III-V COMPOUND SEMICONDUCTORS

SUMMARY

Although there are several wide-bandgap semiconductor technologies emerging (SiC, GaN etc.) the major part of the research work in this sub-theme will focus only on GaN power device technology. This is because RF power devices are one of the most important parts in remote sensing applications and have a direct impact on the overall sensor performance. Compared to Sic materials GaN technology offers a much higher frequency of operation. The proposed research work will range from developing new device topologies to first evaluation prototypes of power amplifiers.

In addition smaller activities are proposed to exploit already existing and matured GaAs technologies in terms of transferring expensive mmW device technologies to a high volume lower cost process (mHEMT technology) and also in terms of developing MMIC’s (Microwave Monolithic Integrated Circuits) to improve current sensor performance.

MILITARY BENEFITS

Current solid-state devices, based on GaAs, are approaching their theoretical limit in output power. Higher microwave power levels are typically required by remote sensing systems, satellite systems, electronic warfare and microwave communication systems. Recent developments in GaN-based materials have yielded pHEMT devices that show maximum RF power densities with up to a 10-fold increase in RF power density when compared to GaAs-based devices.

For remote sensing applications GaN offers not only the potential for solid-state devices with higher output power, but also provides considerable advantages in terms of operating voltage, matching efficiency, life time, robustness and operating temperature. The improved DC-to-RF conversion efficiency of GaN based transmitters offers a major simplification to the whole transmitter chain. Because GaN devices have a similar noise figure to GaAs devices, but offer a much higher power handling capability, the dynamic range of receivers can potentially be increased by a factor of 10. An input protection (limiter) at the receiver will not generally be required, because of the inherent robustness of the GaN devices. Therefore GaN technology (for power and for low noise applications) will have a significant impact on the overall sensor mass, the antenna architecture and the overall system complexity and performance compared to existing GaAs and vacuum tube technology currently used in radar systems.

Early access to GaN solid-state devices will offer major advantages for aerospace, radar and communication systems. Most important are the manufacturing cost of the solid-state devices as they represent a significant part of the overall system cost (e.g. currently up to 50% of the T/R module in a phased array radar). It is therefore essential, when developing semiconductor devices, to take the constraints of a high volume, low cost production line into account. Only device technologies, which are manufacturable on high volume low cost production lines, will be affordable for military applications.

RESEARCH OBJECTIVE

To evaluate GaN devices in terms of their output power, gain and DC-to-RF efficiency for radar transmitter applications.
To evaluate GaN devices for receivers, limiters and switches in terms of their noise figure, linearity, and dynamic range.
To investigate the reliability, ruggedness, robustness (EMP and closely located high power emitters) of GaN devices.
To develop a GaN device technology that is affordable and easily transferable to an industrial high volume production line
To research new device topologies, other than the conventional HEMT, which potentially yield even higher output power, linearity and noise performance.
To develop an affordable GaAs millimetre wave power and low noise m-HEMT technology (100GHz), which is transferable to an industrial high volume production line.
To develop new concepts in circuit design to enhance the overall sensor performance and implement these concepts into existing industrial GaAs MMIC processes.

RESEARCH OUTLINE

The proposed research activity for wide bandgap devices (GaN) will consider the already ongoing research activities in the UK and should enhance the already obtained results. The research in this field has advanced to a stage where power devices are capable almost of matching the existing GaAs pHEMT technology; hence demonstrator circuits should be developed to evaluate the performance parameters, advantages and shortfalls of the new GaN technology. The demonstrators should be representative of a real sensor application. Of special interest are the efficiency, linearity, reliability and ruggedness of the devices when operated at higher output power levels.

Another activity at a more basic research level should investigate new GaN compound materials and device topologies. Recent publications suggest that related materials and other devices (e.g. inverted HEMTs or HBTs) could yield even better performance especially for military applications (higher power density). This is a longer-term research activity that should provide first working devices at lower frequency and power levels.

Many semiconductor companies are retooling for 6″ diameter GaAs wafers, mainly to reduce manufacturing cost and to increase the yield and consistency of their production. GaN’s economic potential is currently limited by the carrier wafer technology (Sic), which is very expensive and not available for the next 7 years in 6″ wafers (NCSR, US-DoD roadmap). In order to be able to provide affordable GaN devices in the near future, it is important to seek other carrier wafer technologies, such as GaN on Silicon, to make the GaN technology directly compatible with already established low cost industrial production methods. Therefore a research activity to provide GaN devices on Si substrate, which can be processed on a 6″ semiconductor production line, is proposed by the Consortium.

The highest operating frequency of 0.25mm p-HEMT structures in GaAs and in GaN is limited to about 50GHz. For higher frequencies the device structure has to be reduced, which implies a lower yielding and more expensive manufacturing process (electron beam). Another option is the use of InP semiconductor material. InP wafers are very fragile, expensive and currently not available in 6″ diameter. The m-HEMT technology is based on InP on GaAs and hence combines the high frequency performance advantage of InP with the GaAs processing compatibility advantage for low cost industrial manufacturing lines. In order to address sensor applications in excess of 100GHz a research activity is planned to evaluate power and low noise m-HEMT technology, that is manufacturable on an industrial production line.

In future sensor systems, even when high performance technologies are available, a significant part of the RF front end will still be realised in traditional GaAs technology. Hence a smaller activity is planned to develop novel circuit concepts, which are tailored to enhance the performance of radar systems. To prove the concept a prototype should be designed and implemented in GaAs MMIC technology.

CO-ORDINATION WITH EXISTING / PREVIOUS RESEARCH

These research activities will build on previous research work in the UK when accessible and available.

Leave a Reply

Your email address will not be published. Required fields are marked *