MILITARY APPLICATIONS have consumed RF/microwave components for as long as electronic equipment has been deployed on the battlefield. Much has changed over the last 50 years but, not surprisingly, many of the design concepts from those earliest military electronics systems are still in use today.
In 1965, frequency-agile radar systems were considered "mature" technology, although the Department of Defense (DoD) was showing renewed interest in frequency agility for proposed new radar systems, as well as a method for updating existing systems. The technology was in favor at the time as a means of thwarting jammers. The basic concept was that if a radar occupied as much bandwidth as possible, the enemy would also have to occupy that bandwidth with a jamming signal to effectively block the operation of the radar. Even now, this poses an expensive and technically challenging proposition for an enemy intent on jamming a radar system. However, if the radar transmitter occupied only a small percentage of its total bandwidth, but moved the location of that occupied bandwidth in a relatively random sequence also known to the radar receiver, the enemy would still need to occupy the full bandwidth with a jammer signal to stop the radar. But the radar's bandwidth and power requirements would be much less than operating over the full frequency band simultaneously.
At that time, magnetrons were the transmit amplifier of choice, and companies were busy working on automatic-frequency- control (AFC) circuits that would allow their high-power electron tubes to operate in a frequency-agile system. One of the more heavily funded technologies in this area included master-oscillator/power-amplifier (MOPA) signal chains, which could be used in retrofitting older systems, as well as on new systems. Emerson Electric, for example, was developing a radar system that would combine frequency agility with moving-target-indication (MTI) capability. (Fig. 1).
Another important electron tube at that time, the traveling-wave tube (TWT), saw a great deal of research funding in the early 1970s due to the importance of the tube to electronic-countermeasures (ECM) applicationsnotably for jammers. Some of the same military requirements that are voiced in current-day design meetings were echoed back then: wider (greater than an octave) frequency coverage, higher output power, smaller size, and lower power consumption. Cost was also an issue, as TWT designers wrestled with using low-cost Alnico permanent-magnet structures for electron-beam focusing in the TWT, or else using samarium-cobalt magnets that cost about 15 times as much (thus raising the cost of the TWT by 15% to 20%) to achieve improved TWT performance. An example of the TWT technology at that time as an air-cooled S-band TWT produced by Microwave Associates (Fig. 2), capable of 400 W CW output power and 800 W pulsed output power.
While high-power tubes drove ground-based ECM and radar systems, Air Force researchers sought lighter-weight microwave circuits that could support airborne military electronics systems. In 1967, for example, companies were pushing the state of the art in microwave integrated circuits (MICs) in order to ship modules for Molecular Electronics for Radar Applications (MERA) use. At that time, Microwave Associates and Texas Instruments were both vying for a major contract to supply MIC modules for the MERA II airborne phased-array radar system. One of the S-band modules developed by Texas Instruments provided a good sample of the level of component/circuit integration possible at that time, with a power amplifier, multiplier, transmit/receive (T/R) switch, balanced mixer, and modulator within the same compact housing (Fig. 3).
Advances in MICs were slow but steady as designers learned to use new circuit materials and processes to advantage. In 1977, for example, Bell Laboratories unveiled a ceramic material, Ba2 Ti9 O20 with extremely high quality factor (Q) of as much as 11,000 at 4 GHz, which would make it attractive for use in filters and oscillators. Making the material even more appealing for high-frequency designers was a dielectric constant of 40 at 4 GHz, along with a temperature coefficient of 2 ppm/C. In comparing filters formed with dielectric resonators based on the material, a 4-GHz six-section waveguide filter measured 5.75 inches long, in contrast to a waveguide filter based on copper, the standard waveguide material of that time (Fig. 4).
That same year, researchers at TRW Semiconductors responded to the military need for higher output power from solid-state devices by developing a 16-cell silicon bipolar transistor reaching a new record (at that time) of 40 W CW output power at 2 GHz. The 16 cells were bonded in parallel and mounted on a beryllium oxide (BeO) substrate to achieve the new solid-state power record (Fig. 5). The 16 cells were actually individual transistors that had been in production for some time, but never used in an application such as this. The devices featured an interdigitated design with 0.07-mil fingers and 0.1-mil spaces.
The idea of combining solid-state devices to achieve higher output levels had been put to work several years earlier at Hughes Research Labs when an X-band power combiner was assembled with 32 Impatt diodes to provide 23.4 W of CW output power at 9.3 GHz (Fig. 6). The solid-state source integrated commercially available Impatt diodes nominally rated at 500 mW output power per diode. The diodes were each mounted in a coaxial module, magnetically coupled to a transverse magnetic (TM) TM020 microwave cavity. A filter was also incorporated in the cavity to suppress unwanted modes. The packaging density of the component was limited only by the physical size of the packaged diodes. Used as an amplifier, it offered 12.5 W output power with 3-dB gain over a 1-dB bandwidth of 125 MHz.
Also at that time, authors from Raytheon Co. projected the growing role of microprocessors in tactical ECM and other military electronics systems. Because of the increasing number of emitters, "smart" electronic systems would be mandatory in military electronics systems architectures, and the authors offered a glimpse of the future by explaining a six-microprocessor ECM module that could track data, manage resources, and analyze received pulse information.
One of the potential applications for the microprocessor module was in the control of agile systems employing multiple voltage-controlled oscillators (VCOs). The microprocessors were developed to control the tuning sequences of the oscillators, calculate their settling times, and handle requests such as overlapping frequency ranges for the multiple VCOs in an ECM system. Obviously, the role of the microprocessor has continued to grow steadily over the years.