Alexander Lee, Darrin Albers, and Steven C. Reising

Lee et al., FR3.T03 IGARSS 2011 Vancouver, B.C. Canada July 29, 2011 1

DEVELOPMENT OF INTERNALLY-CALIBRATED, MMIC-BASED MILLIMETER-WAVE RADIOMETERS OPERATING AT 130 AND 166 GHZ IN SUPPORT OF THE SWOT

MISSION

Alexander Lee, Darrin Albers, and Steven C. Reising

Microwave Systems Laboratory, Colorado State University, Fort Collins, CO

Pekka Kangaslahti, Shannon T. Brown, Douglas E. Dawson, Oliver Montes, Todd C. Gaier, Daniel J. Hoppe, and Behrouz Khayatian

Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA

IGARSS 2011 Vancouver, B.C. Canada

Surface Water and Ocean Topography (SWOT) Mission

Accelerated Tier-2 U.S. National Research Council Earth Science Decadal Survey Mission planned for launch in 2020 (NASA/CNES partnership)

• Oceanography Objectives:– Characterize ocean mesoscale and sub-mesoscale circulation at

spatial resolutions of 10 km and larger (1-cm ht. precision required)• Kinetic energy / Heat and carbon air-sea fluxes• Climate change and ocean circulation• Coastal and internal tides

• Hydrology Objectives:– To provide global height measurements of inland surface water bodies

with area greater than 250 m2 and rivers with width greater than 100 m– To measure change in global water storage in these inland water

bodies and river discharge on sub-monthly to annual time scales

Lee et al., FR3.T03 July 29, 2011 2


Scientific Motivation• Current satellite ocean altimeters include a nadir-viewing, co-located

18-37 GHz multi-channel microwave radiometer to measure wet-tropospheric path delay. Due to the large diameters of the surface instantaneous fields of view (IFOV) at these frequencies, the accuracy of wet path retrievals begins to degrade at approximately 40 km from the coasts.

• Conventional altimeter-correcting microwave radiometers do not provide wet path delay over land.

Along track direction

High frequency footprint

Low frequency footprint

Along track directionAlong track direction

High frequency footprintHigh frequency footprint

Low frequency footprintLow frequency footprint

LandOcean• In order to enable wet path delay

measurements closer to the coastline and increase the potential for over land measurements higher-frequency microwave channels (90-170 GHz) are being considered for the SWOT mission



SWOT Mission Concept Study

Low frequency-only algorithm

Low frequency-only algorithm

Low and High frequency algorithm

Low and High frequency algorithm

High-resolution Weather Research and Forecasting (WRF) model results show reduced wet path-delay error using both low-frequency (18-37 GHz) and high-frequency (90-170 GHz) radiometer channels.



SWOT ACT Objectives

• Develop key radiometer components to enable additional low mass, low power, high frequency radiometers for the SWOT mission

• Design and fabricate a tri-frequency feed horn with integrated triplexer covering 90 to 170 GHz

• Design and fabricate PIN-diode switches and noise diodes for internal calibration from 90 to 170 GHz that can be integrated into the receiver front end

• Integrate and test components in MMIC-based low-mass, low-power, small-volume radiometer at 92, 130 and 166 GHz with the tri-frequency feed horn



PIN-Diode Switch Design

• Microwave switches were designed to cover three frequency ranges of 80-105 GHz, 90-135 GHz, and 160-190 GHz

• Monolithic microwave integrated circuits (MMIC) were realized in microstrip and coplanar waveguide technology

• Fabricated using Northrop Grumman’s 75-μm thick InP MMIC PIN diode process

• PIN diodes used because of low insertion loss and fast switching speeds

• Variations of each SPDT design with PIN diode sizes ranging from 3 to 8 μm were fabricated

• To date, 80-105 GHz and 90-135 GHz switches have been tested; 160-190 switches have not yet been tested



80-105 GHz MMIC Switch

Asymmetric Design

Isolation

Insertion Loss

Common Leg RL

Antenna Leg RL

Measured Performance

Antenna Leg

Common Leg

Reference Leg

1.52 mm

1.37

mm

Common Leg RL (Integrated Ref. Load Version)

Asymmetric design variation with integrated 50-Ω

reference termination


• Microstrip design

• SiN 2-layer MIM capacitors for bypass and DC blocking capacitors

• NiCr thin-film process for resistors

• Radial stubs used to provide well-defined virtual RF shorts

• Antenna and Common legs aligned and Reference leg at a 90°angle



Post-Fabrication On-Chip Tuning of Isolation

• Higher frequency measurements demonstrated isolation was optimized for higher frequency

• By increasing effective electrical length of shunt diode radial stubs, optimal isolation was lowered to frequency range of interest


Tuning ribbon added to shunt diode radial

stub

Isolation (Un-tuned)

Isolation (Tuned)




Symmetric Design

Antenna Leg

Reference Leg

Common Leg

1.52 mm

1.37

mm

Isolation

Insertion Loss

Common Leg RL

Antenna Leg RL


Preliminary tuning of shunt diode radial stub

demonstrates decrease in isolation optimal frequency

Same technology as 80-105 GHz design (microstrip, SiN 2-layer MIM capacitors, etc.)




Symmetric Design • Coplanar waveguide design



Simulated PerformanceAntenna

LegReference

Leg

Common Leg

1.10 mm

0.97

m

m

Insertion Loss

Isolation

Common Leg RL

Antenna Leg RL

Lee et al., FR3.T03 July 29, 2011 10


PIN Diode Switch Results

Switch Insertion Loss

Return Loss Isolation Comments

80-105 GHz <2 dB >15 dB >15 dB

80-105 GHz (Asymmetric) <2 dB >18 dB >15 dB

Isolation >20 dB from 85-103 GHz after on-chip

tuning90-135 GHz <2 dB >15 dB >8 dB

160-190 GHz <2 dB >20 dB >20 dB Simulated Results Only

Lee et al., FR3.T03 July 29, 2011 11


System Block DiagramWaveguide

Components92-GHz

multi-chip module

130-GHz multi-chip module

166-GHz multi-chip module

Common radiometerback end, thermal

control anddata

subsystem

Tri-Frequency Feed Horn

Coupler

Coupler

Noise Diode

Coupler

Noise Diode

Noise Diode

MMIC Components

Lee et al., FR3.T03 July 29, 2011 12

Lee et al., FR3.T03 IGARSS 2011 Vancouver, B.C. Canada July 29, 2011 13

• These Dicke radiometers use four LNAs to provide sufficient signal level at the input to the detector.

• Direct-detection architecture is the lowest power and mass solution for these high-frequency receivers. Keeping the radiometer power at a minimum is critical to fit within the overall SWOT mission constraints, including the power requirements of the radar interferometer.

130- and 166-GHz Radiometer Design


130-GHz Predicted Performance

Lee et al., FR3.T03 July 29, 2011 14

Component Gain (dB) Noise Figure (dB)Cumulative Noise

TemperatureDirectional Coupler -1.3 1.3 101

Waveguide Through Line -0.3 0.3 129Waveguide to Microstrip transition -0.5 0.5 180

Switch -1.9 1.9 438LNA 18 3.3 1267

Interconnect Attenuator -2.8 2.8 1278LNA 18 3.3 1278

Wide BPF -1.25 1.25 1303Attenuator -4.15 4.15 1304



Interconnect Attenuator -2.8 2.8 1307Narrow BPF -3.25 3.25 1308

Waveguide to Microstrip transition -0.5 0.5 1308

Total Gain (dB) 47.70Receiver noise figure (dB) 7.41

Receiver noise temperature (K) 1308



Lee et al., FR3.T03 July 29, 2011 15

0.001 0.010 0.100 1.000 10.0000

0.10.20.30.40.50.60.70.80.9

1

Microwave Humidity Sounder on Orbit at 89 GHzAdvanced Microwave Sounding Unit on Orbit at 89 GHz

Integration Time (Sec)

NE∆T (K)



Lee et al., FR3.T03 July 29, 2011 16

Component Gain (dB) Noise Figure (dB)Cumulative Noise

Temperature

Directional Coupler -2.00 2.00 170

Waveguide to Microstrip transition -0.50 0.50 226Switch -1.90 1.90 509LNA 17.00 3.30 1418

LNA - includes interconnect loss 14.00 3.30 1436Filter -2.20 2.20 1436

LNA - includes interconnect loss 14.00 3.30 1437LNA - includes interconnect loss 14.00 3.30 1437

Filter -2.20 2.20 1437Waveguide to Microstrip transition -0.50 0.50 1437

Total Gain (dB) 49.70Total Noise Figure (dB) 7.75

Total Noise Temperature (K) 1437



Lee et al., FR3.T03 July 29, 2011 17

0.001 0.01 0.1 1 10 1000

0.2

0.4

0.6

0.8

1

1.2

Microwave Humidity Sounder on Orbit at 89 GHz

Advanced Microwave Sounding Unit on Orbit at 89 GHz

Integration Time (Sec)

NE∆T (K)


166-GHz Band Pass Filter:Return Loss

140 145 150 155 160 165 170 175 180-40

-35

-30

-25

-20

-15

-10

-5

0

HFSS Enclosed

HFSS Air

Measured Return Loss

Frequency (GHz)

-Ret

urn

Loss

(dB

)0.94” (2.4 mm)

Lee et al., FR3.T03 July 29, 2011 18

The passive high-frequency microwave components were designed and fabricated in microstrip technology on 3-mil (75 μm) thick alumina substrates.


140 145 150 155 160 165 170 175 180-40

-35

-30

-25

-20

-15

-10

-5

0

HFSS Enclosed

HFSS Air

Measured Insertion Loss

Frequency (GHz)

-Inse

rtion

Los

s (d

B)

160 161 162 163 164 165 166 167 168 169 170-10

-8

-6

-4

-2

0

Frequency (GHz)

-Inse

rtion

Los

s (d

B)

166-GHz Band Pass Filter:Insertion Loss

Lee et al., FR3.T03 July 29, 2011 19


166-GHz Matched Load


Lee et al., FR3.T03 July 29, 2011 20

0.029” (.74 mm)

140 145 150 155 160 165 170 175 180-45-40-35-30-25-20-15-10-50

HFSS enclosedHFSS airMeasured Return Loss

Frequency (GHz)

-Ret

urn

Loss

(dB

)


0.8”(20.3 mm)

1.75” (44.5 mm)

1.45” (36.8 mm

)

0.077”(1.96 mm)

SW LNA LNA BPF-1 BPF-2ATN

WTM

WTM

0.094”(2.39 mm)

50Ω

LNA LNA

130-GHz Multi-Chip Module

Lee et al., FR3.T03 July 29, 2011 21



Lee et al., FR3.T03 July 29, 2011 22



0.093”(2.4 mm)

Lee et al., FR3.T03 July 29, 2011 23


Multi-Chip Module Assembly

Lee et al., FR3.T03 July 29, 2011 24


Summary

Lee et al., FR3.T03 July 29, 2011 25

• The addition of high frequency radiometers on ocean altimetry missions will enable wet tropospheric path delay correction closer to the coastline.

• Key radiometer component technologies are under development to enable additional high frequency radiometers operating at 92 GHz, 130 GHz, and 166 GHz for the upcoming SWOT mission.

• High frequency switches have been designed and fabricated for all three high frequency radiometers.

• Switch testing has been completed on the 92 GHz and 130 GHz switches. The test results show less than 2 dB insertion loss and greater than 15 dB return loss. Additional tuning is required to optimize the isolation.

• Prototype radiometers at 130 GHz and 166 GHz have been designed and are in the process of being fabricated.


Backup Slides

Lee et al., FR3.T03 July 29, 2011 26


Move to Higher Frequency• Supplement low-

frequency, low-spatial resolution channels with high-frequency, high-spatial resolution channels to retrieve PD near coast

• High-frequency window channels sensitive to water vapor continuum

• 183 GHz channels sensitive to water vapor at different layers in atmosphere

22.235 GHz (H2O)

55-60 GHz (O2)118 GHz (O2)

183.31 GHz (H2O)

Lee et al., FR3.T03 July 29, 2011 27


Design Topology

Series-Shunt PIN-diode SPDT Switch Implementation

• Implements both series and shunt diode SPDT configurations together to maximize isolation

• Eliminates the need for quarter-wave transformer (reduces size)

• This configuration was used for SPDT switch designs being presented

RF OUTPUT RF OUTPUT

RF INPUT

Lee et al., FR3.T03 July 29, 2011 28


Design Topology

SPDT Switch Circuit Schematic

Lee et al., FR3.T03 July 29, 2011 29



Symmetric Design

Antenna Leg

Reference Leg

Common Leg

1.52 mm

1.37

mm

Isolation

Insertion Loss

Common Leg RL

Antenna Leg RL

• Microstrip design



• Radial stubs used to provide well-defined virtual RF shorts


Lee et al., FR3.T03 July 29, 2011 30



Measured Results vs. Simulated Results

Lee et al., FR3.T03 July 29, 2011 31


130-GHz Low-Noise Amplifier

• MMIC LNA was packaged in WR-8 and WR-10 housings for characterization over a broad bandwidth.

1

1.5

2

2.5

3

3.5

4

4.5

5

5.5

6

0

5

10

15

20

25

75 90 105 120 135

Noi

se F

igur

e [d

B]

Gai

n [d

B]

Frequency [GHz]

Gain dB G[dB]

NF dB NF [dB]

Lee et al., FR3.T03 July 29, 2011 32


166-GHz Low-Noise Amplifier

• 35-nm process InP HEMT• Three-stage design with separate gate bias for the first stage to optimize low-noise performance• Record low noise temperature of 300 K from 150 - 160 GHz• Chip area of 900 x 560 (μm)2 • The LNA was mounted in optimized WR-08 and WR-05 waveguide housings to test over a broad bandwidth.

Low-Noise Amplifier Layout and Measured Response

Lee et al., FR3.T03 July 29, 2011 33


120

122

124

126

128

130

132

134

136

138

140

-45-40-35-30-25-20-15-10

-50

Measured

HFSS

HFSS air

Frequency (GHz)

-Ret

urn

Loss

(dB

)

130-GHz Band Pass Filter:Return Loss

0.94” (2.4 mm)


Lee et al., FR3.T03 July 29, 2011 34


120 122 124 126 128 130 132 134 136 138 140-40

-35

-30

-25

-20

-15

-10

-5

0

Mea-sured

HFSS

HFSS Air

Frequency (GHz)

Inse

rtion

Los

s (d

B)

125 126 127 128 129 130 131 132 133 134 135-10

-8

-6

-4

-2

0

Frequency (GHz)

Inse

rtion

Los

s (d

B)

Note: Correction for CPW losses included

130-GHz Band Pass Filter:Insertion Loss

Lee et al., FR3.T03 July 29, 2011 35

Alexander Lee, Darrin Albers, and Steven C. Reising

Documents

Transcript of Alexander Lee, Darrin Albers, and Steven C. Reising