185 GHz Monolithic Amplifier in InGaAs/InAlAs Transferred-Substrate HBT Technology

1
185 GHz Monolithic Amplifier in InGaAs/InAlAs
Transferred-Substrate HBT Technology
M. Urteaga, D. Scott, T. Mathew, S. Krishnan, Y.
Wei, M. Rodwell. Department of Electrical and
Computer Engineering, University of California,
Santa Barbara
urteaga_at_ece.ucsb.edu 1-805-893-8044
IMS2001 May 2001, Phoenix, AZ
2
Outline
IMS2001
UCSB

• Introduction
• Transferred-Substrate HBT Technology
• Circuit Design
• Results
• Conclusion

3
Transferred-Substrate HBTs
IMS2001

• Substrate transfer allows simultaneous scaling
  of emitter and collector widths
• Maximum frequency of oscillation
• Sub-micron scaling of emitter and collector
  widths has resulted in record values for
  extrapolated fmax (gt1 THz)
• Promising technology for ultra-high frequency
  tuned circuit applications

30
Mason's
3000 Å collector 400 Å base with 52 meV
grading AlInAs / GaInAs / GaInAs HBT
gain, U
25
20
MSG
Gains, dB
H
f
1.1 THz ??
21
max
f
204 GHz
Emitter, 0.4 x 6 mm2
t
Collector, 0.7 x 6 mm2
I
6 mA, V
1.2 V
c
ce
10
100
1000
Frequency, GHz
4
Ultra-high Frequency Amplifiers
IMS2001

• Applications for electronics in 140-220 GHz
  frequency band
• Wideband communication systems
• Atmospheric sensing
• Automotive radar
• Amplifiers in this frequency band realized in
  InP-based HEMT technologies
• 3-stage amplifier with 30 dB gain at 140 GHz.
• Pobanz et. al., IEEE JSSC, Vol. 34, No. 9,
  Sept. 1999.
• 3-stage amplifier with 12-15 dB gain from
  160-190 GHz
• Lai et. al., 2000 IEDM, San Francisco, CA.
• 6-stage amplifier with 20 ? 6 dB from 150-215
  GHz.
• Weinreb et. al., IEEE MGWL, Vol. 9, No. 7,
  Sept. 1999.
• This Work
• Single-stage tuned amplifier with 3.0 dB gain at
  185 GHz
• First HBT amplifier in this frequency range
• Gain-per-stage is comparable to HEMT technology

5
InGaAs/InAlAs HBT Material System
IMS2001
Layer Structure
Band Diagram
2kT base bandgap grading
Bias conditions for the band diagram Vbe 0.7
V Vce 0.9 V
6
Device Fabrication I
IMS2001
7
Transferred-Substrate Process Flow
IMS2001

• emitter metal
• emitter etch
• self-aligned base
• mesa isolation
• polyimide planarization
• interconnect metal
• silicon nitride insulation
• Benzocyclobutene, etch vias
• electroplate gold
• bond to carrier wafer with solder
• remove InP substrate
• collector metal
• collector recess etch

8
Device Fabrication II
IMS2001
9
Ultra-high fmax Devices
IMS2001

• Electron beam lithography used to define
  submicron emitters and collectors
• Minimum feature sizes
• 0.2 ?m emitter stripe widths
• 0.3 ?m collector stripe widths
• Improved collector-to-emitter alignment using
  local alignment marks
• Future Device Improvements
• Carbon base doping
• na gt1.0 x 1020 cm-3
• significant reduction in Rbb
• DHBTs with InP Collectors
• Greater than 6 V BVCEO

0.3 ?m Emitter before polyimide planarization
0.4 ?m Collector Stripe
10
Device Measurements
IMS2001
DC Measurements
Measured RF Gains

• Device dimensions
• Emitter area 0.4 x 6 ?m2
• Collector area 0.7 x 6.4 ?m2
• ? 20
• BVCEO 1.5 V

• Bias Conditions
• VCE 1.2 V, IC 4.8 mA
• f? 160 GHz
• Measurements of unilateral power gain in
  140-220 GHz frequency band appear to show
  unphysical behavior

11
Amplifier Design
IMS2001
Simulation Results

• Simple common-emitter design conjugately matched
  at 200 GHz using shunt-stub tuning
• Shunt R-C network at output provides low
  frequency stabilization
• Simulations predicted 6.2 dB gain
• Designed using hybrid-pi model derived from
  DC-50 GHz measurements of previous generation
  devices
• Electromagnetic simulator (Agilents Momentum)
  was used to characterize critical passive
  elements

S21
S11, S22
Circuit Schematic
12
IMS2001
Design Considerations in Sub-mmwave Bands

• Transferred-substrate technology provides low
  inductance microstrip wiring environment
• Ideal for Mixed Signal ICs
• Advantages for MMIC design
• Low via inductance
• Reduced fringing fields
• Disadvantages for MMIC design
• Increased conductor losses
• Resistive losses are inversely proportional to
  the substrate thickness for a given Zo
• Amplifier simulations with lossless matching
  network showed 2 dB more gain
• Possible Solutions
• Use airbridge transmission lines
• Find optimum substrate thickness

13
140-220 GHz VNA Measurements
IMS2001

• HP8510C VNA used with Oleson Microwave Lab
  mmwave Extenders
• Extenders connected to GGB Industries coplanar
  wafer probes via short length of WR-5 waveguide
• Internal bias Tees in probes for biasing active
  devices
• Full-two port T/R measurement capability
• Line-Reflect-Line calibration performed using
  on-wafer transmission line standards

UCSB 140-220 GHz VNA Measurement Set-up
14
Amplifier Measurements
IMS2001

• Measured 3.0 dB peak gain at 185 GHz
• Device dimensions
• Emitter area 0.4 x 6 ?m2
• Collector area 0.7 x 6.4 ?m2
• Device bias conditions
• Ic 3.0 mA, VCE 1.2 V

Measured Gain
Measured Return Loss
Cell Dimensions 690?m x 350 ?m
15
Simulation vs. Measurement
IMS2001
Simulation versus Measured Results

• Amplifier designed for 200 GHz
• Peak gain measured at 185 GHz
• Possible sources for discrepancy
• Matching network design
• Device model

16
Matching Network Design
IMS2001
Matching Network Breakout Simulation Vs.
Measurement

• Breakout of matching network without active
  device was measured on-wafer
• Measurement compared to circuit simulation of
  passive components
• Simulations show good agreement with measurement
• Verifies design approach of combining E-M
  simulation of critical passive elements with
  standard microstrip models

S21
S11
S22
Red- Simulation Blue- Measurement
17
Device Modeling I Hybrid-Pi Model
IMS2001
HBT Hybrid-Pi Model Derived from DC-50 GHz
Measurements

• Design used a hybrid-pi device model based on
  DC-50 GHz measurements
• Measurements of individual devices in 140-220
  GHz band show poor agreement with model
• Discrepancies may be due to weakness in device
  model and/or measurement inaccuracies
• Device dimensions
• Emitter area 0.4 x 6 ?m2
• Collector area 0.7 x 6.4 ?m2
• Bias Conditions
• VCE 1.2 V, IC 4.8 mA

18
IMS2001
Device Modeling II Model vs. Measurement
S21

• Measurements and simulations of device
  S-parameters from 6-45 GHz and 140-220 GHz
• Large discrepancies in S11 and S22
• Anomalous S12 believed to be due to excessive
  probe-to-probe coupling
• Red- Simulation
• Blue- Measurement

S11, S22
S12
19
Simulation vs. Measurement
IMS2001
UCSB
Simulation versus Measured Results Simulation
Using Measured Device S-parameters

• Simulated amplifier using measured device
  S-parameters in the 140-220 GHz band
• Simulations show better agreement with measured
  amplifier results
• Results point to weakness in hybrid-pi model
  used in the design
• Improved device models are necessary for better
  physical understanding but measured S-parameter
  can be used in future amplifier designs

20
Conclusions
IMS2001
UCSB

• Demonstrated first HBT amplifier in the 140-220
  GHz frequency band
• Simple design provides direction for future high
  frequency MMIC work in transferred-substrate
  process
• Observed anomalies in extending hybrid-pi model
  to higher frequencies
• Future Work
• Multi-stage amplifiers and oscillators
• Improved device performance for higher frequency
  operation
• Acknowledgements
• This work was supported by the ONR under grant
  N0014-99-1-0041
• And the AFOSR under grant F49620-99-1-0079