An Introduction to
The Design and Testing of Microwave Amplifiers
LABORATORY MANUAL
for the
EE 723 Adjunct Microwave Lab oratory
Patrick Roblin
Asso ciate Professor
The Ohio State University
2003 (12th Edition)
Contents
Acknowledgements 4
1 INTRODUCTION 5
1.1 Lab oratory Overview ...... 5
1.2 The Lab oratory Facilities ...... 5
1.3 Lab oratory Policies ...... 6
1.4 Lab oratory Schedule and Rep ort ...... 6
2 LABORATORY 1:
Microwave Measurement of a Microwave Ampli er 8
2.1 Preparation for Lab 1 ...... 8
2.2 PART I: Frequency and Gain Compression Measurement ...... 8
2.3 PART I I: Measurement of the Noise Figure ...... 12
3 LABORATORY 2:
Measurement of the Scattering Parameters of a Microwave Transistor 16
4 LABORATORY 3:
Design of a Microwave Ampli er 18
4.1 Design ...... 18
4.1.1 Imp ortant Design Constrains: ...... 18
4.1.2 Design Pro cedure: ...... 18
4.1.3 Summary of the Design rules asso ciated with the Fabrication . . . . . 21
4.2 Fabrication Pro cedure ...... 23
4.3 Measurement of Your Circuit...... 24
4.4 Write Up ...... 25
4.4.1 The Lab oratory Rep ort ...... 25
4.4.2 Poster Page ...... 25 2
5 APPENDICES 27
5.1 How to Use ADS ...... 27
5.2 How to Use the Layout Tool ...... 28
5.3 How to Use LineCalc ...... 30
5.4 Tips for Soldering ...... 31
5.5 Caring Ab out Connectors ...... 33
5.6 AT42085 S-Parameters and Noise Parameters ...... 34
5.7 Computer Data Acquisition for 2-Port S-parameters ...... 35
5.8 Loading S-parameter data in ADS and MATLAB ...... 37
5.9 Active Biasing Circuit ...... 41
5.10 Lab oratory Rep ort Forms ...... 43 3
Acknowledgements
This lab oratory would not have been p ossible without the help and supp ort of many
p eople whose contributions I would liketoacknowledge.
Most of all I am very grateful to the Hewlett Packard/AgilentTechnologies Company
for donating our rst Network Analyzer (now retired), the Sp ectrum Analyzer HP8592A and
Noise Figure meter HP8970B. Particular I would like to thank our famous OSU alumini Dr.
John Moll of HP Lab oratories as well as Howard Boyd, Tom Neb el, and Je Skokal of the
lo cal HP sale oce for their active supp ort.
The Network Analyzer HP8753C currently used in the lab was acquired thanks to a
National Science Foundation education equipment grant.
Several parts and supplies have been generously donated by the microwave industry.
I would like to thank the Rogers Corp oration for the donation of microstrip circuit b oards;
and Agilent/Avantek for the donation of microwave transistors.
The rst demonstration microwave ampli er was develop ed for this lab oratory by
Truong X. Nguyen during the course of indep endent study pro jects.
I am most grateful to my graduate students Sung Cho on Kang, Chih Ju Hung, Vakur
Erturk and Sira j Akhtar for their dedicated contribution to the development of this lab ora-
tory since 1992.
Finally I would like to thank Professors. H.C. Ko, B. Mayan, D. Ho dge, and Y. Zheng
for the full supp ort provided to this pro ject as chairman of the EE department. Prof. Klein
on-going supp ort is also acknowledged.
Patrick Roblin
January 2003 4
1. INTRODUCTION
1.1 Lab oratory Overview
The EE 723 adjunct microwave lab oratory is designed to run in parallel with EE 723. An
intro duction to microwave ampli er and oscillator design is given in the EE 723 lecture. In
supp ort to the theory, the simulation and optimization of the microwave ampli ers studied
are p erformed using an advanced microwave circuit simulator (ADS) in our ER4 computer
facilities. To complete the learning cycle the EE 723 adjunct microwave lab oratory provides
the EE 723 students with the opp ortunity to exp erimentally verify the device principles and
design theory of the microwave devices and circuits studied. The two main ob jectives of the
lab oratory are therefore:
1. To intro duce the EE 723 student to the techniques and principles of microwave mea-
surement with a mo dern Network Analyzer.
2. To involve the EE 723 student in the design of an ampli er. The pro ject includes
rst principle design, CAD simulation and optimization, CAD layout, fabrication of a
prototyp e, and testing.
Note that the students who are interested in pursuing a design pro ject of their own will have
the opp ortunity to do so under a follow-up indep endent study. Successful design pro jects
will b e used as demo in the microwave lab oratory.
1.2 The Lab oratory Facilities
The measurement and fabrication facilities of the Education Microwave Lab oratory are lo-
cated in CL 305
The measurement station consists of one HP 8753C Network Analyzer, the HP 8970B Noise
Figure meter, the HP437B Power meter, the HP 8592A Sp ectrum Analyzer, the Colorpro
HP Plotter, printer, the HP 6237B p ower supplies, and the HP 3457A multimeter.
A bench rack is available for storing your circuits under development. Please use the white
stickers available in CL 305 to write your name(s) on your circuits.
The circuits designed by the students are fabricated in CL305 using computer aided ma-
chining by the EE 723 teaching assistant. CL305 features a workb ench, a to ol b ox, various
drills, two di erent soldering irons, and a parts cabinet. 5
1.3 Lab oratory Policies
For a prop er and ecient sharing of the lab oratory equipment a set of rules is necessary. The
latter make up the Lab oratory p olicy. A sample of the Lab oratory Policy is given below.
You are asked to read it and obide by its rules to participate in this lab oratory.
EE 723 Lab oratory Policy
1. Schedule your measurement and fabrication time (CL 305) in advance on the sign up
sheet on the do or of CL 305. You can reserve up to 2 hours/p er week. Op en time is
however available on a rst come, rst served basis.
2. Read your lab oratory manual and make a plan of work b efore coming to the lab.
3. You can obtain the lab oratory key against your I.D card at the reception desk of the
EE department oce. The receptionist should check that you are on the EE 723 class
roster; remind him to do so!.
4. Make sure that the laboratory door is closed when you leave the room even for a short
period.
5. Care for the measurement cables and connectors (refer to App endix 5.5 of your lab o-
ratory manual) and the equipment in general. The measurements will b e made using
SMA test cables. Do not disconnect the SMA test cables from the Network Analyzer.
Dirty cables lead to bad data.
6. Read soldering, drilling and cutting tips in the App endices D, E and F of your lab o-
ratory manual to ensure a prop er use of the to ols and to assure your pro ctection. If
you are uncomfortable with a particular lab oratory pro cedure request help from the
instructor.
7. Rep ort damage or malfunction immediately (you will not b e p enalized) to the instruc-
tor or teaching assistant.
1.4 Lab oratory Schedule and Rep ort
A lab oratory rep ort is due for each lab oratory and should be turned in by the due date
sp eci ed in your EE723 syllabus In addition a poster page summarizing your design should
be included with the last lab oratory rep ort. A few guidelines are given below: 6
use for rst page of your lab rep ort the lab oratory rep ort sheet given in App endix 5.10,
include a table of contents describing the material app ended,
write a caption on each gure and plot,
discuss any problems encountered.
Only one rep ort p er team should b e turned in. The rep orts do not need to b e long but
must be well organized. The rst two lab oratory rep orts do not need to be typ ed. However
to help with clarity please write the lab oratory 3 (design pro ject) rep ort and p oster page
on a word pro cessor (latex is strongly recommended). Note that you can store your ADS
simulation plots in a p ostscript le and include them in your rep ort without cutting and
pasting. Your rep ort will b e archived so that it can be used as a reference by future EE723
students and your p oster will displayed in the display case in the Dreese-Caldwell bridge. 7
2. LABORATORY 1:
Microwave Measurement of a Microwave Ampli er
Lab oratory Goal: This lab oratory is divided into two parts. The purp ose of PartIofthis
lab oratory is to get familiarized with the Network Analyzer and to p erform measurements
on a microwave ampli er. In Part I I of this lab oratory, the measurement of the Noise Figure
of the microwave ampli er is p erformed using a Noise Figure meter. Approximate Duration:
4-5 hours.
2.1 Preparation for Lab 1
1. Preliminary Reading :
Read in the User's guide of the HP 8753B Network Analyzer the following chapters:
- Chapter 1: Op erating the HP 8753B
- Chapter 3: Scattering Parameter Measurements with the HP 8753B
- Chapter 4: Time Domain Analysis
Do not forget to bring your Network Analyzer User's guide with you to the
lab for quick reference.
2. Lab oratory Overview :
Read the remaining of the instructions for Lab 1 b efore coming to your scheduled
lab oratory time. Determine the goals to be attained.
3. Connector and Cable Care :
The measurement of microwave circuit is a science. In this lab oratory and the subse-
quent ones you will physically connecting many microwave devices to the test cables of
the network analyzer. The qualityofyour measurement and the lifetime of the devices
and cables will dep endent critically on this simple pro cess. Please read App endix 5.5
which gives some key advices ab out it. It is indeed imp ortant to learn how to do it
right and develop habits which you can keep during your professional life.
2.2 PART I: Frequency and Gain Compression Measurement
Measurement Pro cedure: 8
1. Clean the SMA connectors of the cables, and standards :
The relatively inexp ensive SMA connectors are not made to be connected more than
a few times. You might notice gold particles on the white dielectric of the cables,
and standards which wil l a ect the quality of your calibration and measurement. The
standards which are sketched below are available in a small wooden box. Clean the
cables and standards (50 and short) using a cotton swab, alcohol and compressed
air. See connector care in App endix 5.5.
2. Power Up :
Turn the Network Analyzer on if it is o or press PRESET if it is already on.
3. Perform a2-Port Calibration of the Network Analyzer
You might nd it convenient to select the SMITH CHART format (p. 10) for all the
scattering parameters b efore calibrating. Simply press MEAS to select a S-parameters
and then press FORMAT and select SMITH CHART.
Do NOT HOLD the cable when you calibrate as this intro duces calibration
instabilities. Just let the coaxial cable lying at rest on the table b efore you
select SHORT, OPEN, 50 LOAD and the four THRU in the calibration
pro cedure.
To start the calibration press CAL and select the 2-p ort Calibration (p. 13). Use the
calibration short, op en, 50 load in the SMA calibration kit (available in the small
wooden box) for the re ection calibration of p orts 1 and 2. A sketch of the standards
is given b elow. Use the thru for the THRU calibration. For the isolation calibration
select Omit Isolation.
Press DONE when done. Save your calibration in a register (p. 31).
Verify the calibration by testing the 50 load, the op en and short and thru connections.
Note that this calibration is only valid for the particular test cables you are using.
Please do not remove the test coaxial cables!. You are now ready to p erform a two-p ort
measurement.
50 Ω OPEN
SHORT THRU
Figure 2.1: Calibration Standards
Before you continue a few comments on calibration are in order. 9
The transmission calibration required the use of a thru standard. A thru is supp osed
to havenolength. Wehave a mating problem since our SMA connectors are b oth male
can b e connected together. Since an error in the thru calibration is not to o critical at
3GHzyou used the short female to female barrel available. A more rigorous approach
at high-frequency requires the use of two equal length barrels. The female to female
barrel is connected b etween p ort 1 and p ort 2 during the thru calibration. The female
to male which provides an equal electric length is connected say to port 1 for the S11
calibration and for the subsequent S-parameter measurements. [Note: Once you have
b ecome familiar with the calibration pro cedure you can avoid the calibration step by
recalling the 2-Port calibration data from a previous calibration whichhave b een stored
on disk for your convenience (check with the instructor to verify if this is available).]
To retrieve calibration data use the following input pro cedure:
Press [LOCAL], select [SYSTEM CONTROLLER].
Press [RECALL], select [LOAD FROM DISK], [LOAD TWOFULL].
Press [LOCAL], select [TALKER/LISTNER].
The last step is intended to free the HPIB (IEEE488) network so that another piece of
instrument (the other Network Analyzer or the Noise Figure meter) can also b ecome
temp orarely a system controller to plot or access the disk drive. Once these calibration
data have been loaded the network analyzer should be calibrated. You can verify it
with the op en short, thru and 50 to make sure. Note: The calibration data are only
valid for the pair of coaxial cables used. The network analyzer will be improp erly
calibrated if someone has changed the coaxial cables. Do not disconnected the coaxial
cables from the network analyzer itself !
4. S-parameters of a microwave ampli er :
We shall now measure the S parameters of a low noise narrow-band microwave ampli-
er.
Clean the connector of the ampli er (see App endix 5.5 on connector care).
Connect the low noise ampli er mo del 13LNA (aluminium body with 2 SMA
connectors, blue lab el) b etween p ort 1 (input) and 2 (output).
Turn on the HP6237B p ower supply. The op erating voltage should already b e set
to 12 V on the lower scale of the power supply meter (Do not increase it as this
would damage the ampli er). Note: The red banana cable should be connected
to +18 V and the black cable to COM. The meter switch should indicate +18 V. 10
Connect the p ower supply to the ampli er. Resp ect the p olarity! Red is p ositive.
Black is negative. To prevent power gain compression of the front end 13LNA
ampli er we need to reduce the power used by the Network Analyzer for the S-
parameter measurements to -10 dBm:
[MENU]
[POWER] [-10] [1]
Find the frequency f at which S is maximum using the marker MKR. Find the 3
0 21
dB bandwidth B of the ampli er. Make a hardcopy output on the plotter (Follow
the instruction given in the User's guide manual p. 17). Record jS j and jS j in
21 12
dB at this frequency and measure jS j and jS j using VSWR units. [Note : The
11 12
network analyser should only b ecome the system controller for the time necessary
to plot (Press SYSTEM and select CONTROLLER). Once you are done free the
HPIB (IEEE488) network (Press SYSTEM and select TALKER/LISTEN) to give
access to the plotter and disk drive to other users.]
5. 1dB Power Compression :
We shall now measure the 1 dB Power compression point at the frequency f of the
0
ampli er under test.
The network analyzer HP 8753B o ers the capability (not usually asso ciated with
a network analyzer) to sweep the RF input power. An accurate measurement of the
output p ower of an ampli er requires a sp ecial calibration pro cedure to account for the
loss through the test set between the ampli er output and the receiver input. For the
sake of simplicitywe shall bypass here this sp ecial calibration. Our measurement should
give us the ampli er's resp onse to a power ramp. Record jS j in dB of the ampli er
21
for -10 dBm input p ower P . [Note : P in dBm is de ned as 10 log(P =1 mW ). From
in in in
this resp onse we shall determine the input p ower at which the gain jS j is compressed
21
by 1 dB.
Measurement pro cedure:
Select a power sweep at the frequency f (e.g. 2GHz)
o
[MENU]
[SWEEP TYPE MENU]
[POWER SWEEP]
[RETURN]
[CW FREQ] [2] [G/n]
Set the stimulus parameters. Power levels must be set so that the ampli er is
forced into power gain compression. The range of the HP 8753's source is from 11
-10 to +10 dBm.
[MEAS]
[S21[B/R]]
[START] [-10] [1]
[STOP] [10] [1]
Use the marker [MKR] to nd the input power for which a 1 dB drop in the
ampli er's gain o ccurs relativetothe small signal gain.
6. Fill the Lab Rep ort # 1 given in App endix 5.10. App end the plots which do cuments
steps 4, and 5.
2.3 PART II: Measurement of the Noise Figure
Low noise RF and microwave ampli ers are used at the frontend of microwavecommunica-
tion systems to amplify the weak signals recieved bytheantenna or dish. The signal recieved
15
from a sattelite are typically in the order of -120 dBm (10 Watt) which corresp onds ap-
proximately to a voltage of 0.2 V across a resistor of 50 .
The Noise Figure F is de ned (see Gonzalez p. 296) as the ratio of the output noise power
P of the ampli er divided by the ampli er gain G and the input noise p ower P .
No A ni
P
No
F =
P G
Ni A
The noise gure F of an ampli er is larger than 1 b ecause it is a measure of the noise added
by the ampli er to the output noise p ower. If the ampli er did not add noise wewould have
F = 1 or 0 dB. The bandwidth B of the ampli er and its noise gure F will p ermit us to
calculate the minimum detectable signal input p ower P (see Gonzalez p. 354).
i;mds
The measurementof the noise gure is quite simple with the HP8970B meter.
1. Power up :
Turn the HP8970B Noise Figure meter by pressing the LINE switch ON. If the noise
gure meter is already on press RESET. The noise Figure Meter p erforms a quick
internal check and frequency calibration.
: 2. Set frequency parameters
The HP8970B measures the noise gure versus frequency from 10 to 1800 MHz. The
default START, STOP and START frequencies are 10, 1600 and 20 MHz resp ectively.
: The noise To extend it to 1800 MHz simply press STOP FREQ 1800 ENTER. [Note 12
gure of the low noise ampli er 13LNA at its center frequency f cannot b e measured
0
as it is ab ove 1800 MHz.]
3. Calibration :
Connect (using a thru) the Noise Figure Meter as shown below
HP 8970B NOISE FIGURE METER
SOURCE INPUT
NOISE SOURCE
Figure 2.2: Noise Figure Meter Calibration
Press CALIBRATE twice. The HP8970B measures its own noise gure. The noise
gure meter should make two frequency sweeps. If an error o ccurs you might have
to increase the number of measurements averaged by pressing INCREASE in the
SMOOTHING blo ck and try again. If after several trials an error is still o ccuring
you mighthave to limit the measurement to 1600 MHz.
4. Measurement of the Noise Figure of the Ampli er :
Once the noise gure meter has b een calibrated it is p ossible to measure the corrected
: This noise gure is noise gure and gain of the device under test (DUT). [Note
corrected b ecause the meter removes its own noise gure from the measurement results.]
Connect the ampli er between the noise source output and the meter input as
shown b elow:
To measure the noise gure and gain press the key NOISE FIGURE AND GAIN.
The INSERTION GAIN displayshows the device gain and the NOISFE FIGURE
display shows the noise gure of the DUT at ro om temp erature at the displayed
frequency.
Connect the HP6237 power supply to the ampli er. Resp ect the p olarity! See
earlier instructions regarding the use of the power supply. Notice the reduction
of the noise gure once the ampli er is powered.
Measure the noise gure and gain at 1800 MHz by entering the frequency. 13 HP 8970B NOISE FIGURE METER
SOURCE INPUT
NOISE SOURCE AMPLIFIER
Figure 2.3: Noise Figure Measurement
Measure the noise gure versus frequency by pressing the SINGLE key to obtain
a single sweep.
FOR UNKNOWN REASONS THE PLOT INSTRUCTIONS BELOW
ARE NO LONGER WORKING. Please simply record the data man-
ually and plot the results in MATLAB!
Plot the data measured. In order to plot, the noise gure meter must be the
controller of the HPIB (IEEE488) network.
Verify that nob o dy is plotting or using the disk and inform the other users that the
noise gure meter will temp orarely b ecome the controller of the HPIB network.
Turn the plotter on and intro duce a sheet of pap er.
Press 48.0 SPECIAL function for the noise gure meter to b ecome the controller.
Press 47.0 SPECIAL FUNCTION to verify that the plotter is on the HPIB net-
work.
Press 25.0 SPECIAL FUNCTION to start the plot.
Once the plot is done press 48.1 SPECIAL FUNCTION to free the HPIB network
from the control of the noise gure meter.
5. Complete the Lab oratory Rep ort #1 with the noise gure data. App end you noise
gure plot. Calculate the p ower of the minimum dectectable input signal P of this
i;mds
amplifer at f (you will need to extrap olate F at f ).
0 0 14 15
3. LABORATORY 2:
Measurement of the Scattering Parameters of a Microwave Transistor
Lab oratory Goal: The goal of this lab oratory is to measure the scattering parameters of
the NPN transistor AT42085 you will use for designing your ampli er. The NPN transistor
AT42085 will b e biased with 10 mA collector current and 8 V collector-emitter p otential. A
microwave test b ed sp ecially develop ed for microstrip transistor will b e used for this purp ose.
You will compare the scattering parameters obtained for your NPN transistor AT42085 will
the scattering parameters quoted by Advantek (see App endix 5.6). You will use these data
for the design of your ampli er.
Measurements:
You must have requested and obtained from your instructor your own AT42085 transistor
b efore starting this lab oratory.
1. Power Up :
Turn the Network Analyzer HP8753C on if it is o or press PRESET if it is already
on.
: 2. Calibration
Due to the cost and fragility of the microwave transistor test-b ed and the op en, short,
thru and 50 ohm standards, we will not p erform the calibration of the Network An-
alyzer with the test bed in this lab oratory. Instead you will retrieve the TWOPORT
calibration data which have been stored in the adjacent Personal Computer. SPE-
CIAL INSTRUCTIONS WILL BE PROVIDED BY THE TA.
Note that the calibration of the Network Analyzer with the test-b ed has set the refer-
ences planes at the level of the microwave transistor itself.
3. Intro duction of your microwave transistor in the test bed :
Gently lift the white lid of the test bed by pressing on its silver button. PLEASE be
careful. Use the tweezers to position your transistor in the test bed. The base of the
transistor is indicated by either a green dot or triangle shap e on the transistor body
or is di erentiated by the triangular shap e of its lead. The collector is on the opp osite
side of the base and the two emitters on b oth sides. Port 1 of the Network Analyzer
should be connected to the input of the transistor (base of the NPN). Port 2 of the
network Analyzer should b e connected to the output of the transistor (collector of the
NPN).
4. Turn on the bias network: 16
The bias network should have the BNC connector lab eled "base" connected to
p ort 1and the BNC connector lab eled "collector" connected to port 2.
Turn on the HP6237 power supply. The op erating voltage should already be set
to 10.1 V on the lower scale of the power supply meter (Do not increase it as
this could damage the biasing network). [Note : The read banana cable should
be connected to +18 V and the black cable to COM. The meter switch should
indicate +18 V.]
Connect the NPN biasing network to the HP6237 power supply. Resp ect the
p olarity (red with red, black with black)! The red part of the banana plug should
be connected to +18 V and the black part to COM.
5. Measurement: Verify that the S scattering is in the order of 7 dB for a wide range
21
of frequencies. This indicates that the transistor provides some gain and is biased
prop erly. Reduce the measurement power level to -10 dBm by pressing MENU, and
then POWER 10
Data Acquisition: Plot the scattering parameters from 300 kHz to 3 GHz on the Smith
Chart. Record the scattering parameters for the frequencies sp eci ed in App endix 5.6.
THIS TASK IS NOW PERFORMED USING COMPUTER DATA ACQUISITION.
Use the sp ecial instruction given in App endix 5.7
6. Using ADS compare on a single Smith Chart the scattering parameters measured with
those given in App endix 5.6. To create the S parameter le for your transistor see
App endix 5.8. 17
4. LABORATORY 3:
Design of a Microwave Ampli er
Lab oratory Goal: The goal of this lab oratory is to design, fabricate and test a microwave
ampli er. Three p ossible designs are prop osed:
- Narrow band/maximum gain (2 GHz center frequency)
-HighQ Narrow band/maximum gain (2 GHz center frequency)
- Narrow band/low noise (1.5 GHz center frequency)
Approximate Duration: 6 hours
4.1 Design
4.1.1 Imp ortant Design Constrains:
1. The ampli er must b e stable for50 source and load imp edances at all frequencies.
2. The ampli er will b e designed using the NPN transistor AT42085 biased with 10 mA
collector current and 8 V collector emitter p otential. Typical values for the scattering
and noise parameters of the AT42085 transistor at this bias p oint are given in App endix
5.6. It is exp ected that the scattering parameters measured in Lab #2 should b e more
accurate than those of App endix 5.6 for your ampli er design.
3. The input and output matching network will be designed for a 50 source and load
imp edance. The matching circuit will be realized using the technique of your choice.
Note however that short circuit stubs and lump ed elements cannot b e used.
4. The microstrip will b e constructed on a Duroid substrate. Get the sp ec of this substrate
from your instructor. The pattern will b e fabricated by the Quick Circuit machine.
5. The biasing of the bip olar transistor will be realized through the built-in
bias Tees of p ort 1 and port 2 of the Network Analyzer (see App endix 5.9)
using the NPN bias network provided in the lab oratory. This active biasing
circuit is analyzed in App endix 5.9.
4.1.2 Design Pro cedure:
The design pro cedure of the ampli er is based on the following steps: 18
1. See the instructor to obtain the values you should use for the dielectric
thickness, dielectric constant and tangent loss and copp er thickness of
r
the Duroid microstrip substrate you will use for your ampli er. For copp er we
have RHO = .84. Using lineCalc calculate the width and length of a 50 line of
electrical length of 360 degree (length equal to one wavelength). .
2. To start your design you need realistic scattering parameters for your transistor which
accounts for the way the transistor is connected on the microstrip b oard. You will use
the transistor S-parameter le you acquired in Lab oratory 2. However the via holes
used to connect the emitter of the transistor to the ground intro duces a
parasitic inductance. It results that the scattering parameters of the transistor plus
connecting bed seen by the circuit departs from those you measured for the transis-
tor. This is very imp ortant as this a ects the ampli er design, stability and noise
performance. Via Hole elements must therefore be introduced between the transistor
emitters and the ground to account for this parasitic inductance in your simulation.
Also the di erence between the width of the transistor pads and the width of your
transmission lines must b e bridged by a tap ered line. TO FACILITATE your design a
transistor fo otprint sub circuit accounting for these parasitics is provided in App endix
5.2.
3. Using ADS calculate the new scattering parameters for the transistor plus its connect-
ing b ed. Make sure that the correct microstrip substrate has been used. Make
sure that the tap ered line connects to the correct width of the line connected to it (for
example your design mightusealineof50 ). Makesuretouseyour own S-parameters
measured in Lab oratory 2. See App endix 5.7 to learn how to save the new transistor
S-parameter data in a le and App endix 5.8 to verify the format used by ADS.
4. Do a rst cut design (see Homework #2) using the Smith Chart and the new scattering
parameters resulting from this circuit.
5. Implement your rst cut design in ADS using transmission lines. Do not optimize
yet this ideal circuit as you still need to design a realistic microstrip layout.
Compare your rst cut design with your Smith Chart design to verify that the rst
cut design is correct. This rst cut design provides you with a starting point for the
design of the ampli er with microstrip lines.
6. Using Linecalc (see App endix 5.3) calculate the width and length of the microstrip line
required for your design (use MLIN in linecalc). Actually or a 50 line you can just
scale the wavelength obtained in the rst step for E=360 by the appropriate electrical
length E you need to implement. 19
7. Implementyour ampli er design in ADS using microstrip lines. You can compare the
microstrip and transmission line implementation but do not optimize yet as this is not
the nal circuit yet. Indeed the nal circuit you are going to optimize should realisti-
cal ly represent the circuit you are going to fabricate. Intro duce MCROSS, MTEE,
MSTEP, MOPEN and VIA2 to account for all the discontinuities present in your
layout (see App endix 5.2). Use the transistor TEST BED: trans_sim.dsn intro-
duced in App endix 5.2 to simulate or optimize your design. Up date FILE=at42085N
in trans_sim.dsn with the S-parameters measured for your own transistor. Once
your circuit has been optimized replace the TEST BED le tran_sim.dsn by
trans_lay.dsn to be able to generate the correct transistor layout. Then you can
generate the layout of your complete circuit using the synchronize command describ e
in App endix 5.2toverify that the layout is acceptable (not to o long) and that all the
discontinuities are prop erly mo deled in your circuit. ALSO your layout should not
violate the design rules given in the next section. You can use MBEND to reduce
the length when appropriates. If you are adding biasing lines (see low noise ampli er
design comment below) includes them in your design to verify if they impact the RF
p erformance and makesure they are prop erly implemented in your layout!
8. The fo otprint for the input and output SMA connectors will b e added to your layout
bytheTA (unless otherwise instructed). JUST makesuretoadda50 ununcomb ered
microstrip line of a quarter of an inch (1/4 in) at b oth the input and output of the
ampli er to p ermit the installation of the connectors without interfering with your
circuits. The connector fo otprint should not a ect your ampli er design apart from
reducing its gain and increase its noise gure due to the SMA connector to microstrip
launcher loss.
9. Once a realistic circuit mo deling of the nal ampli er circuit to be layout has been
implemented then it is worthwhile to optimize your design. Plot jS j (dB) and jS j
21 11
and jS j versus frequency.
22
10. Investigate the stability of the ampli er from .1 to 6 GHz. For this purp ose
generate the stability circles of the ampli er using ADS. This p ermits to verify whether
or not the nal ampli er is stable for the 50 load used at the input and output.
Unconditional stability is usually to o much of a constrain to target. It is generally
preferable to avoid using resistive loading since it reduces the gain unless if you are
purp osely designing a broad band ampli er. If resistive loading is to b e used note that
only series loading can be used since a shunt loading requires a capacitor to prevent
shorting the bias (see the instructor if you absolutly need ashunt loading). A resistor
in series with a high impedance (very narrow) microstrip line (inductance) and DC 20
block capacitor wil l realize a resistive loading with would stabilize the transistor at low
frequency but wil l not decrease its gain at high-frequency.
11. If you are building a low noise ampli er you need to have it t inside a box to shield
it from the environment noise and approach the simulated noise gure. In such a case
the entire circuit should t inside a sp eci c rectangle size with an appropriate safety
margin, say of 1/4 of a inch, from the box's edge. The dimension of the box available
for your design is shown in Figure 4.1.
Note that the box will also impact the synthesis of the microstrip line width and
you you need to set the correct value for HC, W1 and W2 for the distance of the
microstrip line to the cover and wall of the box (see App endix 5.3). Also to measure
its noise characteristic, a an external biasing TEE will need to b e implemented on the
microstrip b oard to bias the transistor (usually the network analyzer is used to p erform
this function). The best approach is to include the bias TEE as part of the ampli er
layout. You can discuss these various options with the instructor.
4.1.3 Summary of the Design rules asso ciated with the Fabrication
Your circuit will b e fabricated using the Quick Circuit machine. Your design should therefore
account for the physical constrains intro duced by the layout and the fabrication.
Read App endix 5.2 for more information on the layout pro cedure. In particular the layout
to ol intro duces the following constrains
Via holes (connection to the ground plane through the microstrip b oard) should be
mo deled using the VIA2 mo del.
All the line discontinuities should b e mo deled using elements such as MSTEP, MCROSS,
MTEE, MBEND, MCLIN, MGAP and so on.
Use an MGAP statement for the fo ot print of chip resistors and chip capacitors.
A few design rules should b e resp ected to p ermit the succesful fabrication of your design:
The minimum line width is 10 mils
The mininum separation b etween two lines (line gap) is 11 mils.
Connectors (SMA launchers) should not b e closer than 2 cm for easing their connection
to the SMA cables. 21 2.375" 2.5" 2.795" 1.5" BOX
Top View Side View
3"
Design Limits 2" 2.375" 2.5 " 2.795" 2.247" Microstrip Board
2.35" 0.127"
2.627" 0.1165" 0.1885" 0.1885" Inside contour of box
Design Limits
Possible layout 2" 1: The microstrip lines should not extend beyond the design limits. Design Limits 2: The center pin of the connectors can be placed anywhere inside the design limits including the edge.
2.35"
Figure 4.1: Box Dimension 22
A 50 line of 1/4 of an inchshould be reserved to attach the SMA Connector (SMA
launcher).
MINIMIZE the area of your circuit. The Quick Circuit fabrication time and therefore
to ol-wear is prop ortional to the amount of copp er to remove! Please minimize the
fo ot-printofyour circuit. For the low-noise ampli er the b ox constrains the size of the
layout.
Please set up an app ointment with the TA b efore the Layout Approval by TA date
sp eci ed in your syllabus. to have your layout checked and approved for fabrication by the
TA.
Exp ect up to a week turn around for fabrication.
The circuits are all fabricated together in a single batch. Your design should
therefore be delivered to the Teaching Assistant by the Final Layout De-
livery date sp eci ed in your syllabus.
4.2 Fabrication Pro cedure
The fabrication of your circuit with Quick Circuit involves multiple steps which are p erformed
by the Teaching Assistant. First your layout is nalized with the addition of the connector
fo otprints and the transistor fo otprint determined and the circuit b ound is determined. The
le is then converted in Gerb er format. The fabrication pro cess is manually designed from
your layout using a to ol call Isolator. Finally the fabrication pro cess is executed.
Your circuit built with the Quick-Circuit machine should already include all the transmission
lines and holes required and no additional drilling or cutting should be required. Therefore
the remaining fabrication steps are relatively few.
List of parts:
Microwave transistor
Microstrip b oard
2SMA connectors
PNP biasing network (provided)
bias TEE are required for oscillators and LNA
Fabrication Pro cedure:
1. Solder the transistor using the soldering iron (see App endix 5.4). Use a heat sink to 23
avoid damaging the transistor. The transistor leads need to be ush with the
microstrip board.
2. Fix the connectors on the circuit b oard. One of the four legs must be removed if this
has not b een done. The connector NEEDs to be ush with the ground plane.
Solder the connectors using the soldering gun (see App endix 5.4). Soldering to legs on
the ground plane should be sucient.
4.3 Measurement of Your Circuit.
Measurement pro cedure:
1. Make a 2 port calibration with the Network Analyzer or recall one saved on the HP
disk drive for the same test cables.
2. Ampli er Biasing:
Connect Port 1 to the input of the ampli er (base of the NPN). Connect Port 2 to
the output of the ampli er (collector of the NPN). The bias network should have
the BNC connector lab eled "base" connected to p ort 1 and the BNC connector
lab eled "collector" connected to p ort 2.
Turn on the HP6237 power supply. The op erating voltage should already be set
to 10.1 V on the lower scale of the power supply meter (Do not increase it as
this could damage the biasing network). [Note : The red banana cable should
be connected to +18 V and the black cable to COM. The meter switch should
indicate +18 V.]
Connect the NPN biasing network to the HP 6237A power supply. Resp ect the
p olarity (red with red, black with black)! The read part of the banana plug should
be connected to +18 V and the black part of the banana plug to COM.
3. Measure and plot jS j in dB versus frequency. Atwhich frequency f is the maximum
21 0
gain? What is the 3dB bandwidth B? Measure and plot jS j in dB.
11
4. Measure and plot jS j and jS j in VSWR units. Checkhowwell the input and output
11 22
networks are matched at the op erating frequency.
5. Trim the op en stub to improvethe matching if necessary. You could also increase the
stubs' length by soldering a piece of copp er tap e. 24
6. If you exp erience diculty (apparent noise) with the measurement of the scattering
parameters with the Network Analyzer the ampli er might be unstable. You can
verify it with the sp ectrum analyzer. The ampli er can be stabilized with a feedback
resistance between collector and emitter or a shunt resistance (in series with a ship
capacitor) b etween base and emitter or collector and emitter.
7. Measure the input p ower for which the power gain drops of1dB.
8. For low noise ampli ers: Measure the noise gure at the design frequency .
4.4 Write Up
4.4.1 The Lab oratory Rep ort
Your rep ort should include:
0. The complete lab oratory 3 rep ort sheet available in App endix 5.10
1. 1. Smith Chart Design
2. ADS circuit le and plot of S11, S22 and S21 (dB).
3. The drawing of the ampli er layout including all dimensions.
4. Plots of measured jS j; jS j; jS j; and jS j (dB or VSWR) versus frequencies.
11 22 12 21
5. Compare the measured and simulated data.
6. Discuss the design and fabrication problems encountered.
4.4.2 Poster Page
A poster page summarizing your design and exp erimental results should be included with
the rep ort. This p oster page should include:
Pro ject Title
Your names
EE723, Fall Quarter 1995 25
A pro ject abstract describing the circuit targeted, the design approach and the p erfor-
mance obtained.
A gure or two with short captions comparing the measured and simulated (ADS)
circuit p erformance.
A gure showing the layout generated by ADS.
A prototyp e Poster page in latex will b e made available in the directory:
~roblin/latex/ . 26
5. APPENDICES
5.1 How to Use ADS
The most current information for using ADS is available at:
http://eewww.eng.ohio-stat e.e du/a ds
Usually to start ADS the following commands are used:
source /apps/ads2002/Startup.EE
hpads
TUTORIAL:
If you are new to ADS you should go through the OSU tutorial:
http://eewww.eng.ohio-stat e.e du/a ds/t utor ial 2.pd f
Check also our lo cal ADS webpage: http://eewww.eng.ohio-sta te.e du/ ads for access
to the online ADS manual and other various tutorials found on the web.
MOVING OR SHARING DESIGNS:
Note that if you are moving your design les to another directory (or want to share them
with someone else) you can copy b oth the .dsn and .ael les of the networks directory of
your pro ject directory. The .atf les are generated automatically.
IMPORTANT NOTE:
It is imp ortant not to exist ADS in an unconventional way. When you do ADS keeps running
in the background even if you log o . We have only ten user licenses and we quickly run
out of them if such pro cesses are running in the background. The consequence is that other
p eople (including yourself ) stop b eing able to use ADS. Please exit ADS using the exit
command in the Pro ject menu.
If ever ADS dies or is exited unconventionaly you need to kill all the unix pro cess which are
still running. For this purp ose typ e:
ps
or typ e:
ps -ef j grep yourusername
Then kill the job number involving your name and hp eesof. For example by typing ps -ef j
grep roblin the user roblin obtained among otherthing:
roblin 14776 14775 0 Jan 5 pts/0 0:00 hpeesofviewer
To kill that job simply typ e:
kill -9 14776 27
5.2 How to Use the Layout Tool
ADS allows for the automatic generation of the layout from the circuit design.
To generate the layout you must have b oth the schematic and layout windows op en. In the
schematic window select the synchronize menu and the rst item on this menu. ADS will
highlight in red every circuit for which it has a corresp onding layout or fo ot print. It do es
not have a fo ot print for capacitors. ADS starts by Port 1 and asks you where you want to
lo cate it. Position (0,0) is just ne. The layout will stop whenever, starting from p ort 1,
ADS runs intoadevice for which it do es not have alayout. If you run again synchronize it
will then ask for a new p osition (x,y) where to initiate the continuation of the layout. The
b est is to instead intro duce elements which sp ecify a fo ot print. For example for a capacitor
you would use a chip capacitors. Simply place in parallel with the ideal capacitor a MGAP
(microstrip gap) element and sp ecify the width of the gap (width of the capacitor).
Note that the synchronize to ol requires that you make use of the discontinuities MSTEP,
MCROSS, MGAP, VIA2 and so on to establish a physical network. This is also required to
obtained a realistic simulation of your circuits as these discontinuities intro duces parasitics
which impact the microwave p erformance of your circuit.
As you go through several layout trials you will want to RESET the layout window. For this
purp ose go to the le menu and select: clear layout and then synchronize again to obtain
a new layout. Be careful if you synchronize in the layout window your schematic
will b e up dated! Save your schematic ahead of time to avoid an unwantedmodi cation of
your schematic.
Sometimes the layout is not what you want. For example one can exp erience two lines
intersecting. In the layout window you can then click on all the items you want to move
together and then move them and rotate them where you want. All the editing commands
apply to these items.
The layout is really o ccuring on several layers. The metal (cond) layer is the most imp ortant
one for our pro cess. However you need also to de ne the b ound (size) of the circuit. In the
Layout window go to the Draw menu and select the Select layer menu time. Select (you
might have to scroll down the menu) the b ound layer (not b ond!). Then again select the
Draw menu and select rectangle. Place your rectangle with the mouse. The Quick Circuit
machine will cut the contour of your circuit using the b ound you sp eci ed. If your circuit is
to b e placed in the box available refer to the section below discussing
Holes are also considered as a layer. There are automatically intro duced by the via holes:
VIA2 (see b elow). 28
Ask you instructor for more information on the substrate you will be using. For the nal
Quick Circuit design we usually use a thin substrate of 30 or 45 mils to obtain smaller
microstrip linewidth. Changing the substrate MSUB a ects the width of all the lines. Such
a change can be easily implemented in your schematic if you de ne a variable say W 1 for
microstrip lines of same width. Then you can select the value typ e of the width to be a
variable (instead of a parameter) and select this variable to be W 1 in the value option
menu.
Via Holes:
Via holes are used to establish a connection to ground. They also a ect the simulation
since a via hole b ehaves as an inductor. You need therefore to include them b oth for the
simulation and the layout. Use the mo del VIA2 available in the microstrip library. W is
the width of the square pad used on the top of the microstrip. The hole should be smaller
than the width of the pad.
So in summary to p erform a layout you need to up dage your schematic with MCROSS,
MSTEP, MGAP, VIA2, bias line (for active circuits) and so on to obtain a realistic design.
When you are done with your layout schedule an app ointment with your instructor to discuss
your layout/schematics and simulation b efore the scheduled fabrication (see EE723 syllabus). 29
5.3 How to Use LineCalc
How to Start and Use LineCalc:
1. You can start LineCalc from the ADS Circuit window from the TOOLS menu.
2. We will most likely use LineCalc to synthesize a microstrip line. Click on Select... and
scroll down the menu to select MLIN (and not MCLIN) and clickon OK.
3. Edit the various substrate parameters using the Mo dify Substrate key (you may need
to scroll the menu or make the window bigger): E is the e ective relative dielectric
r
constant, Mur is the p ermittivity (1.0), H is the thickness of the substrate, Hu is the
p osition of the cover (keep it large if there is none), T the thickness of the copp er line
7 1 1
C ond is the copp er line conductivity: (4:878 10 ! m , T anD is the loss tangent,
R oug h is the ideal surface roughness (Rough=0 is avery go o d approximation).
4. Edit the comp onent parameters. Select the frequency Freq targeted. Wall1 and Wall2
are the distance of the microstrip line from the side metallic walls. Set Wall1 = 0 and
Wall2 = 0 to make them in nite (no walls) except mayb e if you are making a low
noise ampli er going intoabox.
5. Set the characteristic imp edance ZO and e ective electrical length E_Eff you wish to
obtain.
6. Click the up arrow to calculate the width W and length L of the microstrip line. Also
calculated are K_Eff the e ective relative dielectric constant, the line attenuation A_DB
and the skin depth.
7. For more information use the on-line Help command. 30
5.4 Tips for Soldering
Tips for soldering the transistor:
1. Moisten the sp onge of the soldering stand.
2. Turn on the soldering iron.
3. Set the temp erature around 700oF.
4. The right light is blinking when the tip is hot.
5. Before soldering clean the soldering tip with the moisten sp onge.
6. Fast soldering will prevent overheating of b oth the circuit and elements.
7. Avoid breathing the fumes.
8. Do not use to o much solder.
9. When you solder the chip capacitor or resistor, one student can hold the elementwith
tweasers and the other solder it.
10. When you solder the center conductor of the connector, you should solder b oth sides
of the conductor.
11. Be careful when soldering the transistor on the copp er tap e. To prevent damaging
your transistor use a heat sink such as a metal clip or pliers.
Tips for soldering the connectors
To solder the SMA connectors to the microstrip b oard use the soldering gun. BE CARE-
FULL. You must plug it and press on the trigger to active it for the time you
are soldering ONLY.
The SAFETY PRECAUTIONS are:
1. Keep you soldering gun well away from all ammable material.
2. Toavoid burns, always assume that the tip is hot.
3. Be sure the hot metal tip do es not come in contact with the electrical p ower cord.
4. Before making any adjustment-removing or replacing a tip etc- make sure the gun is
unplugged and co ol. 31
5. Release the trigger whenever the tip is not in contact with work. NEVER EVER tap e
back the trigger.
6. Do not hold work in your hand if you can p ossibly avoid it. Use a vise, clamp or pliers.
7. Do not dip the to ol into any liquid.
8. Many materials give o unpleasant fumes when heated-so always work in a well venti-
lated ro om.
9. Clean the tip by wiping it, when hot across a damp sp onge or cloth- placed on a
non- ammable surface, NOT held in the hand.
10. AFTER USE, DISCONNECT the soldering gun. allow the tip to co ol
completely, and store the to ol in a safe place (out of reach of children).
11. Safety goggles are recommanded to prevent hot materials from entering the eyes.
Do not use much solder to hold the connectors. We intend to reuse these connectors. 32
5.5 Caring Ab out Connectors
The relatively inexp ensive SMA connectors we use are not made to b e connected more than
a few times in their lifetime. You might notice gold particles on the white dielectric of the
cables and standards which will a ect the quality of you calibration and measurement. In
such a case clean the cables and standards using a cotton swab and alco ol. Here are some
general care and maintenance rules.
1. Do
a) Keep connectors clean
b) Extend sleeve or connector nut when you store it
c) Place plastic end-caps after you use it
d) Insp ect all connectors carefully b efore every connection
e) Lo ok for metal particles, scratches, dents when you insp ect it
f) Align connectors carefully when you connect them
g) Make preliminary connection lightly
h) Turn connector nut only to tighten
i) Use a torque wrench for nal connection. Use the 5 lb-in torque wrench to connect
a male SMA to a female SMA or a female precision 3.5 mm. Use the 8 lb-in torque
wrench to connect a male precision 3.5 mm to female SMA connectors.
2. Do not
a) Touch mating plate surfaces
b) Set connectors contact-end down
c) Use a damaged connector
d) Apply b ending force to connector
e) Overtighten preliminary connection
f) Twist or screw in connectors
g) Tighten past \break" point of torque wrench 33
5.6 AT42085 S-Parameters and Noise Parameters
! File AT42085
! Vce=8V Ic=10mA
! S parameters, Common Emitter
0.1 .72 - 50 26.52 152 .014 73 .90 -16
0.5 .66 -139 11.23 103 .035 36 .53 -32
1.0 .65 -168 5.95 84 .037 39 .45 -33
1.5 .65 175 4.06 71 .045 46 .43 -36
2.0 .65 163 3.06 60 .054 51 .42 -41
2.5 .66 157 2.51 55 .063 60 .42 -42
3.0 .68 149 2.07 46 .072 65 .41 -48
3.5 .68 141 1.79 38 .085 64 .43 -55
4.0 .69 133 1.57 29 .104 64 .45 -61
4.5 .69 125 1.41 21 .119 63 .46 -66
5.0 .69 114 1.28 12 .139 58 .47 -71
5.5 .71 103 1.17 3 .161 55 .44 -76
6.0 .75 91 1.07 -6 .177 49 .40 -85
! Noise Parameters, Common Emitter
!FREQ Fopt GAMMA OPT RN/Zo
!GHZ dB MAG ANG -
0.1 1.1 0.05 16 0.13
0.5 1.2 0.06 77 0.13
1 1.3 0.10 131 0.12
2 2.0 0.24 -179 0.11
4 3.5 0.46 -128 0.25 34
5.7 Computer Data Acquisition for 2-Port S-parameters
NOTE: A new improved data acquisition to ol is now also available contact the
TA for additional information
Overview
The purp ose of this do cument is to outline the necessary steps to acquire data from the
network analyzer for use in a program such as MATLAB or ADS. The HP 8753C network
analyzer communicates though the GPIB (General Purp ose Interface bus) card installed in
the PC. A graphical interface for GPIB commands has been implemented in LabVIEW to
acquire data. To get started, set the network analyzer to Talker/Listener mo de from the
LOCAL menu button.
Loading a Calibration from the PC
Double click on the SaveRecall.vi icon lo cated on the Windows 95 desktop. Select Run from
the Op erate menu. The red stop sign on the LabVIEW menu to olbar indicates when the
program is nished. When the program is nished, select Exit from the File menu. Press
the LOCAL menu button on the front panel of the network analyzer to regain control of it.
Saving Sparameter Measurements
Lo cate the Daq4SP.vi icon on the Windows 95 desktop. Double click on this icon to start
the data acquisition virtual instrument in LabVIEW. This programs allows up to four mea-
surements; use the scroll bar on the right side of the screen to view the rest of the panel.
The default selections of the Data Format and S Parameter controls are set to measure all
four S parameters in Smith Chart format. You can change the Destination File to the path
where the data is to b e saved for a single measurement. Note that four separate les for data
will be created at the completion of the program. To p erform a measurement select Run
from the Op erate menu. The program may take some time to transfer the calibration les
to the network analyzer. If the le in the Destination File prompt currently exists another
dialogue will app ear, click Replace to overwrite the le. Program execution can b e halted at
anytime by pressing the red stop sign on the LabVIEW menu to ol bar. After the program
has executed, press the LOCAL menu button on the network analyzer to regain control of
it. If you are not making any more measurements from LabVIEW then pro ceed through the
following menu sequence: MENU, TRIGGER MENU, CONTINUOUS. This will removethe
network analyzer from HOLD mo de and allow you to make additional measurements from
the network analyzer.
Data Handling
The default path for the data is on the desktop in a folder called data, it should b e the only 35
path used for hard drivedata. For oppy disks, typ e a:\ file_name in the Destination File
indicator b efore program execution. Two metho ds of transp orting your data to the Region
4 computers are describ ed b elow.
(i) Using FTP to transfer les to Region 4 computers. To enable communications start
eXceed by selecting Programs/eXceed from the start menu lo cated in the lower left corner
of the screen. Select Programs/eXceed/FTP from the start menu. Op en a connection to
a remote account. Enter the server's name i.e. hector.eng.ohio-state.edu, username, and
password. Browse through the folders on the lo cal system to nd the data le. On the
remote system click on the folder that the data le is to be copied to. Drag and drop the
le to the remote host, this can be accomplished by clicking and holding on the left mouse
button and moving the le to the appropriate lo cation.
(ii) Ms-dos disks on Region 4 computers: Unix commands such as mdir, mcd, and mcopy
can b e used to manipulate les on a oppydisk. For example, to retrieve the data from disk
typ e: mcopy a:/test.dat ~/my_dir/logmag.dat
Extensive do cumentation is available in the man pages on these commands.
Imp orting in ADS
Note that for 4 les (one p er S-parameter) must b e transferred. The function glue.pl available
in ~roblin/perl can b e used to write the le in ADS format.
glue.pl S11.dat S12.dat S21.dat S22.dat > myfile.s2p
Imp orting in MATLAB
To load a data le into a vector in MATLAB use the load command. To separate the columns
of the resulting vector use a command similar to x = test(:,1), y = test(:,2). 36
5.8 Loading S-parameter data in ADS and MATLAB
In lab oratory 3 you need to compare the S parameters measured for the device you designed
and fabricated with those you obtained with your simulation. This app endix describ es several
way to do it:
To plot exp erimental data in ADS Move to the data directory of your pro ject directory
(for example HPEESOF/ee723 prj)
cd
cd HPEESOF/ee723_prj/data
You are now going to create a le using an editor (for example vi or emacs) to store your S
parameter data. Call it for example mydata.s2p: The le should lo ok like that
! Comments line: My data file ...
# GHz S MA R 50.0
! SCATTERING PARAMETERS :
2 0.95 -26 3.57 157 .04 76 .66 -14
3 0.93 -40 3.53 147 .05 69 .65 -20
4 0.89 -52 3.23 136 .06 62 .63 -26
Note that in the le ab ove:
Comment lines start with !
The # line de nes the units. In the example shown the Frequency is entered using
GHz. S indicates that S parameters are used. MA indicates that they are entered
using the amplitude (mag[Sij]) and the angle (ang[Sij]). R 50.0 means that 50 ohms is
the reference characteristic imp edance.
The data are intro duced in the following order:
Frequency mag[S11] ang[S11] mag[S21] ang[S21] mag[S12] ang[S12] mag[S22] ang[S22]
Once you have created your le and lo cated it in the data directory you can load it in ADS
by doing the following:
While in the schematics window, click on the library menu and select the menu item:
Linear Data File Elements. In the right side of the dialogue box select then the item:
S2P(2-Port S-parameter File) 37
In the new dialogue b ox which then app ears, clickonValue Options. The le mydata you
have created b efore and which is lo cated in the data directory should app ear. Select it and
click on OK. Then click on Apply and the le name is loaded. Click on OK to intro duce
this new device on your schematic. Then as usual you need to add two p orts and a ground
terminal and save your schematic under a name b efore plotting the S-paramters in a test
window.
Note that you can saveany les in ADS (circuit, layout) into a p ostscript le by clicking on
le and selecting print/plot setup. Select Graphics, File and Postscript Gray Scale and then
OK. Nowwhenyou select plot in the le menuyou will b e prompted for the name of the le
you want to save it to: e.g., mylayout.ps
Plotting ADS Data on MATLAB(Recommended Approach)
It might be more convenient to do the reverse: that is bring the ADS simulation data in
MATLAB. This gives you much more control. Bringing simulated data in ADS for compar-
ison with measured data can give you more control on the plot.
For this purp ose you need to save your simulation data in a le while b eing in ADS. To do
so go in your test window in ADS. Click on Library, scroll down the menu and then clickon
Output Files. Then select SNP (S-parameter File). In the value eld enter the name of the
le for example my le. ADS will automatically add an extension such as .s2p for a 2 p ort
device (my le.s2p), .s4p fora4 p ort device (my le.s4p).
Lo cate the OFILE item in your test window and simulate. It would be convenient to select
a reasonable number of frequency points such as 20 p oints. The le will be created in the
data sub directory of your pro ject directory.
Before loading this le on MATLAB you need to clean it. Use an editor and remove the text
headers:
! Communications Design Suite 5.0 306 Aug 26 1994 (c) 1993 Hewlett-Packard
! Mon Nov 27 11:34:56 1995 lange_tb\Lange
# MHz S MA R 50.0000
! SCATTERING PARAMETERS :
Note that the header sp eci es the data typ e: S stands for scattering parameter, MA indicates
that the scattering parameters are represented by their magnitude and angle. 50 gives the
characteristic imp edance.
For each frequency the data is presented in the following format for a4p ort: 38
freq magS11 angS11 magS12 angS12 magS13 angS13
magS21 angS21 magS22 angS22 magS23 angS23
magS31 angS31 magS32 angS32 magS32 angS33
magS41 angS41 magS42 angS42 magS42 angS43
For each frequency (freq) consolidate with your editor the three lines into a single one
freq magS11 angS11 magS12 angS12 magS13 angS13 magS21 angS21 magS22 angS22
magS23 angS23 magS31 angS31 magS32 angS32 magS32 angS33 magS41 angS41 magS42
angS42 magS42 angS43
To p erform these task automatically you can use the script le s#p2text:
s#p2text
The les my le.s4p and my le.text and s#p2text in the ER4 directory
/tmp_mnt/user2/faculty/ro blin /la tex for you to insp ect them.
Assume you saved your le my le.text You can now load your data in MATLAB using:
>> load myfile.text
Now the frequencies are stored in
>> myfile(:,1)
The amplitude of S11 is stored in
>> myfile(:,2)
To plot simply typ e:
>> plot (myfile(:,1),myfile(:,2) )
To save into an encapsulated p ostscript le typ e:
>> print -deps2 myfile.eps 39
Some of you have several plots they would like to combine together in a single plot and
p ostscript le. This is easily done with MATLAB. Four small plots (2x2) would b e generated
using:
>>subplot(2,2,1), plot (myfile(:,1),myfile(:,2) )
>>subplot(2,2,2), plot (myfile(:,1),myfile(:,4) )
>>subplot(2,2,3), plot (myfile(:,1),myfile(:,6) )
>>subplot(2,2,4), plot (myfile(:,1),myfile(:,8) )
Typ e help plot or help subplot to check how to set your titles and axes. 40
5.9 Active Biasing Circuit
(from Ralph S. Carson, High Frequency Ampli ers, 1982)
An active biasing circuit is shown in Fig. 5.16a. Here, pnp transisto or Q , assumed to op erate
1
in its active region, helps stabilize the op erating p oint of transistor Q , hence the name active
2
biasing. The basic op eration of the circuit is as follows. If I increases, more voltage of the
C 2
p olarity indicated in Fig. 5.16 app ears across R , and this decreases the forward bias for
3
the emitter-base circuit of Q , so I decreases. Since I = I = I , a decrease in I
1 E 1 C 1 B 2 1 E 1 E 1
causes I to decrease. But a decrease in I leads to a decrease in I ,so this opp oses the
B 2 B 2 C 2
increase in I assumed originally. Therefore bias stabilization is achieved.
C 2
The analysis of the active biasing circuit cannot followthe T -equivalent metho d b ecause of
the active device Q . Instead, straightforward circuit analysis can be carried out after sub-
1
stituting an appropriate circuit mo del for each transistor, as shown in Fig. 5.16b. Neglecting
I ,forCase I,we nd, from Fig. 5.16b
co
I = I = I
C 2 2 B 2 2 1 E 1
Since ' 1, this b ecomes
1
I ' I (1)
C 2 2 E 1
Also,
I = I + I ' I + I = I (1 + )
R E 1 C 2 E 1 2 E 1 E 1 2
3
Using KVL around the V R R lo op and substituting (1 )= 1=(1 + ), we obtain
CC 1 2
!
1
(R + R )+I I R = V (2)
1 2 E 1 R 1 CC
1
1+
1
and, using KVL around the R R V lo op yields
2 3 BE 1
(R ) I (1 + )R = V (3) I
2 E 1 2 3 BE 1 R
2
Solve 2and 3 for I and substitute that result into 1 to obtain, for negligible I ,
E 1 co
i h
R
2
V
V
BE 1 2 CC
R +R
1 2
(4) I ' (1 + )
C 2 1
R +(1+ )(1 + )R
4 1 2 3
where R = R R =(R + R ). If (1 + )(1 + )R R ; I is indep endent of and
4 1 2 1 2 1 2 3 4 C 2 1
varies as =(1 + ),and bias stabilization is obtained. It is seen that the collector current
1 2 41
I do es not dep end on V b ecause the current source I , in the simpli ed transistor
C 2 BE 2 1 E 1
mo del is indep endent of V .
BE 2
Therefore for large and wehave(1+ ) (1 + ) R =R and the collector current
1 2 1 2 4 3
of the transistor Q (microwave transistor) is
2
R
2
V V
CC BE 1
R +R
1 2
I =
CC
R 3
SMA E SMA B C
E Q2
Port1 Port 2
RF RF
Network Analyzer HP 8753C
DC bias DC bias
BNC BNC
Q1
R3 R2 Vcc
R1
IR3 I C2
Vcc R3 RF IE1 R IR2 Choke 2 Cdecoupling I IB1 B2 Q1 RF out Q2 RF Choke Microwave I R1 R1 Transistor RF in
Cdecoupling
Cdecoupling
Figure 5.1: (a)Overal active biasing circuit and (b) active bias analysis 42
5.10 Lab oratory Rep ort Forms 43
Name:
Date:
EE 723 Adjunct Lab oratory | LABORATORY #1 Rep ort
Step 4) of Part I:
Center frequency: f =
0
Bandwidth: B=
S21(f0) dB =jS (f )j =
21 0 dB
(Include S plot in app endix)
21
jS (f )j =
12 0 dB
jS (f )j (SW R)=
11 0
jS (f )j (SW R)=
22 0
Step 5 of Part I:
Include the swept power gain compression plot in app endix.
jS (P = 10dBm)j (in dB) =
21 in
P ( 1dB drop of jS j)=
in 21
Step 6 of Part II:
Include the Noise Figure versus frequency plot in app endix. Noise Figure at 1.8 GHz (with
Power on):
F (1.8 GHz) =
Minimum detectable input signal (See Gonzalez, p. 354):
P dBm =
i;mds
Discussion/Comments: 44
Name:
Date:
EE 723 Adjunct Lab oratory | LABORATORY #2 Rep ort
Include the Smith Chart plots of S ; S ; S ; and S for the AT42085 from 300 kHz to 3
21 12 22 11
GHz measured with the network analyzer.
Include a listing of the scattering parameters for the frequencies given in App endix 5.6.
Using ADS compare in a single Smith Chart plot, all the measured scattering parameters
with those given in App endix 5.6. App endix 5.6 gives the pro cedure to load your data in
ADS. 45
Name:
Date:
EE 723 Adjunct Lab oratory | PROPOSAL for LABORATORY #3
Pro ject Title:
Design Goals:
Design Pro cedure:
List of parts:
1AT42085 transistor
Duroid circuit b oard
2SMA connectors 46
Name:
Date:
EE 723 Adjunct Lab oratory | LABORATORY #3 Rep ort
Ampli er characteristics:
center frequency: f =
0
3dB bandwidth: B =
jS (f )j =
21 0 dB
(Include S21 plot in app endix)
jS (f )j =
12 0 dB
jS (f )j (SW R)=
11 0
jS (f )j (SW R)=
22 0
Optional: Include the swept power gain compression plot in app endix.
jS (P = 10dBm)j (in dB) =
21 in
P ( -1 dB drop of jS j )=
in 21
Noise Figure at 1.5 GHz (with Power on):
Minimum detectable input signal (See Gonzalez, p. 176):
P =
i;mds 47