Download this article in .PDF format
This file type includes high resolution graphics and schematics when applicable.

A deluge of information is currently available about GaN’s performance and its potential in the marketplace. Some sources claim that it will overtake both the GaAs and LDMOS markets, among others. Lance Wilson, Research Director of RF Devices for ABI Research, sums up these considerations with the following: “Engineers need to look at device technology very clear eyed. There is a lot of hype right now about how GaN technology can do no wrong. GaN is not viable for every application and it can have cost issues. Above 3 GHz for high-power RF, though, GaN is generally a no-brainer.”  

When asked about competition between LDMOS and GaN, Lance responded, “GaN is starting to capture some wireless-infrastructure sockets-- mostly for remote radio heads and active antennas. But silicon LDMOS still rules this space. It will be hard for GaN to really dislodge LDMOS for lower-frequency markets due to cost, but the GaN vendors are making progress in that regard GaN is still several times the price for die than LDMOS. As you go up in frequency, where the packaging and assembly become more costly than die costs, the cost difference starts to level out. For instance, an S-band-radar LDMOS device costs about the same as a S-band-radar GaN part.”

Of course, much hype has been devoted in recent years to whether GaN can replace GaAs in the bulk of its application. Lance summarizes, “Where the present application power line between GaN and GaAs lies is still a point of discussion. Some say 10 W and some say as much as 35 to 40 W, but 20 to 25 W seems to be a reasonable space.” His guess is that applications under 25 W or so are likely to remain GaAs.

The development of GaN on SiC devices is a main target for its potential to enable new high-power and high-frequency markets. This could include wireless backhaul, military/defense, satellites, and many more. The main drivers from these markets require this device technology to lower costs as the cost of SiC as a material and the small wafer sizes are limiting the cost conscious markets. Higher switching frequencies along with the inefficiencies caused by normally-on switches are other characteristics necessary for larger adoption of this technology.

As this technology naturally is very robust, reliable, radiation hard, and has a high breakdown voltage these are targeted characteristics for cost effective advancement. GaN and Si have significant thermal mismatch properties, which prevents easy integration. Development is being focused on enhancing the capability to incorporate GaN and Si for drop-in swap operation with Si. Increasing the availability of material sources and limiting defects in GaN technology are other aspects of the technology needing solutions. As GaN technology is invested in, many of these barriers are likely to fall to the wayside and will open doors for new RF possibilities.

Bright new technologies will always be driven forward with promises of enhancing existing RF industries and helping to create new ones. When considering device technologies, however, it is key that RF and microwave engineers look beneath the claims to find the option that is best for their application. Due diligence is required while weighing all of the specifications and parameters. As RF technology advances, it is likely that the present device technology will be supplanted by higher-performance, more cost-efficient or more reliable technologies. New device technologies shine in enabling advanced RF operation for fresh applications, which can best take advantage of these technologies’ unique benefits.

Download this article in .PDF format
This file type includes high resolution graphics and schematics when applicable.