Distributed Amplifier Monolithic Microwave Integrated ... · A very broadband distributed amplifier was designed using a ... mobility transistor ... which uses the input and output
Post on 10-Jun-2018
219 Views
Preview:
Transcript
Distributed Amplifier Monolithic Microwave Integrated
Circuit (MMIC) Design
by John E. Penn
ARL-TR-6237 October 2012
Approved for public release; distribution unlimited.
NOTICES
Disclaimers
The findings in this report are not to be construed as an official Department of the Army position
unless so designated by other authorized documents.
Citation of manufacturer’s or trade names does not constitute an official endorsement or
approval of the use thereof.
Destroy this report when it is no longer needed. Do not return it to the originator.
Army Research Laboratory Adelphi, MD 20783-1197
ARL-TR-6237 October 2012
Distributed Amplifier Monolithic Microwave Integrated
Circuit (MMIC) Design
John E. Penn
Sensors and Electron Devices Directorate, ARL
Approved for public release; distribution unlimited.
ii
REPORT DOCUMENTATION PAGE Form Approved OMB No. 0704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the
data needed, and completing and reviewing the collection information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing the
burden, to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302.
Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently valid
OMB control number.
PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.
1. REPORT DATE (DD-MM-YYYY)
October 2012
S
2. REPORT TYPE
Final
3. DATES COVERED (From - To)
4. TITLE AND SUBTITLE
Distributed Amplifier Monolithic Microwave Integrated Circuit (MMIC) Design
5a. CONTRACT NUMBER
5b. GRANT NUMBER
5c. PROGRAM ELEMENT NUMBER
6. AUTHOR(S)
John E. Penn
5d. PROJECT NUMBER
5e. TASK NUMBER
5f. WORK UNIT NUMBER
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)
U.S. Army Research Laboratory
ATTN: RDRL-SER-E
2800 Powder Mill Road
Adelphi, MD 20783-1197
8. PERFORMING ORGANIZATION
REPORT NUMBER
ARL-TR-6237
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES)
10. SPONSOR/MONITOR’S ACRONYM(S)
11. SPONSOR/MONITOR'S REPORT
NUMBER(S)
12. DISTRIBUTION/AVAILABILITY STATEMENT
Approved for public release; distribution unlimited.
13. SUPPLEMENTARY NOTES
14. ABSTRACT
A very broadband distributed amplifier was designed using a 0.13-µm gallium arsenide (GaAs) pseudomorphic high electron
mobility transistor (PHEMT) process from TriQuint Semiconductor. The design and fabrication of this circuit was performed
as part of the fall 2011 Johns Hopkins University Monolithic Microwave Integrated Circuit (MMIC) Design Course, taught by
the author. The design approach is applicable to very broadband, low noise MMICs that could be used for a variety of radio
frequency (RF) and microwave systems.
15. SUBJECT TERMS
MMIC, distributed amplifier
16. SECURITY CLASSIFICATION OF: 17. LIMITATION
OF
ABSTRACT
UU
18. NUMBER
OF
PAGES
28
19a. NAME OF RESPONSIBLE PERSON
John E. Penn a. REPORT
Unclassified
b. ABSTRACT
Unclassified
c. THIS PAGE
Unclassified
19b. TELEPHONE NUMBER (Include area code)
(310) 394-0423
Standard Form 298 (Rev. 8/98)
Prescribed by ANSI Std. Z39.18
iii
Contents
List of Figures iv
1. Introduction 1
2. Distributed Amplifier Design 1
3. Sonnet Simulations 6
4. Testing 13
5. Conclusion 17
6. References 18
List of Symbols, Abbreviations, and Acronyms 19
Distribution List 20
iv
List of Figures
Figure 1. Distributed amplifier with three stages. ...........................................................................2
Figure 2. Distributed amplifier with three stages using lumped element feeds. .............................2
Figure 3. Gain measurements vs. simulations for a 0.5-µm PHEMT distributed amplifier (2006). ........................................................................................................................................3
Figure 4. S-parameter performance of an ideal distributed amplifier design (thin lines) vs. the TriQuint elements (thick lines). ...........................................................................................4
Figure 5. S-parameter performance of an ideal distributed amplifier design (thin lines) vs. the DC to 30 GHz amplifier with TriQuint elements (thick lines). ...........................................4
Figure 6. Schematic of an ideal distributed amplifier design. ........................................................5
Figure 7. Schematic of the TriQuint element distributed amplifier design. ...................................5
Figure 8. Layout plot of the TriQuint element distributed amplifier designs on a 54x54 mil die. ..............................................................................................................................................6
Figure 9. Sonnet electromagnetic (EM) layout plot of the DC-30 GHz distributed amplifier. ......7
Figure 10. Sonnet EM simulation of the DC-30 GHz distributed amplifier with a 31-GHz stability problem. .......................................................................................................................8
Figure 11. Sonnet EM layout plot of the DC-30 GHz distributed amplifier “redo.” ......................9
Figure 12. Sonnet EM simulation of the DC-30 GHz distributed amplifier “redo” with improved stability. .....................................................................................................................9
Figure 13. Layout plot of the modified TriQuint element distributed amplifier designs (54x54 mil)...............................................................................................................................10
Figure 14. Sonnet EM current plot of the modified distributed amplifier at 30 GHz. ..................11
Figure 15. Sonnet EM current plot of the modified distributed amplifier at 2 GHz. ....................11
Figure 16. Sonnet EM current plot of the original distributed amplifier at 30 GHz. ....................12
Figure 17. Sonnet EM current plot of the original distributed amplifier at 2 GHz. ......................12
Figure 18. Noise figure and gain performance of the DC-30 GHz distributed amplifier design. ......................................................................................................................................13
Figure 19. Simulated noise figure: distributed amplifier designs (red-1G+, black-DC+). ...........14
Figure 20. Measured output power, gain, and efficiency performance of the 1‒30 GHz distributed amplifier design at 10 GHz. ..................................................................................14
Figure 21. Simulated output power, gain, and efficiency performance of the 1‒30 GHz distributed amplifier design at 10 GHz. ..................................................................................15
Figure 22. Measured output power, gain, and efficiency performance of the DC-30 GHz distributed amplifier design at 10 GHz. ...................................................................................15
v
Figure 23. Simulated output power, gain, and efficiency performance of the DC-30 GHz distributed amplifier design at 10 GHz. ..................................................................................16
Figure 24. A 3-D view of the MMIC distributed amplifier layout (1 GHz+). ..............................16
vi
INTENTIONALLY LEFT BLANK.
1
1. Introduction
Very broadband gain can be achieved with distributed amplifiers. The idea of using transmission
lines to connect the input and output feeds of parallel gain stages, each with small but broadband
gain, was proposed in 1936, but gained popularity after a 1948 paper from Bill Packard (co-
founder of Hewlett-Packard Co.) et al. using vacuum tubes for the gain stages (1). While many
monolithic microwave integrated circuits (MMICs) are optimized for a particular band to achieve
optimal performance, there are many applications that require a broadband response from DC to
very high microwave frequencies: such as fiber-optic drivers, electronic warfare (EW) receivers,
and other broadband radio frequency (RF) sensors or communications systems.
The distributed amplifier designs documented here were fabricated as part of the Johns Hopkins
University (JHU) MMIC Design Course in fall 2011. A similar distributed amplifier design
using TriQuint’s TQPED 0.5-µm gallium arsenide (GaAs) pseudomorphic high electron mobility
transistors (PHEMTs) was designed and fabricated along with the student designs for the 2006
JHU MMIC Design Course. That earlier 0.5-µm PHEMT design worked well from 1 to 10 GHz
and was published in the November 2007 issue of Microwaves and RF (2). This distributed
amplifier design uses a much higher frequency TriQuint TQP13 0.13-µm PHEMT process,
resulting in more than two decades of gain bandwidth. Depending on the frequency, space
available, etc., the “transmission line” feeds of the distributed amplifier can be distributed,
typically microstrip, or the feed can be a lumped element transmission line equivalent (3). This
distributed amplifier uses the lumped element transmission line equivalent approach, but could
have used microstrip line feeds.
2. Distributed Amplifier Design
In the distributed amplifier, a signal travels down the input transmission line feeding each of
several parallel gain stages (figure 1). Each gain stage has a small amount of gain, but the gain is
broadband, summing up at the output transmission line to provide an amplified signal.
Typically, there is a cutoff frequency limiting the number of stages that can be paralleled.
Depending on frequency, size available, device characteristics, etc., it may be beneficial to
replace the distributed transmission line with a lumped element equivalent of a transmission line,
which uses the input and output capacitance of a transistor gain stage tuned to inductances in the
feed lines to control the impedances and gain over a broad operating band (figure 2).
2
Figure 1. Distributed amplifier with three stages.
Figure 2. Distributed amplifier with three stages using lumped element feeds.
The drain (output) and gate (input) feed lines of the distributed amplifier design are terminated in
a 50-ohm resistor. This tends to dissipate half the output power, resulting in lower efficiency.
There is a method using tapered output lines to try and reduce this 3-dB loss, but the
implementation can be difficult to achieve, except in simulations using ideal elements. In the
lumped element approach, the inductances in the gate and drain lines are adjusted to create a
matched 50-ohm input and output impedance over a broad range with relatively flat gain. The
number of transistor stages and the size of the device can be tuned to optimize the performance.
Figure 3 shows the good agreement between simulations of the previous 0.5-µm PHEMT
distributed amplifier of 2006 and measurements of the fabricated circuit.
3
dB(S(2,1))
dB(DA3ms3_3V..S(2,1))
dB(DA3ms3V..S(2,1))
dB(Damp3B_tqs_mln_msdev..S(2,1))
1 2 3 4 5 6 7 8 9 10 110 12
-5
0
5
10
15
-10
20
freq, GHz
dB
(S(2
,1))
dB
(DA
3m
s3_3V
..S
(2,1
))dB
(DA
3m
s3V
..S
(2,1
))dB
(Dam
p3B
_tq
s_m
ln_m
sdev..S
(2,1
))
Forward Transmission, dB
dB(S(2,1))
dB(DA3ms3_3V..S(2,1))
dB(DA3ms3V..S(2,1))
dB(Damp3B_tqs_mln_msdev..S(2,1))
ADS/Sonnet
Measured
Figure 3. Gain measurements vs. simulations for a 0.5-µm PHEMT distributed amplifier (2006).
Using a much higher frequency 0.13-µm PHEMT process from TriQuint, an updated distributed
amplifier was designed. A lumped element approach was maintained with some initial tuning of
the transistor size along with the drain and gate feed inductors to obtain a first cut of the
amplifier design. Then, ideal elements were replaced with actual “lossy” passive MMIC
elements and re-tuned based on the simulations. Two nearly similar versions of the distributed
amplifier were created with one subtle difference. The first design uses a spiral inductor
connected in parallel to the 50-ohm drain termination, which reduces DC power consumption.
This limits the low end frequency response to about 1 GHz, and, as the design has two decades
of bandwidth, the resonance of this inductor causes a slight ripple in the gain (~20 GHz), as
shown in the preliminary S-parameter simulation of figure 4. The simulation compares the
amplifier using TriQuint’s TQP13 passive elements, without interconnect, versus ideal lumped
elements. In the alternate layout, the DC supply current through the 50-ohm drain termination
reduces the power efficiency, but allows gain at much lower frequencies than supplying DC
current through the inductor. Figure 5 shows a similar simulation comparing the alternate DC to
a 30-GHz amplifier using TriQuint’s TQP13 passive elements, without interconnect, versus ideal
lumped elements. The preliminary schematic using ideal elements is shown in figure 6, with the
“lossy” TQP13 schematic in figure 7. Again, the only difference in the two designs is the DC
supply in the upper top left of the schematics. Originally, an external DC gate bias was to be
supplied, but it can be difficult to make a broadband DC supply for a distributed amplifier over
two decades of bandwidth. The gate (input) DC supply was simplified by terminating the
50-ohm load resistor to a substrate ground via to provide a fixed 0 V “Vgs”. This removes the
flexibility of controlling the drain current, but it also removes any resonances and potential
4
instabilities that might be created in the gate DC bias supply. For these TQP13 PHEMT devices,
0 V “Vgs” bias provides good gain, and good noise figure at a moderate DC power consumption.
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Frequency (GHz)
SParam_1G_Damp
-20
-15
-10
-5
0
5
10
15
DB(|S(2,2)|)Damp11
DB(|S(1,1)|)Damp11
DB(|S(2,1)|)Damp11
DB(|S(2,1)|)Damp11_idl
DB(|S(1,1)|)Damp11_idl
DB(|S(2,2)|)Damp11_idl
Figure 4. S-parameter performance of an ideal distributed amplifier design (thin lines) vs.
the TriQuint elements (thick lines).
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Frequency (GHz)
SParam_DC_Damp2
-20
-15
-10
-5
0
5
10
15
0.7963 GHz-10.1 dB
31.13 GHz-10.14 dB7.42 GHz
-9.689 dB
0.315 GHz11.14 dB
20.05 GHz9.983 dB
DB(|S(2,2)|)
Damp11_2
DB(|S(1,1)|)
Damp11_2
DB(|S(2,1)|)
Damp11_2
DB(|S(2,1)|)
Damp11_idl
DB(|S(1,1)|)
Damp11_idl
DB(|S(2,2)|)
Damp11_idl
Figure 5. S-parameter performance of an ideal distributed amplifier design (thin lines)
vs. the DC to 30 GHz amplifier with TriQuint elements (thick lines).
5
DC_VID=V5Sweep=NoneV=3 V
INDID=L10L=1000 nH
1
2
3
TQP13_PHSS_T3iID=PHSSi1W=WfNG=4TQP13_PHSS_T3_MB=PHSS_T3
TQP13_SVIAID=TQ_VIAL_1
1
2
3
TQP13_PHSS_T3iID=PHSSi2W=WfNG=4TQP13_PHSS_T3_MB=PHSS_T3
TQP13_SVIAID=TQ_VIAL_2
1
2
3
TQP13_PHSS_T3iID=PHSSi3W=WfNG=4TQP13_PHSS_T3_MB=PHSS_T3
TQP13_SVIAID=TQ_VIAL_3
TQP13_SVIAID=TQ_VIAL_4
TQP13_SVIAID=TQ_VIAL_5
1
TQP13_PADID=TQ_BP_2
1
TQP13_PADID=TQ_BP_3
1
TQP13_PADID=TQ_BP_4
TQP13_HF_CAPID=C1C=12 pFW=175 um
TQP13_HF_CAPID=C2C=12 pFW=175 um
TQP13_HF_CAPID=C3C=12 pFW=175 um
TQP13_HF_RESWID=R2R=50 OhmW=20 umTYPE=NiCrSTAT=No StatisticsR_STAT=0R_TOL=0
TQP13_HF_RESWID=R3R=50 OhmW=20 umTYPE=NiCrSTAT=No StatisticsR_STAT=0R_TOL=0
1
2
3
TQP13_PHSS_T4mID=PHSSi4W=WfNG=2TQP13_PHSS_T4_MB=PHSS_T4
1
2
3
TQP13_PHSS_T4mID=PHSSi5W=WfNG=2TQP13_PHSS_T4_MB=PHSS_T4
1
2
3
TQP13_PHSS_T4mID=PHSSi6W=WfNG=2TQP13_PHSS_T4_MB=PHSS_T4
1
2
3
TQP13_PHSS_T4mID=PHSSi7W=WfNG=2TQP13_PHSS_T4_MB=PHSS_T4
TQP13_SVIAID=TQ_VIAL_6
1
2
3
TQP13_PHSS_T3iID=PHSSi8W=WfNG=4TQP13_PHSS_T3_MB=PHSS_T3
INDID=L2L=100 nH
INDID=L5L=LLd2 nH
INDID=L6L=LLd2 nH
INDID=L7L=LLd nH
INDID=L8L=LLd nH
INDID=L9L=LLd nH
INDID=L1L=LLg2 nH
INDID=L3L=LLg2 nH
INDID=L11L=LLg nH
INDID=L12L=LLg nH
INDID=L13L=LLg nH
PORTP=1Z=50 Ohm
PORTP=2Z=50 Ohm
Ld=0.3
Ld2 = Ld / 2
Wf=40
Lg=0.3
Lg2 = Lg / 2
W=10
L1d=102 L1g=80
L1d2=62 L1g2=71L2d2=78 L2g2=83
L2d=137 L2g=101
LLd2=0.235
LLd=0.595
LLg2=0.195
LLg=0.235
Figure 6. Schematic of an ideal distributed amplifier design.
DC_VID=V5Sweep=NoneV=3 V
28.4 mA
3 V
0 V
INDID=L10L=1000 nH
28.4 mA 3 V
INDID=L11L=1000 nH
0 mA
0 V
DC_VID=V7Sweep=NoneV=0 V
0 mA
0 V
1
2
3
TQP13_PHSS_T3iID=PHSSi1W=WfNG=4TQP13_PHSS_T3_MB=PHSS_T3
TQP13_SVIAID=TQ_VIAL_1
7.11 mA
1
2
3
TQP13_PHSS_T3iID=PHSSi2W=WfNG=4TQP13_PHSS_T3_MB=PHSS_T3
TQP13_SVIAID=TQ_VIAL_2
7.1 mA
1
2
3
TQP13_PHSS_T3iID=PHSSi3W=WfNG=4TQP13_PHSS_T3_MB=PHSS_T3
TQP13_SVIAID=TQ_VIAL_3
7.09 mA
TQP13_SVIAID=TQ_VIAL_4
0.000119 mA
TQP13_SVIAID=TQ_VIAL_5
0 mA
1
TQP13_PADID=TQ_BP_1
1
TQP13_PADID=TQ_BP_2
0 mA
1
TQP13_PADID=TQ_BP_3
0 mA
1
TQP13_PADID=TQ_BP_4
0 mA
TQP13_HF_CAPID=C1C=12 pFW=175 um
0 mA
2.93 V
0 V
TQP13_HF_CAPID=C2C=12 pFW=175 um
0 mA
0 V
6.2e-6 V
TQP13_HF_CAPID=C3C=12 pFW=175 um
0 mA
0 V
TQP13_HF_CAPID=C4C=12 pFW=175 um
TQP13_HF_RESWID=R1R=2000 OhmW=4 umTYPE=NiCrSTAT=No StatisticsR_STAT=0R_TOL=0
TQP13_HF_RESWID=R2R=50 OhmW=20 umTYPE=NiCrSTAT=No StatisticsR_STAT=0R_TOL=00.000119 mA
6.13e-6 V
1.42e-9 V
TQP13_HF_RESWID=R3R=50 OhmW=20 umTYPE=NiCrSTAT=No StatisticsR_STAT=0R_TOL=0
1.06 mA
2.95 V
TQP13_MRIND2ID=L1W=10 umS=10 umN=24LVS_IND="LVS_Value"
1
2
3
TQP13_PHSS_T4mID=PHSSi4W=WfNG=2TQP13_PHSS_T4_MB=PHSS_T4
3e-5 mA
7.11 mA
7.11 mA
8.53e-5 V
1
2
3
TQP13_PHSS_T4mID=PHSSi5W=WfNG=2TQP13_PHSS_T4_MB=PHSS_T4
2.97e-5 mA
7.1 mA
7.1 mA
8.52e-5 V
1
2
3
TQP13_PHSS_T4mID=PHSSi6W=WfNG=2TQP13_PHSS_T4_MB=PHSS_T4
2.95e-5 mA
7.09 mA
7.09 mA
8.5e-5 V
TQP13_MRIND2ID=L12W=10 umS=10 umN=5LVS_IND="LVS_Value"
21.3 mA
2.93 V
2.94 V
TQP13_MRIND2ID=L14W=10 umS=10 umN=5LVS_IND="LVS_Value"
3e-5 mA
6.19e-6 V
6.2e-6 V
TQP13_MRIND2ID=L13W=10 umS=10 umN=5LVS_IND="LVS_Value"
14.2 mA
2.93 V
TQP13_MRIND2ID=L15W=10 umS=10 umN=5LVS_IND="LVS_Value"
5.97e-5 mA
6.18e-6 V
TQP13_MRIND2ID=L16W=10 umS=10 umN=5LVS_IND="LVS_Value"
28.4 mA
TQP13_MRIND2ID=L17W=10 umS=10 umN=5LVS_IND="LVS_Value"
0 mA2.93 V
TQP13_MRIND2ID=L18W=10 umS=10 umN=5LVS_IND="LVS_Value"
0 mA
TQP13_MRIND2ID=L19W=10 umS=10 umN=5LVS_IND="LVS_Value"
0.000119 mA6.16e-6 V
TQP13_MRIND2ID=L20W=10 umS=10 umN=5LVS_IND="LVS_Value"
7.08 mA
TQP13_MRIND2ID=L21W=10 umS=10 umN=5LVS_IND="LVS_Value"
8.92e-5 mA
1
2
3
TQP13_PHSS_T4mID=PHSSi7W=WfNG=2TQP13_PHSS_T4_MB=PHSS_T4
2.94e-5 mA
7.08 mA
7.08 mA
8.5e-5 V
TQP13_SVIAID=TQ_VIAL_6
7.08 mA
1
2
3
TQP13_PHSS_T3iID=PHSSi8W=WfNG=4TQP13_PHSS_T3_MB=PHSS_T3
TQP13_MRIND2ID=L3W=10 umS=10 umN=16L1=350 umL2=200 umLVS_IND="LVS_Value"
27.3 mA
INDID=L2L=1000 nH
PORTP=1Z=50 Ohm
0 mA
0 V
PORTP=2Z=50 Ohm
0 mA
0 V
Ld=0.3
Ld2 = Ld / 2
Wf=40
Lg=0.3
Lg2 = Lg / 2
W=10
L1d=102 L1g=80
L1d2=62 L1g2=71L2d2=78 L2g2=83
L2d=137 L2g=101
Figure 7. Schematic of the TriQuint element distributed amplifier design.
The layout of both distributed amplifiers on a 54x54 mil die is shown in figure 8. At the top of
the die is the DC to 30-GHz distributed amplifier and the bottom of the die is the 1 to 30 GHz
version.
6
Figure 8. Layout plot of the TriQuint element distributed amplifier designs on a 54x54 mil die.
3. Sonnet Simulations
Both designs are essentially identical except for the spiral inductor in the drain bias. Sonnet was
used to electromagnetically simulate the layout, which showed a potential stability problem at
higher frequency (~30 GHz) (figures 9 and 10). One concern is the relatively large decoupling
7
capacitor on the drain DC bias and its relatively close spacing to the spiral inductors of the drain
feed line. To improve the design, the 12-pF large capacitor was split into two 6-pF capacitors to
allow a shorter path to ground for the higher frequency operation while maintaining a 12-pF total
capacitance for the lower frequency operation. Also, the capacitor was moved further away from
the spiral inductors of the drain feed to reduce parasitic coupling. Lastly, 5-ohm resistors on the
drain were added in the simulation, which significantly reduced the previous stability problem
(~30 GHz) (figures 11 and 12). A modified layout with the stabilizing resistors and modified
capacitor is shown in figure 13 for the same 54x54 mil Die.
Figure 9. Sonnet electromagnetic (EM) layout plot of the DC-30 GHz distributed amplifier.
8
Figure 10. Sonnet EM simulation of the DC-30 GHz distributed amplifier with a 31-GHz stability
problem.
The current view of Sonnet can help give insight into the parasitic coupling of the actual layout.
At 2 GHz, the electric field strength around the two 6-pf capacitors is approximately symmetric
but at 30 GHz, the field is stronger for the left side of the first 6-pf capacitor, particularly at the
substrate via a ground connection. Propagation tends to take the shortest path, which led to the
premise that at higher frequencies the smaller 6-pf capacitor would be better than a much larger
12-pf capacitor. The larger 12-pf capacitor would also have a lower resonant frequency. Some
more simulations could be beneficial to determine a more optimal approach. Figures 14, and 15
show the field strength around the two 6-pf capacitors at 2 and 30 GHz using the current viewer
in Sonnet. Likewise, the original layout with a single 12-pf capacitor is shown in figures 16 and
17 for comparison.
9
Figure 11. Sonnet EM layout plot of the DC-30 GHz distributed amplifier “redo.”
Figure 12. Sonnet EM simulation of the DC-30 GHz distributed amplifier “redo” with improved stability.
10
Figure 13. Layout plot of the modified TriQuint element distributed amplifier designs (54x54 mil).
11
Figure 14. Sonnet EM current plot of the modified distributed amplifier at 30 GHz.
Figure 15. Sonnet EM current plot of the modified distributed amplifier at 2 GHz.
12
Figure 16. Sonnet EM current plot of the original distributed amplifier at 30 GHz.
Figure 17. Sonnet EM current plot of the original distributed amplifier at 2 GHz.
13
4. Testing
When the fabricated student designs were returned, unfortunately, the original distributed
amplifier designs were inadvertently submitted, rather than the modified designs. During testing,
both distributed amplifiers showed promise at lower bias voltages and at their nominal design
voltages, but appeared to be conditionally stable at certain intermediate voltages. These
oscillations appeared to be well above 45 GHz, but would disappear near the nominal design
voltages of 3 V (1–30 GHz) and 4 V (DC-30 GHz), and higher DC biases. Figure 14 shows a
plot of measured performance for the two designs, which differ only by the DC supply bypass
inductor. Noise figure and gain was measured at 4-V bias for the DC-30 GHz version, showing
slightly higher noise figure than expected, but decent, at 3 dB from 15 to 26 GHz (highest
frequency measured) (figure 18). Simulations of noise figure show a similar shape but were
expected to be better, closer to 2 dB (figure 19). Efficiency and output power versus input power
are plotted in figure 20 for the more efficient 1–30 GHz amplifier at 10 GHz showing a
respectable 20% power added efficiency (PAE) at 2-dB gain compression. The simulated
performance is reasonably similar as shown in figure 21. Likewise, the performance of the less
efficient DC-30 GHz distributed amplifier is shown in figure 22. Those measurements are also
very similar to the simulations shown in figure 23. A three-dimensional (3-D) view of the layout
of the distributed amplifier (1‒30 GHz) is shown in figure 24.
0
1
2
3
4
5
6
7
8
9
10
0.0 3.0 6.0 9.0 12.0 15.0 18.0 21.0 24.0 27.0
dB
Freq
Dist Amp 4VNFCorr
GainCorr
Figure 18. Noise figure and gain performance of the DC-30 GHz distributed amplifier design.
14
Figure 19. Simulated noise figure: distributed amplifier designs (red-1G+, black-DC+).
17.68
19.03
10.179.65
7.98
16.8
20.5
0
5
10
15
20
25
-4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12
Pin
DA1 10GHz Meas 114.5V ~43mA
Pout
Gain
PAE
Figure 20. Measured output power, gain, and efficiency performance of the 1‒30 GHz distributed
amplifier design at 10 GHz.
15
Figure 21. Simulated output power, gain, and efficiency performance of the 1‒30 GHz distributed
amplifier design at 10 GHz.
13.68
9.36
7.58
11.7
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
-4 -3 -2 -1 0 1 2 3 4 5 6 7
Pin
DA2 10GHz Meas 114V ~38mA
Pout
Gain
PAE
Figure 22. Measured output power, gain, and efficiency performance of the DC-30 GHz distributed
amplifier design at 10 GHz.
16
-5 -4 -3 -2 -1 0 1 2 3 4 5 6 7
Power (dBm)
Pcomp_DA_1G_DC
0
5
10
15
p3
p2
p1
5.93 dBm13.86 dBm
5.98 dBm11.96
5.99 dBm7.886 dB
DB(|Pcomp(PORT_2,1)|)[*,X] (dBm)Damp11_lay_mlin_PO.AP_HB
PAE(PORT_1,PORT_2)[*,X]Damp11_lay_mlin_PO.AP_HB
DB(PGain(PORT_1,PORT_2))[*,X]Damp11_lay_mlin_PO.AP_HB
p1: Freq = 10 GHz
p2: Freq = 10 GHz
p3: Freq = 10 GHz
Figure 23. Simulated output power, gain, and efficiency performance of the DC-30 GHz distributed
amplifier design at 10 GHz.
Figure 24. A 3-D view of the MMIC distributed amplifier layout (1 GHz+).
17
5. Conclusion
Both versions of the distributed amplifier worked well, with the 1–30 GHz version providing
higher output power and efficiency, while the DC-30 GHz version had gain below 1 GHz and
avoided the gain ripple from the resonance due to the DC bypass inductor. There appeared to be
some conditional stability concerns, which might make the design less robust over process
variation in repeat fabrications. Hopefully, there will be a future fabrication of the modified
designs to test for improved stability.
18
6. References
1. Ginzton, E. L.; Hewlett, W. R.; Jasberg, J. H.; Noe, J. D. Distributed Amplification. Proc.
IRE 1948, 956–69.
2. Penn, J. E. Designing MMIC Distributed Amplifiers. Microwaves & RF November 2007.
3. Penn, J. E. Convert Distributed MICs to MMICs. Microwaves & RF July 2003.
19
List of Symbols, Abbreviations, and Acronyms
3-D three-dimensional
EM electromagnetic
EW electronic warfare
GaAs gallium arsenide
JHU Johns Hopkins University
MMIC monolithic microwave integrated circuit
PAE power added efficiency
PHEMTs pseudomorphic high electron mobility transistors
RF radio frequency
20
1 DEFENSE TECHNICAL
(PDF INFORMATION CTR
only) DTIC OCA
8725 JOHN J KINGMAN RD
STE 0944
FORT BELVOIR VA 22060-6218
1 DIRECTOR
US ARMY RESEARCH LAB
IMAL HRA
2800 POWDER MILL RD
ADELPHI MD 20783-1197
1 DIRECTOR
US ARMY RESEARCH LAB
RDRL CIO LL
2800 POWDER MILL RD
ADELPHI MD 20783-1197
1 DIRECTOR
US ARMY RESEARCH LAB
RDRL CIO LT
2800 POWDER MILL RD
ADELPHI MD 20783-1197
12 DIRECTOR
US ARMY RESEARCH LAB
RDRL SER
PAUL AMIRTHARAJ
RDRL SER E
ROMEO DEL ROSARIO
BEN HUEBSCHMAN
JAMES WILSON
TONY IVANOV
JOHN PENN (3 HCS)
ROBERT PROIE
ROBERT REAMS
PANKAJ SHAH
ED VIVEIROS
2800 POWDER MILL RD
ADELPHI MD 20783-1197
top related