1045G01Z-0.032F1.5
N1050G03X-0.0039Y-0.081410.1979J0.2234F3.0
N1055G01X-0.0825Y0.0355
N1060X-0.0736Y0.0249
N1065G03X0.125Y-0.069110.1986J0.1629
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
131
N1070GO1X1.8778
N1075G03X1.8947Y-0.0825I0.1679J0.1944
N1080GO1X1.967
N1085X1.9219Y-0.0507
N1090G02X1.8941Y-0.0274I0.1239J0.176
N1095GO1X0.125
N il 00GO2X-0.0415Y0.051410J0 .2 152
N1105GO1X-0.0825Y0.1006
N il 10Y0.1656
N1115X-0.0094Y0.0779
N1120G03X0.125YO.014210.1344J0.1099
N1125G01X1.9123
N1130G03X1.9458Y-0.016710.1334J0.1111
N1135G01X2.0393Y-0.0825
N1140X2.il 16
N1145X1.9698YO.0174
N1150G02X1.9335Y0.0558I0.076J0.1079
N1155G01X0.125
N1160G02X0.0227YO. 104410.JO. 132
N1165G01X-0.0825Y0.2307
N1170X-0.1448Y0.3674F2.0
N1175X-0.0825Y0.3964
N1180X0.0245Y0.5487
N1185G02X0.064Y0.5852IO. 1005J-0.0692
N1190G03X0.0655Y0.5878I-0.0015J0.0026F0.8
N1195G01Y1.2395F2.0
N1200G03X0.064Y1.2421I-0.003JO.F0.8
N1205G02X0.0311Y l.2710.061JO. 1057F2.0
N1210G01X-0.0756Y1.3987
N1215G02X-0.0825Y1.4079I0.0939J0.0778
N1220G01X-0.1444Y1.4377
N1225X-0.1419Y1.5113
N1230X-0.0825Y1.5459
N1235X0.0191Y1.7234
N1240G02X0.125Y1.784810.1059J-0.0606
N1245G01X4.275
N1250G02X4.3809Y1.7234IO.J-0.122
N1255GO 1X4.4825Y l.546
N1260X4.5419Y1.5114
N1265G00Z1.0
N1270X2.1579Y-0.1449
N1275GO0Z0.0625
N1280G01Z-0.032F1.5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
132
N1285X2.129Y-0.0825F2.0
N1290X1.9755Y0.0255
N1295G02X1.9401Y0.0643I0.0702J0.0998
N1300G03X1.9375Y0.0658I-0.0026J-0.0015F0.8
N1305G01X0.125F2.0
N1310G02X0.0304Y0.110710.JO. 122
N1315G01X-0.0799Y0.2432
N1320G02X-0.0825Y0.2464I0.0937J0.0781
N1325G01X-0.1462Y0.2723
N1330GOOZ1.O
N1335X4.5445Y1.4369
N1340GOOZ0.0625
N1345G01Z-0.032F1.5
N1350X4.4825Y l .4073F2.0
N1355G02X4.4756Y1.3982I-0.1006J0.069
N1360G01X4.3688Y1.2698
N1365G02X4.336Y1.2421I-0.0938J0.078
N1370G03X4.3345Y1.2395I0.0015J-0.0026F0.8
N1375G01YO. 1878F2.0
N1380G02X4.2125Y0.0658I-0.122J0.
N1385G01X2.4625
N1390G03X2.4599Y0.0643I0.J-0.003F0.8
N1395G02X2.4245Y0.0255I-0.1057J0.061F2.0
N1400G01X2.271Y-0.0825
N1405X2.2421Y-0.1449
N1410G00Z1.0
N1415M09
N1420G00Z1.0
N1425GO0X-8.0Y0.
N1430GOOZO.O
(SD1DUR0ID)
(TOOL NUMBER:4)
(SPINDLE RPM:2000)
N1450M06
N1455S2000M03
N1460G00Z1.0
N1465X1.95Y0.6
N1470GO0ZO.1
N1475G81X1.95Y0.6R0.1Z-0.03F1.1
N1480G80
N1485GO0Z1.O
N1490X2.05Y1.22
N1495G00Z0.1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
133
N1500G81X2.05Y l .22R0.1Z-0.03F1.1
N1505G80
N1510GOOZ1.O
N1515X2.4Y0.55
N1520G00Z0.1
N1525G81X2.4Y0.55R0.1Z-0.03F1.1
N1530G80
N1535M09
N1540GOOZ1.O
N1545G00X-8.0Y0.
N1550GOOZO.O
(HL2)
(TOOL NUMBER:5)
(SPINDLE RPM.-2000)
N1570M06
N1575S2000M03
N1580GOOZ1.O
N1585X1.95Y0.6
N1590G00Z0.1
N1595G83X1.95Y0.6R0.1Z-0.3641Q0.0036F1.4
N1600G80
N1605G00Z1.0
N1610X2.05Y1.22
N1615G00Z0.1
N1620G83X2.05Y1.22R0.1Z-0.3641Q0.0036F1.4
N1625G80
N1630GO0Z1.O
N1635X2.4Y0.55
N1640G00Z0.1
N1645G83X2.4Y0.55R0.1Z-0.3641Q0.0036F1.4
N1650G80
N1655M09
N1660G00Z1.0
N1665G00X-8.0Y0.M05
N1670GOOZO.O
N1675M2
%
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
134
CNC Cover Drawings
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission
135
4.400
3 .263
1 .1 0 9
0.06
1.850
$5a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
136
CNC Cover Prqgram
%COVER CNC PROGRAM
N15G00Z1.0
N20G00G90X0Y0
N25G00Z0
N30T1M6
N35G43H1Z1.0
N40X1.447Y0.8845S2000M03
N45G00Z0.1
N50G01X0.947Z-0.027F18.0
N55X1.447Z-0.154
N60X0.947Z-0.281F20.0
N65X3.453
N70Y0.903
N75X2.565
N80G02X2.1864Y1.0041IO.J0.7595
N85X1.9375Y0.96551-0.2489J0.7834
N90G01X0.947
N95Y0.8845
N100Y0.718
N105X3.6196
N110G02X3.6195Y0.73I0.5929J0.012
N115G01Y1.0695
N120X2.565
N125G02X2.2254Y1.1764I0.J0.593
N130X2.2Y1.186910.2371J0.6111
N135X1.9375Y1.132I-0.2625J0.6006
N140G01X0.7805
N145Y0.73
N150G02X0.7804Y0.718I-0.593J0.
N155GO 1X0.947
N160Y0.5515
N165X3.8252
N170G02X3.786Y0.73I0.3873J0.1785
N175G01Y1.236
N180X2.565
N185G02X2.3047Y1.3247IO.J0.4265
N190X2.2233Y1.361I0.1578J0.4628
N195G01X2.1767
N200G02X1.9375Y1.2985I-0.2392J0.4265
N205G01X0.6016
N210G02X0.614Y1.19631-0.4141J-0.1022
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
137
N215G01Y0.73
N220G02X0.5748Y0.5515I-0.4265J0.
N225G01X0.947
N230Y0.385
N23 5X4.0775
N240Y0.4199
N245G02X4.0198Y0.5555I0.26J0.1908
N250X3.9525Y0.73I0.1927J0.1745
N255G01Y1.4025
N260X2.565
N265G02X2.3 856Y1.474310.J0.26
N270X2.2717Y1.5275I0.0769J0.3132
N275G01X2.1283
N280G02X1.9375Y1.4651-0.1908J0.26
N285G01X0.3225
N290Y1.4533
N295G02X0.3574Y1.3931I-0.26J-0.1908
N300X0.4475Y1.19631-0.1699J-0.1968
N305G01Y0.73
N310G02X0.3802Y0.5555I-0.26JO.
N315X0.3225Y0.41991-0.3177J0.0553
N320G01Y0.385
N325X0.947
N330G00Z1.0
N335M09
N340GO0Z1.O
N345G00X-8.0Y0.
N350GOOZO.O
(PK1_MAINPKT)
(TOOL NUMBER:2)
(SPINDLE RPM:2000)
N370M06
N375S2000M03
N380GO0Z1.O
N385X0.947Y0.2025
N390G00Z0.0625
N395G01Z-0.281F1.5
N400X0.9525Y0.1962F3.0
N405G03X0.9754Y0.187510.0229J0.0256F1.2
N410G01X4.275F3.0
N415Y0.5025
N420G02X4.2125Y0.6108I0.0625J0.1083
N425G01Y0.6675
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
138
N430G02X4.15Y0.73I0.J0.0625
N435G01Y1.6
N440X2.565
N445G02X2.5025Y1.6625IO.J0.0625
N450GO1X2.4625
N455G02X2.3542Y1.725I0.J0.125
N460G01X2.0458
N465G02X1.9375Y1.6625I-0.1083J0.0625
N470G01X0.125
N475Y1.3708
N480G02X0.1875Y1.2625I-0.0625J-0.1083
N485G01Y1.2588
N490G02X0.25Y1.196310.J-0.0625
N495G01Y0.73
N500G02XO. 1875Y0.6675I-0.0625J0.
N505G01Y0.6108
N510G02X0.125Y0.5025I-0.125JO.
N515G01Y0.1875
N520X0.9754
N525X1.0754
N530G03X1.0983YO. 196210.J0.0344
N535GO1X1.103 8Y0.2025
N540GOOZ1.O
N545X0.195Y0.9527
N550GOOZO.O625
N555G01Z-0.03F1.5
N560X0.1939Y0.9554F3.0
N565G03X0.1875Y0.9527I-0.0027J-0.0027F 1.2
N570G01Y0.5972F3.0
N575X0.1975
N580Y1.3083
N585X0.1875
N590Y0.9527
N595Y0.8527
N600G03X0.1928Y0.8448I0.0086JO.
N605G01X0.1961Y0.8441
N610GOOZ1.0
N615X4.2025Y l.6267
N620GO0Z0.0625
N625G01Z-0.03F 1.5
N630Y1.6525F5.0
N635X4.1767
N640G02X4.2025Y1.6267I-0.0367J-0.0625
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
139
N645G01X4.2067Y1.6319F3.0
N650G03X4.2125Y1.651I-0.0286J0.019F1.2
N655G01Y1.6625F3.0
N660X2.525
N665Y1.6525
N670X4.14
N675GO2X4.2025Y1.59I0.J-0.0625
N680G01Y0.6733
N685X4.2125
N690Y1.651
N695Y1.6625
N700X4.124
N705G03X4.105Y l.656710.J-0.0344
N710G01X4.0998Y1.6525
N715G00Z1.0
N720X2.233Y1.7566
N725G00Z0.0625
N730G01X2.108Z0.0033F1.5
N735X2.233Z-0.0558F4.2
N740X2.108Z-0.115F5.0
N745X2.292
N750X2.2Y1.8213
N755X2.108Y1.7566
N760X2.1132Y1.7523F3.0
N765G03X2.1322Y1.7466I0.019J0.0286F1.2
N770GO1X2.322F3.0
N775Y1.7477
N780X2.2Y1.8336
N785X2.0764Y1.7466
N790X2.1322
N795X2.2322
N800G03X2.2512Y1.752310.J0.0344
N805G01X2.2564Y1.7566
N810G00Z1.0
N815X0.1225Y0.3901
N820GO0ZO.0625
N825G01X0.0457Y0.3462Z0.027F1.5
N830X0.1225Y0.3901Z-0.0085F4.3
N835X0.0457Y0.3462Z-0.044
N840X0.1225Y0.3901Z-0.0795
N845X0.0457Y0.3462Z-0.115F5.0
N850X0.0818Y0.3668
N855X0.0867Y0.3584
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
140
N860Y0.3753
N865X0.0809Y0.3684
N870X0.0818Y0.3668
N875X0.0457Y0.3462
N880X0.1283Y0.2018
N885Y0.4908
N890X0.0305Y0.3728
N895X0.0457Y0.3462
N900X0.0446Y0.3395F3.0
N905G03X0.0491Y0.3202I0.0343J-0.0023F1.2
N910G01X0.125Y0.1875F3.0
N915X0.1383
N920Y0.5017
N925X0.1243
N930X0.0184Y0.3738
N935X0.0491Y0.3202
N940X0.0987Y0.2334
N945GO3X0.1132Y0.2197I0.0298J0.0171
N950G01X0.1195Y0.2173
N955G00Z1.0
N960X4.2805Y0.3518
N965GOOZ0.O625
N970GO1Y0.4768Z0.0033F1.5
N975Y0.3 518Z-0.0558F4.2
N980Y0.4768Z-0.115F5.0
N985Y0.2173
N990X4.3697Y0.3 73
N995X4.2833Y0.4768
N1000X4.2805
N1005X4.2763Y0.4716F3.0
N10 10G03X4.2705Y0.4526I0.0286J-0.019F1.2
N1015G01YO. 1875F3.0
N1020X4.275
N1025X4.3818Y0.3741
N1030X4.288Y0.4868
N1035X4.2705
N1040Y0.4526
N1045Y0.3526
N1050G03X4.2763Y0.3336I0.0344JO.
N1055G01X4.2805Y0.3284
N1060GO0Z1.0
N1065X0.1193Y l .4968
N1O70GO0Z0.0625
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
141
N1075G01X0.0489Y1.5555Z0.027F1.5
N1080X0.1193Y l .4968Z-0.0085F4.3
N1085X0.0489Y1.5555Z-0.044
N1090X0.1193Y l .4968Z-0.0795
N1095X0.0489Y1.5555Z-0.115F5.0
N1100X0.0809Y1.5288
N1105X0.0788Y1.5264
N il 10X0.0883Y l.5129
N1115Y1.5377
N1120X0.0809Y1.5288
N1125X0.0489Y1.5555
N1130X0.0264Y1.5285
N1135X0.1241Y1.3895
N1140X0.1299
N1145Y1.6519
N il 50X0.0489Y1.5555
N1155X0.0423Y l .5542F3.0
N1160G03X0.0257Y1.5432I0.0098J-0.0329F1.2
N1165G01X0.0138Y1.529F3.0
N1170X0.1189Y1.3795
N1175X0.1399
N1180Y1.6619
N1185X0.1245
N1190X0.0257Y1.5432
N1195X0.0138Y1.529
N1200X0.0607Y1.4623
N1205G03X0.0763Y1.4501I0.0281J0.0198
N1210G01X0.0828Y1.4482
N1215G00Z1.0
N1220M09
N1225G00Z1.O
N1230GOOX-8.0Y0.
N1235G00Z0.0
(PK1_GND1)
(TOOL NUMBER:3)
(SPINDLE RPM:2000)
N1255M06
N1260S2000M03
N1265G00Z1.0
N1270X0.0625Y0.925
N1275G00Z0.1
N1280G81X0.0625Y0.925R0.1Z-0.0756F1.5
N1285G80
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
142
N1290G00Z1.0
N1295X1.1312Y1.7875
N1300G00Z0.1
N1305G81X1.1312Y1.7875R0.1Z-0.0756F1.5
N1310G80
N1315GOOZ1.0
N1320X3.2688
N1325G00Z0.1
N1330G81X3.2688Y l .7875R0.1Z-0.0756F1.5
N1335G80
N1340GO0Z1.0
N1345X4.3375
N1350GO0Z0.1
N1355G81X4.3375Y1.7875R0.1Z-0.0756F1.5
N1360G80
N1365GOOZ1.0
N1370Y0.925
N1375G00Z0.1
N1380G81X4.3375Y0.925R0.1Z-0.0756F1.5
N1385G80
N1390GOOZ1.0
N1395X3.2688Y0.0625
N1400GOOZ0.1
N1405G81X3.2688Y0.0625R0.1Z-0.0756F1.5
N1410G80
N1415GOOZ1.0
N1420X2.2
N1425G00Z0.1
N1430G81X2.2Y0.0625R0.1Z-0.0756F1.5
N1435G80
N1440G00Z1.0
N1445X1.1312
N1450GO0Z0.1
N1455G81X1.1312Y0.0625R0.1Z-0.0756F1.5
N1460G80
N1465M09
N1470G00Z1.0
N1475GO0X-8.0Y0.
N1480GOOZO.O
(HL1)
(TOOL NUMBER:4)
(SPINDLE RPM:2000)
N1500M06
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
143
N1505S2000M03
N1510G00Z1.0
N1515X0.0625Y0.925
N1520GOOZO.1
N1525G83X0.0625Y0.925R0.1Z-0.421Q0.0421F2.1
N1530G80
N1535G00Z1.0
N1540X1.1312Y1.7875
N1545G00Z0.1
N1550G83X1.1312Y1.7875R0.1Z-0.421Q0.0421F2.1
N1555G80
N1560G00Z1.0
N1565X3.2688
N1570G00Z0.1
N1575G83X3.2688Y1.7875R0.1Z-0.421Q0.0421F2.1
N1580G80
N1585G00Z1.0
N1590X4.3375
N1595G00Z0.1
N1600G83X4.3375Y1.7875R0.1Z-0.421Q0.0421F2.1
N1605G80
N1610G00Z1.0
N1615Y0.925
N1620GOOZO.1
N1625G83X4.3375Y0.925R0.1Z-0.421Q0.0421F2.1
N1630G80
N1635GO0Z1.0
N1640X3.2688Y0.0625
N1645G00Z0.1
N1650G83X3.2688Y0.0625R0.1Z-0.421Q0.0421F2.1
N1655G80
N1660GO0Z1.0
N1665X2.2
N1670GO0ZO.1
N1675G83X2.2Y0.0625R0.1Z-0.421Q0.0421F2.1
N1680G80
N1685GO0Z1.0
N1690X1.1312
N1695G00Z0.1
N1700G83X1.1312Y0.0625R0.1Z-0.421Q0.0421F2.1
N1705G80
N1710M09
N1715G00Z1.0
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
144
N1720G00X-8.0Y0.M05
N1725GOOZO.O
N1730M2
%
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
145
FET Rise and Fall Time Data
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
146
c c «
„.s
ww •
S
o
§
• • **
1777
■^
0
M
A
1 0 DV
g
w
V
w
i
?
«:t =
^o>a
o<
U
l
>
i*
ii
F£V
S.fc GHz
Q.3
ao
?
0 .0
3 *0
CJ
a
co»
i“
0
ae
<r
i-
U
al
«i9 aa
0 a
<4. O
«4. O
■
Vo
aa
in
o ini
•
•
i i
, >M
°U >
«>
I
-V
>~ s> 9 o9 _Oa f
*
«b* a* -S'-S’
U B a01 _c
ui ix
a
°
in in
■"
a0a
CO
1
§
.
I
I
. • ' • m
o
■ >^ «^ i o•
a x a « o
x a » ■»
t. •O) • *•»5 .1
3> l
°\2 8 ,§ 5
<§£
5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
147
oin S
«
<D .
a.
13
<
rJ
to
fUl
U.
u
X >
a
at
u■5i
in
r»
9
•«
r>
•
■ m nwnpr Further reproduction prohibited without perm ission.
R eproduced with perm ission of the copyright owner. Further repr
148
i i s
n
o n !
C
oo
o
•
• —n —
!i o o o
S ------
5k w
U>
4#
(0
I
U
o<l
s
>
<
u
II
i-s
c
£
0 .0
3
o
C
l
a
u
?
««
r
J3 «
o£~
as
Ul
►-
zUl
u
c
a
a
a
ei
•
O) Of
uU
l ^Ul
Hot doff
in
- 40.000
c
ao
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
149
oo
3■ W0 W0 WO
to
(I
il
H i
?
X§
r«
ua
?
a.
Io
U
■ 8-
oc
«fca
4D
»
ut
m
u
*9>O
a
»Q
°uiI
V,
X. >
4> 8
C Q
«n in
mai
Ul UJ
«> ^• cni aa
(71•
•
R eproduced with permission
n fih e copyright owner. F u rtn e n e p
P*
150
C 0 .1
lOBlO
• NBh
•
•
s■• A• A• A•
i « +*# ?+*
9
V
J l ,
o M flL
-r
‘
«••IU«O>&
U
ul
Ul
il
“■s
?
r*
U
O)
c
u.
■g -
£ "*
U
ul
&
4o>
%
u
*8
• 9
in
\
i
>v>
*»
«< a
|S
cn in
is
0) 0)
*0 D 5?
U
l Ul #A • Q
•^• «0>*'^•i NI OO
O 4* • t o
x o>> «r
O)
?ii
••OL)£*9> 8, 8 $
”62
£ 4
Reproduced w ith p e n s i o n o, the copyright owner. Further reproduction p ro h M e d without permission.
151
nS
m
“• VSI "•
4•0 NOO
S. . .
S m^ a
i■ ^u w
o o^
«•
«•
40 4
»
*t »i. £s .
>
o
11
m
c
0.0
3^
o
u
O
)
c
Q.
3O
O
4o>
0
4* O
&o
• 6o
«*.
(4 o
° «I
> v >
-> 5v,
I*
«•
• O
•C O
0 ) CD
Ul Ul
in in
Reproduced wrth permission
C the
copyright owner. Further reproduction prohihited without permission.
152
•c •a -«
c
S
8.2
” cw
«n
5<«
I • ■•
Oo o
J -----V
A
A
A
«*«•
■x (L
•
-
A
I
0 CL
L V >
U
o< l
<
fcl
U
>
<
Si
g-s
?
V5°.
^ at
U
a
<3^
V”
£
?
r
o•
£ **
J )
>
m
z
U
l
u
■
o
\
A
£ 8
•
° in
E
X>
A
a
in
_
■
SB.
5
cn in
o
T
I
6
5
U
_
_
O ®
m
v
ui ui
. a •
>
I
8Q
_
C
8
cm a
> '* i o
T0•*««•
I IO
* 8 ,8 -*
a> l c «*■
5*2 8,8$
£ -5 * 5 2
5£ 2
•
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner
-8 .0 0 0
na
Further reproduction prohibited without permission
Chonnal 1
S. 00 n a/d!v
S a n a ltlv lty
5 .0 0 aV /d lv
CENTER
Q ffa a t
-5.00000 aV
RR
F £T 9.2G ,H z
Proba
lil
10.000
Coupllog
de
Trigger Hodai Edga
On tha Nagatlva Edga o f Channoll
T rlg g ar L av al(a )
Chonnal 1 - -2.50000 aV (no lee r a ja o t OFF. coupling OC)
H o idoff ■ 40.000 no
no
lapadanca
50
AVERACE - 84
44.000
r la a tla a
f o i lt la o
Vp-p
na
Naoauraaanta
(e l)
2.308
na
(e l)
088
pi
12. 1875 aV
(e l)
Ul
w
154
C C •
C
IV £j
so
J. ■ •
•
^
J333
S i
fLo«►>a .
s<
s5
11
Is
c
?
ur *
O
a
t
4~
u
■8
(C
uT
&
o
cc
U
k-l
2
U
ul
>
5\
c
oa
o>a
ui ui
in
c
a
oo
CD
I
r Further reproduction prohibited without permission.
Reproduced w«h p e n s i o n of >de copyriph, owner. Fodder repr
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
F E T 9 .5 GHz
-0 .0 0 0
na
Chonnal 1
5 .00 n a/div
S a n a ltlv lty
5 .00 aV /d lv
CENTER
O ffa s t
-15.0000 aV
RR
Proba
la I
19.000
Coupling
do
Triggar Modai Edga
On tha Nagotlwa Edga o f Chonnal)
Triggar Laval(a)
Chonnal1 ■ -2.50000 aV (noiaa ra ja o t OFF. coupling DC)
Hoi d o ff - 40.000 na
na
lapadonca
50
AVERAGE - 04
44.000
r la a t la a
fo lltla a
Vp-p
na
Maoauraaanta
(e l) “ 1.0BB
na
te l) - 1.BB7
na
te l) - 33.1250 m\
156
uoo
W
W
W
>
o
II
£-3
it
01
c
§-«
uo
g*
so
o
Ji «
o•
£~
Ul
u
cn o)
ui ui
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
157
Schottky Diode Rise and Fall Time Data
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Scho1+ky
-44.000
na
Chonnal 1
3 .0 0 n a /d lv
S a n a ltlv lty
S.00 aV /d lv
CENTER
O ffa a t
9.00000 aV
OR
-10.000
Proba
li 1
Coupling
do
Triggar Nodai Edga
On tha Nagofclva Edga o f Chonnal1
Triggar Laval(a)
Chonnal 1 - 3.75000 aV (nolaa r a ja o t OFF. ooupllng DO
Ho I d o ff - 40.000 na
na
lapadonca
SO oha
AVERAGE - 04
0.000
r la a t ia a
fa lltla a
Vp-p
na
Haoouraaanta
(o il - 3.100
na
(o il - 2.574
na
( a ll - 21.0930 a>
159
!
I
D vU O
<9
AVERAGE - 04
m
8
?
ao
r
U
a
i9 -*
r
S. 00 na/dlv
CENTER
RR
*10.000
na
I2-31
o
*
>.>
9
•«
>>
\u
—
tf 8
Is.
in n
%
..
in m i.
o c
-o *d
in
r* _
,»o
va
a n
a
ui ui
•
• >w o
44.000
n»
83 a ■ d
* 8 ,5 - ;
fc5J
« fc s* Jt'
1 1 2S
i- *4C*L5*6
a »- xO
Pi.rthpr reoroduction prohibited without permission.
Reproduced with permission of the copyright owner. Further re p ro d u ce
160
-3
c*»l z
C
^ jm
riS
■•
I: 33
s
i£
Ul
IB
U«
>
l«
§*s
g
^ d
Oo g
r
0 .0
J'
?
o-
4i a
>» >
*» > **
2• g* «
111ui
UJA 8r*• O
—
5
3
«
««* io_t4»z 9 I Oi•
1 Hiii
|
a
.b - J S S
St 2
Reproduced with permission o. the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
S c.h o ii
-44 .0 0 0
na
Channel 1
5 .0 0 n a/d lv
S a n a ltlv lty
5 .0 0 oV/div
CENTER
O ffaat
9.00000 oV
RR
-19.000
Proba
Coupling
do
111
Trlggac Modai Edge
On the Nagatlva Edga o f Chonnall
Trlggar Laval (a)
Chonnall “ 3.75000 oV (nolaa re je o t OFF. coupling OD
H o idoff - 40.000 no
no
lopadoneo
50
AVERAGE • 84
6.000
Naoauraaanto
r ia a t la a
(o l) ■ 2.970
na
fo llfelo a
(o l) “ 2.591
na
Vp-p
(o l) ■ 15.0000 oV
Os
CD
-o
o
Q .
C
o
CD
Q.
■a
CD
iission of the copyright owner. Further reproduction prohibited without permission.
S c h o tt* / 1 0 6 fa.
-44 .00 0
to
Chormal 1
5 .0 0 TO/div
S a n a ltlv lty
5 .0 0 aV /dlv
CENTER
O ffa a t
0.60000 aV
RR
-18.000
Proba
lil
Coupling
do
Trlggar Nodai Edga
On tha Nagotlva Edga o f Chonnall
Trlggar L aval(a)
Chonnall ■ 3.75000 aV (nolaa ra jo o t OFF. ooupllng 0 0
Hoi d o ff * 40.000 na
na
lapadcnoa
50 aha
AVERAGE - 64
6 .0 0 0
r la a t la a
f o lltla a
Vp-p
(o il * 5.102
na
to ll * 2.530
na
(e l) - 17.6125 aV
On
hi
163
c c m
x8
S 55 .
0
I"
‘
5ooo
9
a
a
a
L
• •
«
M
i
W
M
1 SS
«•
«•
lie
i. <•■>
O
<
s-s
ftl
coupling
QC)
0-3
3
rajaot
OFF.
s-
40»gS
nV (noloa
<ri
v
»*>
«> v5
3m t ^
a t o»
■o *u
Ul III
I*. • •
cn m
■8
*
o *»
Trlggar L aval(a)
Chonnall ■ 5.00000
Hoi doff > 40.000
na
OQ
* s.
0»
0) 0
•»
L £«•
in
5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission
zs
^ r> 1
w• w ^ •
cm
O
•
I O
L
| . ..
SIS
5S wo o vo
II
!» L
Ill
u
>
if
is
o>o
3’
ua
?
vn
o•
M»
ft
t!
u
•H
sc
<•
«s*
I I
X. >
* I
«• TS
>S
»
mm naa ii
*xi *u
m m
y O
u ji »u i
H
ton
|
go
>j;
•r*' -
5
i th is
Reproduced with permission of the copyright ow ne r Further reproduction prohibited without permission.
165
i n
a—n n!
i ririJ:
• I I
A
#«•••<
SI
« •
•»
5*L ?V- >e£
S
i
f |
< s
\r c
m
ll
coupling
I
o
OFF.
k~
rojoot
S
oV (noloo
s
u
t)
2
s
m
Uf Ul
cnn
3.75000
^ '
Q
-S
o
V/>
g
DC)
0 .0
Reproduced with permission of the copyright owner. Further reproduction prohibited w ithout permission.
166
I
ri «2
♦
i
0
0
-S
z
a
>
§
in
»\
u ut
• • n
>
2
I
w
167
n a
••
* fi
n
tola*
o !l
rajaot
OFF.
coupling
DO
>
>s
m
Trlggar Laval (a)
Chonnall ■ 3.75000
Hoi doff - 40.000
no
aV {nolaa
111
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission
$ c.k o ltt/ 9 * 4 Hz.
-44.000
na
C h on nall
S. 00 n a/dlv
S a n a ltlv tty
3 .00 aV /dlv
CENTER
O ffa a t
0.60000 aV
RR
-10.000
Proba
la 1
Coupling
do
Trlggar Modoi Edga
On th a Nogotlva Edga o f Chonnall
Trlggar Laval (a)
Chonnall ■ 3.73000 aV (nolaa r a ja o t OFF. coupling OD
Hoi d o ff - 40.000 na
na
AVERAGE - 04
SO
6.000
r la a t la a
fo llt la a
vp-p
na
Haoauraaanto
(ol) ■ 3.066 na
( o l ) “ 2.060
na
(o l) “ 14.8430 aV
Marquette University
This is to certify that we have examined
this copy of the
dissertation by
Juan R. Luglio, B.S., M.S.
and have found that it is complete
and satisfactory in all respects.
The dissertation has been approved by:
Dissertation Director, Department of Electrical and Computer Engineering
Committee Member
Committee Member
Committee Member
Committee Member
y '
Approved on Afo vevu b er
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
IMAGE EVALUATION
TEST TARGET (Q A -3 )
t>.
k
v.
w
%
150mm
IIS/MGE. In c
1653 East Main Street
Rochester, NY 14609 USA
Phone: 716/482-0300
Fax: 716/288-5989
0 1993. Applied Image. Inc.. All Rights Reserved
with permission of the copyright owner. Further reproduction prohibited without permission.
Документ
Категория
Без категории
Просмотров
0
Размер файла
4 605 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа