Microwave Circulators Evolve to Meet True Surface Mount Requirements of Advanced Systems
If it’s solving design challenges of directivity and throughput in 5G communications applications or providing enhanced anti-jamming and tracking agility, phased array antennas, or actively electronically steered arrays (AESAs), are essential to meeting the application demands of these emerging technologies. Phased array antennas and AESAs are now deployed in ground-based, aerospace, and naval military applications, as well as commercial point-to-point/point-to-multipoint (PTP/PTMP) radio and satellite communications (SATCOM) applications. A critical component to designing compact and high-performance transmit-receive (TR) modules for the latest antenna arrays are circulators and isolators. For extremely low-cost consumer applications, such as a wireless router, quasi-stripline circulators may be adequate for some vendors. However, when performance and high reliability (Hi-Rel) features are necessary, ferrite-based circulators and isolators are still a necessity. Due to recent advances with developing true surface mount technology ferrite-based circulators and isolators, there is now no longer a compromise between circulator performance and size, weight, power, and cost (SWAP-C).
The ability of magnetic materials to manipulate the propagation of electromagnetic energy has been known for over a century. Moreover, ferrite-based RF circulators, an application of ferromagnetic resonance, have been in use for over half a century. Aside from the development of various packaging technologies, however, there have not been substantial changes to how ferrite-based microwave circulators are designed and assembled. Notwithstanding, there have been attempts to produce surface mount technology (SMT) ferrite-based microwave circulators. Though innovative, the prior attempts at achieving industrial, let alone military and SATCOM grade, SMT circulators have not been able to leverage the full advantages of highly reliable and precise autonomous surface mount assembly and quality control—autonomous tape and reel solder reflow assembly.
For many decades, market demands have allowed for purely waveguide, coaxial, or drop-in stripline circulator assemblies. Emerging markets and recent technological innovations have changed the landscape of needs drastically, and ferrite-based circulator technology has been slow to meet these demands. Amongst a wide group of small size, lightweight, power efficient, and low cost (SWAP-C) RF and microwave microelectronics, ferrite-based circulators have been a clunky necessity, sometimes dominating valuable PCB real estate and driving up device costs. This is especially true for phased array antennas and actively electronically steered arrays (AESAs) used in commercial, aerospace, military, and SATCOM applications, which often require transmit-receive modules numbering in the hundreds and thousands.
In order to avoid doubling the number of antennas in phased arrays, ferrite-based microwave circulators and isolators are used to enable duplexing of the antenna port. Essentially, the magnetic properties of a circulator enable non-reciprocal transmission of signals over a certain frequency range. Typical circulators for RF and microwave applications are 3-port devices, where port one allows transmission to port two, port two allows transmission to port three, and port three allows transmission to port one. Additionally, some applications require complete isolation, or unidirectional transmission, from one port to another, and the third port of a circulator is terminated with a matched impedance to achieve this.
Practical Limitations and Considerations of Ferrite-based Microwave Circulators
The frequency behavior of a circulator, namely the bandwidth, is dependent upon the size of the magnetic material, “Y-junction” stripline, and impedance matching network used in the design. The size of the stripline is a function of the frequency, which leads to decreased sizes for circulators at higher frequencies. Certain design aspects also affect the relative size of the matching network, which could contribute substantially to the overall footprint of a circulator. Depending on the type of ferrite and assembly techniques used, the height of a circulator assembly may also be a function of the ferrite material, and can almost be similar to the width and length dimension of the component.
A common design approach for a RF circulator is to sandwich a stripline transmission line, “Y-junction,” between two ferrite discs. The ferrites are cut, lapped, and polished specifically to enable the intimate contact between assembled components which support low loss, counterrotating waves that form a nonreciprocal electromagnetic transmission pattern. In order to achieve the desired high isolation, low-insertion loss, low VSWR, reasonable power handling, and wide bandwidth, the geometry of the ferrites and transmission lines suffer constraints that usually end with design trade-offs and large circulator devices.
For X-band (8 GHz to 12 GHz) and higher frequency applications, the size of a circulator is relatively small, though still large compared to other microelectronic devices in SMT packages. For C-band, L-band, S-band, or commercial cellphone and ISM-bands—between 600 MHz and 6 GHz—the size of a circulator assembly is many times the size of other RF components. Depending upon the application, design, and ferrite-based circulator packaging, the circulator itself can be up to half the footprint of the PCB, and be the single most expensive component.
The challenges with developing true SMT ferrites stem largely from the intrinsic physics of the device and the practical requirements of assembling relatively cumbersome ferrite parts. As the magnetic field’s influence on the transmission path of an RF signal is what enables the unique benefits of a circulator, preventing external magnetic flux from increasing or decreasing the established magnetic field is also important. To achieve this, external ferromagnetic shielding may be required, along with the additional size and weight that iron, steel, and other ferromagnetic shielding materials bring.
Ferrite-based circulators are constructed of fragile materials and are prone to performance variations based on their fabrication process. Moreover, the size and geometries of ferrite-based circulators also mean that the component is potentially subject to a substantial portion of the shock and vibration that an assembly may experience during use. Considering all of these factors is necessary when designing in an RF circulator, which also contributes to additional system-level design overhead.
Types of Microwave Circulator and Isolator Packaging
Not including novel true SMT microwave circulators, there are several typical fabrication and packaging approaches for circulators. One approach places a ferromagnetic material inside a waveguide structure that is either terminated in waveguide flanges or with waveguide-to-coaxial conversions for coaxial connector ports. Another is the assembly of a metallic housing around a specially crafted and coaxial connected circulator structure. Though these options are capable of handling high powers and enabling many high-performance features, they are not applicable to PCB designs, automated fabrication, automated assembly, or capable of meeting the SWAP-C requirements of the latest applications.
More compact and flexible connection circulator fabrication and assembly options include drop-in, microstrip, and surface mounted (SMD). Drop-in circulators are assembled in a compact package that leverages through-hole assembly technology. Though able to achieve good performance metrics and handle mid-range power transmissions, drop-in circulators are still handcrafted and installed manually by a trained assembly technician. Also, drop-in circulator models require a cut-out section in a PCB and tapped holes for screw-mounting the circulator housing. Once this is done, the circulator leads must be bent and soldered in such a way that the heat applied to the circulator leads doesn’t damage or demagnetize the circulator’s ferrite.
Microstrip or stripline based circulators are typically more compact circulator designs that may even have wrap-around traces that can be soldered to a PCB. However, stripline circulators scale significantly with size and still require specialized fabrication and assembly. Lastly, SMD circulators are very compact designed microstrip circulators that are designed with wraparound traces that resemble surface mount package sizes and footprints. These circulators have previously been the most compact, but still aren’t capable of being designed into standard surface mount packages, and can’t be automatically assembled.
Some of the limitations that prevent microstrip and SMD packaged RF circulators from being true surface mount technologies involve the material properties of the circulator manufacture, the unique device geometries, the high variations from part to part, and the cost of producing these circulators in large numbers. For example, many microstrip and SMD circulators rely on adhesives to maintain close contact between the ferrites and microstrip segments. These adhesives typically can’t withstand the high temperatures of reflow solder, even if the ferrite material wouldn’t be harmed in the process.
A cut-out section is also often necessary to accommodate the height of a SMD circulator, which is also not conducive to automated assembly or reflow soldering. Furthermore, the geometries and part-to-part variations prevent true tape-and-reel automated pick-and-place, and these devices tend to struggle with yields and rework costs. To meet datasheet performance, some microstrip and SMD devices may also require ribbon bonding to the substrate, which increases costs, process development, and turnaround time on product delivery.
What is the Difference Between Surface Mounted Devices (SMD) and True Surface Mount Technology (SMT)
Though some of the fabrication and assembly challenges with SMD circulators were mentioned previously, there are many other factors that need to be in place to enable the full set of SMT assembly benefits. Also, the market forces that are opening the door for true SMT circulators are only likely to compound in the future, which indicates that it will be increasingly worthwhile to have capable SMT circulators available to adapt to these market trends. Of these trends, the growing integration of digital electronics and RF electronics on the same platform and the demand for more highly integrated RF circuits—either RF integrated circuits (RFICs) or monolithic microwave integrated circuits (MMICs)—play a large role in leveling cost pressures on microwave circulators.
If a complete and high performing microwave design, based on highly integrated components on an industry standard substrate, can be built, then applications that require huge numbers of modules can be built with mass manufacturing techniques instead of handcrafted custom fabrication. Overall, costs and component traceability and reliability would be much greater than with current designs. As most other RF and microwave devices and components either have IC versions or can be built using metal traces on PCB, microwave circulators are often a limiting factor in low cost, no compromise manufacture of RF modules.
Often not brought to light, the qualification/certification testing and early design and simulation stages of product development are also benefited by this approach. In the case of product design and characterization, a highly reliable automated manufacturing and packaging process eliminates many of the process variables and piece work that are currently necessary to build non-SMT circulators. Beyond lowering costs, this also enables quality control analysis and the evolution of the process to enhance yields, reduce manufacturing times, and decentralize manufacturing expertise. Certification and quality assurance testing could also be more easily automated, which would greatly benefit meeting the latest FCC, industrial, military, and space qualification standards. Also, smaller package sizes leading to reduced lead lengths could result in better electromagnetic compatibility and electromagnetic interference (EMC/EMI) performance, as smaller lead lengths could result in less radiated and conducted emissions related to the circulator component.
Possibly as important, a more reliable process would allow for model development and the use of sophisticated modeling tools that also automate many design tasks. Either with RF circuit design simulation tools or 3D EM simulation tools, a consistent manufacturing and packaging process could open the door to faster design cycles, as the device models would be more reliable. As circulators are often part of both the transmit and receive signal chains, the ability to accurately predict circulator behavior could prevent troubleshooting challenges and redesign in larger assemblies reliant on circulator performance. These benefits could lead to faster time to market and reduce the need for redesign cycles.
Circumventing Circulator PCB Design Challenges
Enabling true SMT compatibility—namely compatibility with automated pick-and-place machines, tape-and-reel loader systems, and solder reflow—requires a circulator to be designed in such a way that the device can be reliably installed in a standard SMT package. Furthermore, the circulator SMT package traces would have to be flush with the lead when automatically placed, and the circulator would have to be able to withstand standard solder reflow temperature profiles that far exceed operating temperature ranges.
Though the details are proprietary, TRAK, a brand of Smith’s Interconnect, was able to solve the assembly challenge of developing a circulator technology that can be installed within a standard 0.25” square base and 0.66” high package. This X-band SMT circulator presents a 10% bandwidth (from 8.5 GHz to 9.5 GHz) with insertion loss at 0.45 dB maximum, return loss at a 20 dB minimum, isolation better than -20 dB, and power handling up to 20 Watts from -40C to +85C. One of the critical design challenges was designing a fabrication process that can withstand the temperature profile of reflow soldering without changing the electric performance.
Addressing another challenge, the development required the ability to rapidly optimize the design using full EM simulation, Ansys HFSS. The very tight tolerances and precision needed to adequately simulate a circulator of these dimensions would typically require days of calculations with a standard computer simulation. Hence, TRAK invested in a high-performance computing cluster that enabled much faster turnaround on simulation results, and the fine tuning of the circulator’s device performance was able to be brought to industry standards.
Though the applications for true SMT microwave circulator technology can span aerospace, SATCOM, wireless communications, and potentially more, the root driver for the development of this technology was to reduce the costs and footprint size of phased array radar TR modules. The constraints on the technology for cost and size reduction are stringent, as circulator devices are competing with standard microwave switches. These switches have much poorer insertion loss and don’t allow for duplex communication, however, the cost and size of previous microwave circulators is often prohibitive to use in phased array systems with hundreds and thousands of elements. These needs steered the development of the industry’s first true SMT circulators, and will likely continue to push the boundaries of integration and performance in shrinking package sizes.
Further developments with this SMT circulator technology platform are planned to include a 0.35” square package X-band circulator with 30% bandwidth, three times the bandwidth of the currently available technology. Additionally, development of an S-band model in a 0.75” square package and 30% bandwidth is also planned.