Report

Teaser, summary, work performed and final results

Periodic Reporting for period 1 - GRACE (GaN mm-wave Radar Components Embedded)

Teaser

Environmental impact from the aviation industry is a topic of intense discussions. In concrete numbers air transport emissions are relatively low (2% of the total but emitted at high level in the atmosphere), the aviation community is committed to reducing their environmental...

Summary

Environmental impact from the aviation industry is a topic of intense discussions. In concrete numbers air
transport emissions are relatively low (2% of the total but emitted at high level in the atmosphere), the
aviation community is committed to reducing their environmental impact in light of increasing demand for extra
passenger capacity. In 2001, the Advisory Council for Aeronautical Research in Europe (ACARE) defined
stringent targets to be achieved by aircraft technology maturing in 2020 compared to state of-the-art technology of
the year 2000.

The climate change is an important concern, ACARE target a 50% reduction of CO2 emissions from the commercial flying
segment, 40% by greener technologies and another 10% by a more efficient Air Traffic Management (ATM).
The project we are detailing here is inserted in the ATM segment. The main application target is an Enhanced Flight
Vison System (EFVS) that permits take-off, landing and taxiing in poor visibility conditions. EFVSs are traditionally based
on a forward looking infrared (IR) camera which gives a thermal image of the world. However, IR sensors are sensitive
to environmental factors such as rain and fog. Further, they are not as good as Radars for motion
detection. High-resolution motion detection may be fulfilled with a mm-wave (mmW) radar.

In contrast to passive RF imaging systems that can be implemented with only a low noise amplifier and a detector, an
active imaging system, e.g., a radar system, requires also a Power Amplifier (PA) and a Signal Source (SS) but offers
a much higher detection range and a much better definition. Short gate-length GaN HEMT technology is the most
promising candidate to deliver mmW PAs and SSs. The only process commercially open in Europe is the D01GH process
from OMMIC, which is a 100 nm gate-length process on Si substrate. OMMIC also has a 60 nm gate length process
D006GH in their roadmap.

The packaging of W-band components is traditionally based on chip-and-wire technology which provides the technical
requirements but at a high cost. Significant cost reduction may be achieved with standard surface mount technologies
(SMT) used at lower frequencies. Most mmW packages are different types of Quad Flat No (QFN) leads packages.
In this project, we intend using a fan-out wafer level (FOWL) packaging approach offered by Fraunhofer IZM.

In summary there are four primary technical objectives to be addressed in the GRACE project:
• Design of a 93-100 GHz Power Amplifier (PA) MMIC with an output power in the range of 500mW to 1W
• Design of a signal-source MMIC covering 93-100 GHz with state-of-the art far-carrier phase noise performance
• Development of a FOWL packaging flow for the PA MMIC
• Development of a FOWL packaging flow for the signal source MMIC

Work performed

Overall, the project has made a very good progress in achieving the project objectives. the first round of MMIC design was
closed by delivery of chips at the final date of the reporting period. Also the packaging design under the lead of Fraunhofer IZM
is effectively materialized.

In WP1, system level trade-offs were reconsidered and it was agreed that the project should aim at multiple version chips. On
the one hand side highest possible performance without regards to complexity and on the other hand side –good
performance/complexity trade off. OMMIC and MC2 reviewed performance of PAs designed in TWEETHER for input. Signal source
specifications were revised by solid feasibility analysis. Fraunhofer IZM investigated the thermal behavior of the proposed package
configuration for the embedding of the PA.

In WP2 device characterization validated that the models in OMMIC’s D006GH design kit are sufficiently accurate within expected
process variations with exception for parasitics that were compensated for. MC2 has designed four different versions of PAs.
Chalmers has designed a TRL cal kit including through lines and 50 ohm load as well as several VCOs, frequency multipliers,
frequency dividers, and buffer amplifiers. The pdk models converge well and simulations predict promising and reasonable results.
Fraunhofer IZM has designed, modeled and simulated planar and vertical interconnects in the 93-100 GHz frequency range for the
characterization of the FOWLP packaging technology applied for the fabrication of RF packages embedding both VCO and PA.
Planar resonators were desugbed for the characterization of the electromagnetic properties (permittivity and loss tangent) of
the RDL material up to 94 GHz. Finally IZM has applied the interconnects models to the design, modeling, simulation and optimization
of the complete RF signal path. Chip-package co-design was applied for the design of external resonators in the package operating at
12 GHz and 23 GHz for the VCO chip. The full packaging structures (with both PA and VCO) including RF signal path, DC supply and
external resonator were designed and simulated.

In WP3 “Manufacturing of MMICs and FOWLPs” OMMIC delivered wafers according to original time plan in month 12.
Measurement results will be used for package design optimization. OMMIC has shipped first MMICs to MC2 and Chalmers.

WP5 “Project management and dissemination” has progressed according to project plan.

Final results

The GRACE project aims at significantly improving performance and reducing cost of mmW radar components for next generation
enhanced flight vision systems (EFVS). Critical building blocks are monolithic microwave integrated circuit (MMIC) power amplifiers
and signal sources. Critical technologies are a short-gate length wide-band gap semiconductor technology, e.g., OMMIC’s D01GH
or D006GH GaN HEMT technology and a suitable surface mount packaging technology, e.g., a fan-out-wafer-level (FOWL) packaging.

The goals of GRACE as set out on components level, e.g., for the MMIC PA and the MMIC SS are very competitive as compared to
state of the art, for the respective functionality. A fully integrated 100 GHz PA with an output power level of 1 W is near the
state-of-the art in the open research, and it is also a functionality needed in telecommunications, high resolution radars and
passive imaging products. In this perspective the impact could not be higher. Regarding the signal source, the target phase noise
performance has not been reached at mmW frequencies, and certainly not in MMIC technology. The functionality is also increasingly
demanded as the interest in mmW commercial applications is growing.

On top of the high ambitions with respect to development of MMIC technology and circuit design, the GRACE project also aims at
packaging the MMICs, using a fan-out-wafer-level (FOWL) packaging flow. The FOWL package is optimized to manage the thermal
requirements of the PA and SS MMICs as well as the low loss required at mmW frequencies. Today, there are no standard packaging
solutions available for mmW components, and certainly not for power components with high thermal constraints. In this perspective,
the success of the GRACE project would have significant impact, not only for EFVS systems as aimed in this topic. It will also be very
important for the European competitiveness in the growing millimetre-wave communication and sensing industry.

Website & more info

More info: http://www.chalmers.se/en/departments/mc2/research/grace/Pages/News.aspx.