GaN: The Technology of Choice for Next Generation Advanced Radar Systems
By Integra Technologies, Inc.
Radar, an acronym for RAdio Detection and Ranging1, is a method for detecting the presence of distant objects called targets. The targets could be defined in military terms such as aircraft, ships, vehicles and missiles, or in commercial applications such as precipitation or terrain. A radar system generates energy in the form of high frequency radio waves which is directed in a narrow beam toward the target with directional antennas. Energy reaching the surface of a target will be scattered and the amount of energy reflected back to the receiver of the radar system can be analyzed. The scattered radar signals that are reflected back towards the transmitter are the desirable ones that make radar work. Analyzing the reflected radio waves reveals the characteristics of the target such as the distance, direction and speed. The time between the transmission and reception of the energy and the direction that energy is received determine the target’s position in relation to the radar (radiolocation1) and the shift in frequency determines the velocity (Doppler Effect1.) As certain materials absorb or reflect different amounts of energy, it is possible through signal processing techniques to determine the material composition of the target as well. This reflected energy can be converted into images on a scope or projected onto a map.
The diversity of radar applications today is staggering. The military uses include surveillance, early warning systems, aircraft identification and tracking, marine navigation, collision avoidance and fire control, to name a few. Commercially, radars are used for air traffic control and scheduling, automotive collision avoidance, automotive automatic parking systems, astronomy, vehicle speed monitoring, weather formations including clouds and precipitation, foliage-penetration for surveillance, and ground-penetration for geological observations and mapping.
Radar systems come in a variety of configurations in the emitter, the receiver, the antenna, frequency and even the strategy of scanning the area.
Radar systems generally work by connecting an antenna to a powerful radio transmitter to emit a high frequency signal. The transmitter is then disconnected and the antenna is connected to a sensitive receiver. The receiver utilizes a LNA which amplifies the faint echos returned from the target objects and sends these signals to a display. In order to scan an area effectively, the radar antenna must be physically moved to point in different directions. Typically an antenna is physically rotated in a circle to scan in all directions. However, the physical rotation of the antenna is a physical limitation and a costly component as well.
A method was developed where the beam rapidly can be manipulated from one direction to another without mechanical movement of the large antenna structure. The signal source is split into multiple paths, some of which are electrically delayed, and sent to separate individual antennas typically spaced in an array. Since the delays could be easily controlled electronically, allowing the beam to be steered very quickly without moving the antenna, the system is called Electrically Scanned Array (ESA). Each antenna acts as its own individual radar. Each antenna element emits radio waves that will combine in an interference pattern at different points in front of the antenna. The interference patterns between the individual signals can be controlled to reinforce the signal in certain directions and mute it in all others. Steering can also be accomplished by placing a phase shifter behind each radiating element in the array. The beams are formed by shifting the phase of the individual signals to produce an interference pattern. This system is called Phased Array Radar (PAR). In a phase array configuration, there is no physical movement of the antenna needed to focus the radar waves on the target as all the elements of the array are fixed in position and the beam is steered electronically.
While the total output power of a phased array radar system may be in the many KW range since there are potentially hundreds of elements in the array, it is desirable to have relatively small output power devices. In PAR systems with so many elements it is necessary to keep the physical size of the transistor small. A typical transistor is pre-matched inside of the package and has an impedance of a few ohms at the transistor leads. The remaining matching of the transistor to a 50 ohm system requires additional passive matching components on a dielectric pcb. The input and output matching networks that are required are repeated for each element in the array. Integra Technologies Inc. has developed the MPAG2735M30 (shown in Figure 1) specifically for phased array systems with the size of the device limited to the size of the package. This device is a self-contained Miniature Power Amplifier (MPA) where the entire matching circuit is housed within the package. Thus, the device impedance at the leads is already 50 ohms and perfectly suited to implement in a 50 ohm system with no additional matching required. This approach is ideal for phased array applications as the size of each element is now simply the size of the package, which results in smaller and cheaper system implementation. Figure 1 is an example of a 30W output power GaN/SiC HEMP transistor. This device is characterized under pulsed signal conditions with a 300us pulse width and 10% duty cycle and produces 11 dB of power gain and 50% efficiency at rated output power. The transistor operates in the S-band from 2.7 – 3.5GHz and is ideal for phased array radar applications.
Modulation – CW vs Pulse
In a radar system, the emitted signal comes in a variety2 of forms, including pulsed, unmodulated continuous wave (CW) and modulated CW. The types of pulsed radar systems include the simple pulse radar, pulse Doppler radar, high range resolution, and pulse compression based systems. One of the most typical radars is the simple pulsed based system. This system consists of a pulse of certain duration that is repeated at regular intervals. A typical pulsed based system transmits pulses at a particular repetition rate. The time interval between transmission and reception of the pulses is used to determine the target’s range.
When the radar utilizes the frequency shift of the echos of the pulses it is called Pulse Doppler radar. Pulse-Doppler is a 4D radar system capable of detecting a target’s 3D location (position) and its velocity. The radar transmits short pulses of radio frequency waves which are partially bounced back by the targets. In a typical system, the received energy from a dozen or more pulses are combined using digital signal processing (DSP) based on the Doppler effect to resolve the key characteristics. The Doppler effect is utilized to reject unwanted noise and clutter in the signal. Doppler radar types use a very high pulse repetition rate for the measurement of velocity. Pulsed systems with a lower pulse repetition rate have a more accurate range measurement. A trade-off exists between the pulse repetition rate and the accuracy of measurements such as velocity and range. A firm range measurement is obtained by transmitting multiple waveforms with different PRRs and resolving the final measurement.
When very short pulses are used, a very accurate range measurement is obtained in high-range resolution radar. Pulse compression radar is similar to high range resolution radar, but uses longer pulses for greater range. The long pulses are modulated in frequency or phase which is compressed in the receiver otherwise overwhelming the receiver with too much energy which can potentially damage the receiving equipment.
Pulsed radars can be quite complex and costly as they require many components such as the pulse generator, phase/frequency modulator, and circuitry to control the pulse duration. Usually requiring high power transmitters when built for maximum range, they are generally very powerful, with output power ratings in the 10s or 100s of KW. Applications for pulse radar systems are generally used when it is necessary to detect targets in a certain area as well as determine range, bearing and velocity. If only one of these characteristics is needed, there are alternative systems that will meet this requirement.
For other applications the level of complexity and power of a pulsed based radar system is unnecessary. A simpler system is to transmit a known fixed stable frequency as a continuous wave (CW). The advantages of this system are that the energy source is not pulsed and is easier to generate, which makes it much simpler to manufacture and operate. The total power on target is often greater than pulses as the transmitter is broadcasting continuously. CW radar systems are generally simpler, less costly, and more compact than pulsed radar systems.
Many radar applications, however, do not require all the information normally supplied by pulse radar systems. Extremely simple applications may require only that the radar indicate detection (the presence of a reflecting target object at any range), velocity (instantaneous change of rate of speed), or range (distance to the target object). For many such simple applications, a type of CW radar is often used. CW radar transmits and receives at the same time at a constant frequency and amplitude. Doppler is used to resolve velocity, but this system suffers from poor range measurement capabilities. Since the radar transmits continuously there is no basis for the time delay between pulses (there are no pulses) and therefore distances can not be measured directly. The transmitter is the strongest source of RF energy in the vicinity of the receiver unless there is a large physical barrier along the line of sight, such as a mountain. The CW radar must depend on the Doppler shift3 of the received signals to differentiate between the echos and the strong transmit signal which can overload the receiving detection equipment. While the system is fundamentally incapable of measuring range, the system is very capable of measuring instanteous rate of change or velocity of the target. Therefore these unmodulated CW systems are generally used for speed gauges. The advantage of the unmodulated CW radar is the ability to handle targets of nearly any velocity. These types of radar equipment have been specialized for speed measurements. Some of our readers may be familiar with these types of low cost radars as these are most often utilized as speed gauges by police departments!
Unmodulated CW radar, unlike a pulsed radar, cannot determine target range based on the time delay between transmit and reception of a radar pulse reflected from the target. By imposing a frequency modulation (FM) on the CW carrier however, a timing mark can be created that will provide a reference from which target range can be derived. A more complicated system has been developed to detect the range in a CW configuration by modulating the CW signal4. A frequency modulated CW radar is one in which a ranging capability is provided by frequency modulation of the transmitted signal. In the modulated CW radar, the transmitted signal is constant in amplitude but modulated in frequency. The most common type is the frequency modulated CW system (FM-CW). Frequency-modulated (FM) continuous wave radars are capable of determining distance directly. If the frequency of the CW radar is changed with time, then the reflected signal will be a different frequency which is better distinguished from the strong CW signal at the transmitter. In this system, the signal is not a continuous fixed frequency, but varies up and down over a fixed period of time. Range is determined by evaluating the frequency spread of the received signal. The difference in the frequency is proportional to the distance to the target and can be used to determine the range of the target object. It is also possible to use a CW radar system to measure range instead of velocity by frequency modulation, the varying the frequency of the transmitted signal systematically. The return signal is measured and the time delay between transmission and reception will determine the range. Of course, the amount of frequency modulation must be significantly greater than the expected Doppler shift or the results will be affected.
Advanced Radar Systems
As advances continue in the signal processing arena, making faster and cheaper DSP units, more and more post processing can be done on the received radar signals. A type of radar requiring post processing using DSPs is the Synthetic-Aperture Radar (SAR5). SAR is usually implemented by mounting a single beam-forming antenna on a moving platform where the target is repeatedly illuminated with pulses. Many echo signals arrive at the receiver although the antenna has been moving. The signals are coherently detected and post processed together and resolved into an image of very high precision. The relative motion of the antenna platform provides distinctive coherent signal variations that obtain much finer spatial resolutions than conventional beam scanning methods. The fine resolution capabilities make this radar useful for commercial mapping applications and military surveillance applications.
The Power of Radar Systems
One of the key specifications of a radar system is the output power, which can be a few hundred watts to over a hundred kilowatts. More power on target is always desirable. The higher the power level, the greater the distance the high frequency radio waves will travel. For the military, more power equals more range where the target can be identified at a greater distance. For commercial applications, more power can mean different things. The higher power level can increase the signal to noise ratio, leading to a higher resolution system and more accurate data received.
For all systems, the RF output power is converted from a DC power source. The efficiency of a system is the amount of DC power converted to useful RF power compared to that power which is wasted or consumed. Power consumption is critical for several reasons: in commercial systems, wasted power is money wasted; in military systems, the extra power that is wasted generates heat and leads to premature failure; in portable systems, wasted power lessens the battery life; in fixed systems, the heat generated by the excess power is handled with expensive and large cooling systems. For this variety of reasons, efficiency is a key factor in high power radar designs.
The Importance of Frequency
The performance of radar systems is affected by the frequency of operation. The radio waves travel through the atmosphere, which has different rates of absorption at different frequencies. Therefore key performance metrics like range and resolution are dependent on the amount of received reflected power, which is influenced by the frequency. Certain frequency bands specific to radar systems are specified by the ITU, International Telecommunication Union, and are recognized internationally, as shown in Table 1.
Widely available silicon devices are suitable for the UHF and L-band and the lower end of the S-band. For the C-band and X-band, non-silicon technologies are required. One of the features of wide bandgap compound semiconductor technology like GaN is the high electron mobility produces devices that operate at higher frequencies. One application of the C-band radar system is for meteorological uses to monitor precipitation and weather forecasting. A pulse-Doppler system is employed due to its high detection capability in high-clutter environments.
An example of a C-band pulsed GaN/SiC HEMT transistor is shown in Figure 2.
Figure 2 is an example of a 25W output power GaN/SiC HEMP transistor. This device is characterized under pulsed signal conditions with a 300us pulse width and 10% duty cycle and produces 13 dB of power gain and 50% efficiency at rated output power. The transistor operates in the C-band from 5.2 – 5.9GHz and is ideal for weather radar applications.
Radar Evolution and Current Market Drivers
The Role Technology Plays
With the wide variety of radar system types at varying power levels and frequencies, the choice of technology plays a large role in meeting the requirements at a reasonable cost. RF power transistors fabricated with silicon bipolar are ideal for many radar systems. The bipolar technology has very high power density, especially when pulsed. The high power density leads to more power in a small package and keeps costs down. BJTs have a simple DC biasing scheme, additionally reducing complexity and cost. The drawbacks to silicon bipolar junction transistors are the relative low gain (typically less than 10dB) and inherently non-linear response.
Figure 3 is an example of a 150W RF output power silicon bipolar transistor. The device is characterized under pulsed signal conditions with a 330us pulse width and 10% duty cycle and operates in Class C mode with 50V bias supply voltage. The transistor operates in the L-band from 1.2 – 1.4GHz and is ideal for ATC radar applications.
Another silicon technology readily being used is LDMOS. The LDMOS layout is spread out, offering a lower W/mm solution that is useful for longer pulse or CW signals. Silicon LDMOS devices have high gain, high linearity and in many cases, higher efficiency than their bipolar counterparts. The high linearity and CW capabilities enable LDMOS to be used in many advanced radar system architectures that bipolar can not satisfy. However, using silicon for both BJT and LDMOS limits the frequency of operations to the UHF, L and S-bands. For higher band radar a new technology is required.
Figure 4 is an example of a 150W RF output power silicon LDMOS transistor. The device is characterized under pulsed signal conditions with a 300us pulse width and 10% duty cycle operating with 32V drain voltage and Class AB mode with 40mA IDQ. The transistor operates in the S-band from 2.7 – 3.1GHz and is ideal for ATC radar applications.
For frequency operation above 4GHz, a wide bandgap compound semiconductor technology7 is necessary, such as GaAs, GaN8 or SiC. We will focus on GaN with a SiC substrate for improved thermal performance as a reasonable compromise between performance and price of these technologies. GaN on SiC has very high performance metrics compared to silicon technology9. GaN/SiC has a high power density similar to bipolar and exhibits higher gain and efficiency than LDMOS. The high power density produces devices with lower parasitic capacitance per mm of die periphery. The low capacitance translates to higher impedances on the RF load line, enabling higher performance at a greater bandwidth of frequency compared to silicon technologies.
Figure 5 is an example of a 150W RF output power GaN/SiC HEMT transistor. The device is characterized under pulsed signal conditions with a 300us pulse width and 10% duty cycle and operates in Class AB mode with 120mA of IDQ at 36V supply bias voltage. The transistor operates in the C-band from 5.2 – 5.9GHz and is ideal for weather radar applications.
The high power density also increases the system efficiency. GaN on SiC technology produces higher power devices and uses smaller die than silicon technologies which require fewer devices to combine in a system. Combining fewer devices reduces combining losses which decreases wasted power and translates into a higher system efficiency. The smaller form factor also keeps the system to a reasonable size introducing a cost savings.
Why Integra GaN/SiC for Radar Systems?
Wide bandgap semiconductor material, such as GaN, exhibits a very high power density compared to silicon. This feature is electrically desired, but mechanically special care must be taken to distribute the heat generated in such a small area as heat is one of the major contributors to early failures in solid state technologies. Therefore, the GaN High Electron Mobility Transistor is typically epitaxially grown on an insulated material such as SiC. The improved thermal conductivity of the SiC substrate material allows the safe operation of GaN/SiC devices up to 225C10 junction temperature compared to silicon technology which is safely rated at 200C. The ability to operate a transistor at a higher operating temperature without negative effects on reliability make wide bandgap GaN/SiC a superior choice compared to silicon technologies.
GaN also exhibits an extremely high breakdown voltage compared to silicon technologies. This is due to the high field strengths supported by the material, as noted by Weitzel9. Voltage breakdowns of >300V have been reported for GaN/SiC HEMT RF power devices11. For linear operation in classic modes of operation like Class A and AB, the typical breakdown voltage must be greater than twice the operating voltage to account for the voltage swings during normal operation and prevent premature failure due to breakdown. The technology is ideal for the high efficiency circuit techniques, like switching modes of operation of Sokal12 and popularized by Raab13-15. These switching modes of operation place a high demand on breakdown voltage requirements. These advanced circuit techniques to improve efficiency require breakdown voltages that are multiples of the operating bias voltage. The class E and F operation is normally narrowband with silicon, but with the lower capacitance per mm in GaN/SiC is ideal for this mode of operation.
Conclusion: The Future Is Now
GaN/SiC has a high electron mobility allowing high frequency operation. The high power density leads to small, high power parts that cover a wide bandwidth of frequencies. The smaller form factor uses fewer devices per system and has less combining losses at the system level. The devices exhibit a higher efficiency per device and the high voltage breakdown allows the use of high efficiency schemes, improving the efficiency performance even more. All of these traits enable GaN/SiC RF power devices to exhibit high power, efficiency and impedance up to 50 ohms which reduce the complexity in design advanced radar systems like Phased Array Radars. The devices exhibit a high reliability with the high operating temperature performance and the ability to sustain high voltage fields throughout the material. The high efficiency performance allows fewer devices per system, creates less heat which lowers the operating temperature, reduces heatsink requirements and consumes less DC power, saving on operating costs. In effect, GaN/SiC is a green technology which should be explored more for future systems.
The superior performance of the GaN/SiC technology over silicon is apparent for many applications; however, the underlying cost of fabricating on SiC substrate material is the major limitation to implementing these systems across all systems at all frequencies. Thus, the bipolar and LDMOS solutions will serve markets where the requirements are satisfied with a silicon technology.
2 Camacho, Joseph P., 20-Year Federal Spectrum Requirements Forecast for Radar Bands, US Dep’t of Commerce, NTIA Special Publication 00-40, pp. 1- 6, May 2000.
3 Collins, John D., Continuous Wave Radar with Ranging Capability, US Patent 4,620,172, Oct. 28th 1986.
4 Wise, John. C., “A Perspective on EW Receiver Design,” Technical Report APA-TR-2009-1102, November 2009.
5 Chan, Y.K., Koo, V.C., “An Introduction to Synthetic Aperture Radar (SAR),” Progress in Electromagnetics Research B, Vol. 2, pp. 27-60, 2008.
6 Camacho, Joseph P., 20-Year Federal Spectrum Requirements Forecast for Radar Bands, US Dep’t of Commerce, NTIA Special Publication 00-40, page 6, May 2000.
7 P. Parikh and J. Palmour, “Gallium Nitride Wide Bandgap Devices – Technology, Potential, and Markets,” Compound Semiconductor Outlook 2002 – USA, Gorham Conference, Feb. 26, 2002.
8 Mishra, U.K., et al, “AlGaN/GaN HEMPs – an overview of device operation and applications,” IEEE Proceedings, Vol. 90, Issue 6, pp. 1022-1031, June 2002.
9 C.E. Weitzel, “RF Power Devices for Wireless Communications,” IEEE MTT-s Digest, Volume 1, 2-7 June 2002 Pages: 285-288.
10 Milligan, et al, “SiC and GaN Wide Bandgap Device Technology Overview,” IEEE Radar Conference, pp. 960-964, 2007.
11 Cree Inc., www.cree.com
12 N. O. Sokal and A. D. Sokal, “Class E—A new class of high efficiency tuned single-ended switching power amplifiers,” IEEE J. Solid-State Circuits, vol. SSC-10, pp. 168–176, June 1975.
13 F. Raab, Class-E, Class-C, and Class-F Power Amplifiers Based Upon a Finite Number of Harmonics, IEEE Transactions on Microwave Theory and Techniques, Vol. 49, NO. 8, August 2001 pp. 1462-1468
14 F. Raab, “Idealized operation of the class E tuned power amplifier,” IEEE Transactions on Circuits and Systems,NO. 24, pp. 725-735 (1977).
15 F. H. Raab, “Maximum efficiency and output of class-F power amplifiers,” IEEE Trans. Microwave Theory Tech., vol. 47, pp. 1162–1166, June 2001.
Integra Technologies, Inc.
this article to a friend!