Selection of PCB Materials for 5G - FEBRUARY 2018 - Rogers Corporation
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
E-BOOK Selection of PCB Materials for 5G FEBRUARY 2018 S P O N S O R E D B Y 1 MICROWAVE JOURNAL n JUNE 2011
Table of Contents 3 Introduction Pat Hindle Microwave Journal, Editor ENHANCED MOBILE BROADBAND 4 MASSIVE MACHINE ULTRA-RELIABLE LOW TYPE COMMUNICATION LATENCY COMMUNICATION 5G: Is it Ready for Take Off? James Kimery National Instruments, Austin, Texas RO3003 RO4350B RO4003C PTFE/WG Dk 3.2 PTFE/WG/Ceramic FR-4 1.04 9 1.03 Change from 25°C 1.02 Opportunities for High Frequency 1.01 1.00 0.99 Materials in 5G and the IoT 0.98 0.97 –50 –25 0 25 50 75 100 125 150 Temperature (°C) Art Aguayo Rogers Corp., Chandler, Ariz. 13 Paving the Way for 5G Wireless Networks The ROG Blog is contributed by John Coonrod and various other experts from Rogers Corporation, providing technical advice and information about RF/microwave materials. 15 Transformation to 5G: PCB Advantage Tony Mattingly Rogers Corp., Chandler, Ariz. 4 Mil LCP Laminate Effective Dielectric Constant vs. Frequency for Various Copper Foil Types on 50 Ω Microstrip Transmission Line 2.8 ED-3.0 µm RMS ED-1.5 µm RMS Effective Dielectric Constant RT-0.7 µm RMS RT-0.5 µm RMS 2.7 18 2.6 2.5 Selecting Circuit Material for the 2.4 2.3 Different Spectra of 5G Power Amplifiers 2.2 0 10 20 30 40 50 60 Frequency (GHz) John Coonrod Rogers Corp. 2
Introduction PCB Materials Considerations for 5G With the first 5G NR standard recently approved by the 3GPP at the end of 2017, many companies are racing to design 5G radio products that will demand wider bandwidths, higher frequencies, enhanced carrier aggregation and support of massive MIMO. AT&T and Samsung plan to launch 5G mobile services and Verizon plans to launch 5G Fixed Wireless Access in the US this year while South Korea will be demonstrating 5G at the upcoming Winter Olympics. The demanding performance requirements of 5G will push the limit of PCB designs from antennas to control functions to amplifier circuits. PCB effects such as copper surface roughness, Dk variations, thermal dissipation, passive intermodulation, coefficient of thermal expansion and thickness variations, will affect 5G designs more than previous generations that had less stringent performance criteria. PCB designers will have to consider many of these effects in their designs so need to learn about the various tradeoffs with different types of PCB materials and processes. 5G mmWave applications will have very different materials needs than sub 6 GHz applications so these articles also address the requirements for higher frequency designs. At mmWave frequencies, the effects of small variations in material parameters will have even more affect on the circuit’s performance. For higher frequencies into the mmWave range, designers typically need to minimize Dk, copper roughness and thickness variations as they can all negatively affect performance. These considerations must be incorporated into the design using accurate models and simulations in order to avoid costly iterations. This eBook introduces 5G objectives and goals, opportunities for high frequency materials in 5G and IoT applications, materials effects for 5G designs, PCB antenna design considerations, and material selection for the different spectrum of 5G power amplifiers. The effects of various material parameters on loss, thermal dissipation and reliability are discussed in these articles to provide designers with the tools needed to account for these affects. Microwave Journal has put together this collection of articles covering these topics. Rogers Corporation has contributed many of the articles as a leading PCB material manufacturer offering their expertise and experience in this field to educate designers and manufacturers about various PCB material considerations in 5G designs. Pat Hindle, Microwave Journal Editor 3
5G: Is it Ready for Take Off? James Kimery National Instruments, Austin, Texas N ever has there been so much hype the 3GPP also set key performance metrics and attention paid to a new wire- for ultra-reliable low latency communica- less standard than with 5G. 5G has tions (URLLC) and increased connectivity, generated a lot of interest, due to setting the stage for billions of connected its potential transformational impact on both devices for massive machine-type communi- consumers and businesses across the globe. cation (mMTC). Has the hype been overdone? Let’s look at Today’s wireless standards do not and where we have come from, where we are to- never have addressed latency and broad day and speculate a bit as to what the future connectivity. Latency is quite important be- may hold. cause lower latency will not only improve the common data experience for users but HOW WE GOT HERE create new applications that rely on fast A little over 10 years ago, Apple intro- network response. Low latency and, specifi- duced the iPhone and opened our eyes to cally, deterministic low latency responsive- the potential of smart devices combined ness provide the foundation for “control” with wireless broadband data. In 2016, Cisco applications over the network. Combining published their Global Mobile Traffic Fore- robots, drones, cars and other devices that cast, estimating that over 1.5 billion smart “move” with low latency wireless commu- devices were sold globally. The report also nications makes controlling these devices estimated that by 2021, the world will con- from a remote location possible, potentially sume over 49 exabytes of data per month —a 7x increase over the usage in 2016. The acceptance, adoption and pervasiveness of smart devices astounds and has been, in ENHANCED MOBILE and of itself, transformational. 5G aims to BROADBAND go further. Broadband wireless data will con- tinue to draw attention, and the world’s stan- dardization bodies shaping 5G have taken notice. At the outset of the 5G standardization MASSIVE MACHINE ULTRA-RELIABLE LOW kickoff, the 3GPP outlined three key per- TYPE COMMUNICATION LATENCY COMMUNICATION formance metrics (see Figure 1). The 3GPP defined the enhanced mobile broadband (eMBB) use case and attached a perfor- mance target of greater than 10 Gbps peak data rate to expand broadband data servic- es. In response to several industry groups, s Fig. 1 5G use cases defined by 3GPP. WWW.MWJOURNAL.COM/ARTICLES/29474 4
RAN RAN RAN RAN #80 #74 #75 #78 (Rel-15 Completion) 2016 2017 2018 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 5G Study 5G NR Work Item 5G NR NSA 5G NR SA Further Evolution Completion Completion Stage 3 Completion Stage 3 Completion for Non-Standalone 5G-NR for Standalone 5G-NR NSA Option 3 Family ASN-1 Rel-15 ASN-1 for SA & NSA NSA = Non-Standalone = EPC Core (”Option 3”) & LTE Anchor SA = Standalone s Fig. 2 Timeline for 5G radio access network standardization. The first NR specification release is scheduled for late 2017, with updates during 2018. impacting construction, medicine, manufacturing, re- but the shape and structure of 5G NR has been crystal- tail services and safety. Latency in this context also in- izing for several months. “Final” solutions will inevitably cludes delivering timely information from the cloud or need some tweaking to meet the standard; however, deployed sensors to the brains of these devices, so that progress toward the creation of the ecosystem has al- decisions can be made on the fly to enhance safety. In ready started with a path toward commercialization. this case, the data is delivered real-time, and the control Service operators have announced 5G plans in all mechanism is deployed on the device. shapes and sizes: SKT and KT are gearing up for 5G In 2015, when the 3GPP kicked off the 5G standard- trial services to accompany the 2018 PyeongChang ization effort, the group outlined the timelines and Winter Olympics in South Korea. In the U.S., Verizon key performance objectives for this new standard. The has aggressively purchased spectrum in the 28, 37 and 3GPP stated unprecedented guiding principles for the 39 GHz bands and driven the development of the 5GTF definition process to follow. First, 3GPP broke compat- standard, primarily for fiber to the premises (FTTP) ap- ibility with prior releases, setting a goal of forward com- plications. Verizon has been trialing pre-commercial patibility. By breaking with LTE and prior-generation equipment in 11 cities in the U.S. since the beginning standards, the 3GPP opened a path for innovation to of this year and announced plans for initial commercial meet these very difficult objectives. Second, the 3GPP deployments in 2018. T-Mobile, the big winner in this divided 5G into phases. The first phase, or Phase 1, fo- year’s FCC auction, won 31 MHz of spectrum around cused on mobile access below 40 GHz and set a frame- 600 MHz and announced plans to build a “nationwide” work for Phase 2 to investigate spectrum above 40 GHz. 5G network using their newly purchased spectrum. In all, the 3GPP has been working on 3GPP release 15, Sprint has approximately 120 MHz of spectrum in the also known as 5G New Radio (NR) Phase 1, with an ex- 2.5 GHz band and has been working with Qualcomm pected release date of June 2018 (see Figure 2). and SoftBank, its parent company, to plan 5G rollouts in 2019. Meanwhile, AT&T has announced plans for IoT WHERE WE ARE TODAY services in spectrum it currently owns and acquired Fib- As I write this article, the 3GPP is closing in on the first erTower to obtain licenses at 24 and 39 GHz. draft of the physical layer of 5G NR Phase 1, targeted for Since the 3GPP kicked off the 5G standardization ef- December 2017. This first draft is critically important, as fort in 2015, the mMTC use case has been deprioritized. it establishes the foundation upon which semiconductor, The 3GPP continues to evolve LTE; in release 14, the device, infrastructure, test and measurement and other 3GPP made several enhancements to LTE specifically wireless ecosystem players will plan and build their busi- targeting the mMTC use case, with development of the nesses. Until this point, the development has evolved us- NB-IoT and LTE CAT-M standards. The mMTC use case ing system prototypes for field trails with service operators. elevates connectivity as a goal, driving device manufac- With a firmer standard in place, the players can develop turers to incorporate wireless capabilities into many de- tangible plans and targets for product and service rollouts. vices not previously connected, expanding their utility. Interestingly, there have been announcements re- We have seen a glimpse of the possibilities with new IoT garding the availability of 5G technologies—specifi- devices, but there are significant challenges: there is no cally by Intel and Qualcomm—and these early develop- pervasive or ubiquitous wireless IoT standard. As such, ments are intended to seed other companies to drive there are challenges with interoperability and seamless adoption. It may seem strange to announce products connectivity to infrastructure and smart devices. With compliant with a standard before the standard is final, the 3GPP addressing the mMTC use case in release 14 5
and delivering a comprehensive and widely supported standard, time will TABLE 1 tell whether further enhancements LTE VS. 5G NR PHASE 1 are needed in a future evolution of LTE NR Phase 1 5G. Number of Streams SISO SISO per Polarization per Antenna Panel WHAT WILL 5G LOOK LIKE? BW 20 MHz 800 MHz As noted, the 3GPP plans to final- Subcarrier Spacing 15 kHz 240 kHz (Maximum) ize 5G NR Phase 1, 3GPP Release 15 FFT Size 2048 2048, 4096 by the end of 2017, with the ASN.1 Number of Occupied ~1600 (FFT Size 2048) ratification in June of 2018. The 3GPP Subcarriers 1200 ~3300 (FFT Size 4096) has started on the path of defining the transformational radio access net- Spectral Occupancy 90% 98% work by including wider bandwidths, Slot Duration 0.5 ms 125 µs [14 Symbols for 120 kHz Subcarrier Spacing] essential for faster data rates. New [7 Symbols] spectrum has been identified to de- Antenna Omni 64 Beams (for SS Block) ploy these wider bandwidth systems. The 3GPP has also reduced the sym- NSA and will surely improve data bol timing compared to LTE, to en- rates and latency to be much closer EPC able shorter transfer time intervals to the 5G targets. The 3GPP is tar- (see Table 1). In addition, the 3GPP geting December 2018 to wrap up has aligned on a self-contained sub- Phase 1, release 15, which will in- UP UP frame, which enables transmission clude both NSA and SA. CP + and reception in a single subframe The 3GPP has accomplished for time-division duplexing (TDD) sys- much in a very short time. In the tems. With this initial work, the 3GPP short-term, 5G deployments be- has addressed faster data rates and low 6 GHz may look a lot like LTE CP + UP lower latency. Perhaps most impor- on steroids, i.e., faster data and NR gNB LTE eNB tantly, a new, flexible numerology will lower latency. The first NSA deploy- enable operators to accommodate ments may provide noticeable per- Data Flow Aggregation Across different types of devices and support formance enhancements over LTE, LTE eNB and NR gNB via EPC diverse use cases. and the lightning fast data speeds (a) For mmWave, the 3GPP has iden- will likely appear when network op- tified specific frequency bands and erators deploy mmWave technolo- NextGen Core incorporated beam management gies and the new 5G core network, and control for phased array anten- needed for SA operation. What is nas (PAA). Although the stage is set clear is that this is just the begin- for mmWave deployments, many ning. Future evolutions and itera- practical challenges remain for wide- tions seem inevitable. spread adoption (more on this later). eLTE eNB The 3GPP defined two main net- CHALLENGES AHEAD eLTE eNB Connected to NextGen Core work architectures for Phase 1. With As the 3GPP finalizes the formative (b) the non-standalone (NSA) architec- 5G specification, the path forward is ture, the 5G NR uses the existing LTE not unimpeded. 5G has the poten- radio access network and evolved tial to be a “game changer,” but the NextGen Core packet core or EPC (see Figure 3a); transformational impact must come NSA includes two additional options with extensive help from a diverse set UP UP (see Figures 3b and 3c). The second of players. Potential challenges exist CP + main architecture, named standalone in three high level areas: mmWave, (SA), uses the 5G NR and a new 5G network topology and ecosystem. core (see Figure 4). NSA enables op- erators to offer 5G services sooner, mmWave CP + UP taking advantage of the existing in- The 3GPP chose to incorporate NRgNB NR gNB frastructure to deliver services in the mmWave technologies into the eLTE eNB short-term, since investments for SA standard due to the scarcity of avail- Data Flow Aggregation Across are expected to be much larger and able spectrum below 6 GHz. More eLTE eNB and NR gNB via NextGen Core will take more time. In the standards spectrum equates to faster data (c) meetings, NSA has been a focus be- speeds. Although the 3GPP will cause of the immediate opportunity specify 5G technologies for use in s Fig. 3 Initial 5G deployments will use spectrum below 6 GHz, the 3GPP the existing LTE radio access network and, perhaps, a narrower scope. SA and EPC (a). The NSA specification will deliver more 5G benefits than is relying on mmWave, with its copi- includes architectural options using the new 5G core (b) and (c). 6
ous spectrum, to meet its goals for tically, into the software deployed 3GPP, but the implication was clear. the eMBB use case. Over the last on these systems. To support mo- To describe this flexible network couple of years, several researchers bility, the protocol software must topology, the wireless industry de- have prototyped mmWave systems switch the beams in less than 200 fined the term “network slices.” extensively, but the early prototypes ns to maintain the link, requiring fast Network slices describes the ability were big, bulky and used very new switching technology in the antenna of a service operator to “slice” the technologies such as PAAs. assemblies and the software archi- network to tailor a unique set of ser- PAAs overcome the free space tectures that program them. vices for users, creating diverse ap- path loss associated with mmWave The testing of PAAs and the sys- plications and use cases and charg- transmission and reception using tems that incorporate them is being ing appropriately for the services. multiple antenna elements and investigated and poses new chal- With network slices, a company or beamforming to enhance gain. lenges for the test and measure- individual could purchase a service With their benefits, PAAs also pose ment industry. As PAAs are often or a set of services to meet specific system challenges, as the control of integrated with their transceivers to needs. For example, consider a the beams must be incorporated minimize loss, cable access to these company that outfits a factory with into the standard and, more prac- modules and the arrays that incorpo- 1,000 connected sensors. The com- rate them will not exist. Over-the-air pany may expect to pay less than (OTA) techniques for testing PAAs $40 per month for unlimited data, NextGen Core are being explored by several com- since those sensors do not transmit panies, with proposals submitted or receive the type of data that we to the 3GPP RAN4 working group consume on our smartphones. for incorporation into the standard. While the 3GPP has, indeed, OTA introduces new variables to the made great progress toward de- test equation; most significant is the fining a radio access network ca- need to minimize test time and test pable of achieving these lofty goals, cost. Test and measurement compa- network operators must be able NR gNB nies must deliver fast, cost-effective to fairly charge for these services NR gNB Connected to NextGen Core solutions to the wireless industry to and conserve valuable network re- (a) facilitate the development of the sources to sustain a healthy ecosys- mmWave ecosystem. tem and enable all contributors to Even with PAAs in both the user prosper. To facilitate the creation NextGen Core equipment (UE) and infrastructure of network slices, the 3GPP has en- (i.e., gNodeB), mmWave propaga- hanced the split architecture of the tion is limited, even at the lower control and user planes to enable P +U mmWave frequencies. Denser de- separate control and data paths. UP ployment of the infrastructure is a This is just the foundation. Network CP foregone conclusion that will likely slicing also depends on implement- signal more costly rollouts of the ing infrastructure elements beyond CP + UP technology and systems. To ad- the physical layer of the protocol eLTE eNB dress the density challenge, re- stack. Network technologies such NR gNB searchers are exploring new tech- as virtual EPC, network function vir- niques for mesh and integrated tualization (NFV), software defined Data Flow Aggregation Across NR gNB and eLTE eNB via NextGen Core access backhaul (IAB) to minimize networking (SDN) and mobile edge (b) the cost of denser deployments computing (MEC) are components by utilizing the 5G gNodeBs al- and services necessary to move net- ready deployed. IAB would reduce work slicing forward. Without these NextGen Core the cost of running fiber to each technologies, all data and control mmWave access device; however, traffic must aggregate at the core P the technique may introduce more network, potentially crippling the in- UP +U latency. dustry’s ability to meet the goals of CP data throughput, end to end (E2E) Network Slicing latency and massive connectivity. One of the more impactful ob- CP + UP servations for 5G transformation is Creating the Ecosystem NRgNB NR gNB NR gNB that the networks must morph and 5G’s success or failure will de- scale to optimize resources to sup- pend on the creation of an eco- Data Flow Aggregation Across port new applications and services. system. The 5G ecosystem must NR gNBs via NextGen Core (c) “Scalable” networks was not an ex- extend beyond the traditional plicit goal outlined by the ITU or the wireless value chain of usual par- s Fig. 4 The SA specification assumes Next Generation Mobile Networks ticipants: the service providers, the 5G NR and a new 5G core (a), with two additional options to handle (NGMN) Alliance when the 5G stan- semiconductor companies, infra- deployment scenarios (b) and (c). dardization process was kicked off by structure manufacturers and test 7
and measurement companies, to Markit is not alone in their predic- ership, 5G has unstoppable mo- name a few. Application software tion, as the world’s economic lead- mentum. and service providers, cloud and ers continue to invest in 5G with a The industry has made great cloud infrastructure, vertical inte- myriad of funding and regulatory strides in moving the 5G agenda grators, software companies and support. These global leaders are forward, achieving key milestones even car, drone, appliance, medi- betting on 5G to catalyze GDP in both the standardization process cal device and construction manu- growth and create economic pros- and technology development. De- facturers must be an integral part perity. While 2035 is 20 years away, monstrable commitments from aca- of the 5G landscape to realize the the 5G foundation is being laid demia, industry and governments true economic potential. Creating today. Creating an ecosystem for worldwide have created forward an ecosystem does not occur over- these applications will take invest- momentum, yet there remains much night. The initial deployments of ment, dedication and perseverance to do. 5G below 6 GHz may have 5G services to enable the ecosys- and, above all, time. The various a shorter runway to deployment, tem to grow and evolve are critical. ecosystem players must step up to but mmWave is very important to the plate to realize the vast poten- the overall 5G vision. The next year A 5G FUTURE tial that 5G promises. should provide a better picture of Qualcomm recently commis- As noted, initial rollouts are the 5G timelines and potential, and sioned a study by IHS Market to scheduled for next year, and more the world will be watching the prog- assess the economic impact of 5G. meaningful deployments will begin ress on the challenges outlined in IHS Markit speculates that 5G will in 2019. To realize 5G’s potential, this article. The major players have become a general purpose tech- significant innovations must occur all anted up, but there are real and nology (GPT), a development so in semiconductors and packaging hard problems to solve for 5G to impactful that it becomes a catalyst technology, system and network live up to its promise. 2018 should for socio-economic transformation. topologies and architectures and, be an interesting year and this time For perspective, other examples in of course, the important verticals next year, a clearer picture of the fu- our history cited as GPTs include that will take advantage of this new ture should materialize. the printing press and electricity. network of 5G capabilities and ser- The central question is not IHS Markit expects 5G to contribute vices. It will not be easy, but with whether 5G will be impactful, the $12.3 trillion—yes, that is a “t”—to the commitment of the industry question is when?■ the global economy by 2035. IHS and the world’s government lead- 8
Opportunities for High Frequency Materials in 5G and the IoT Art Aguayo Rogers Corp., Chandler, Ariz. N Enhanced Mobile Broadband ot a day goes by that we don’t encounter the terms 5G and IoT Gigabytes in a Second (Internet of Things). Indeed, the 3D Video, UHD Screens future is about super-connectivity Work and Play in the Cloud and the promise it gives us, once everything Smart Home/Building Augmented Reality is “talking” with each other. Connected Industry Automation appliances to connected cars, making life Voice Mission Critical easier as we prepare for smart homes, smart Smart City Future IMT Application cities and smart everything. It is difficult to Self Driving Car imagine this future—yet so was imagining 2015 when looking through the view of the Massive Machine Type Ultra-Reliable and Low Latency internet in the year 2000. Communications Communications There appear to be as many definitions (a) of 5G and IoT as forecasts and opinions when discussing the potential benefits and Peak Data Rate (Gbps) User Experienced Data Rate (Mbps) relevant business cases of these technolo- 20 100 gies. The International Telecommunication IMT-2020 10 Union (ITU) has been working on defining Area Traffic 1 Spectrum Efficiency what 5G IMT-2020 will be from a technical Capacity 10 perspective, or at least how it will differ in (Mbps/m) 1 1× 3× 0.1 performance from 4G (IMT-Advanced). The 1× 350 term 5G IMT-2020 was coined in 2012 by 100× 10× 400 500 the ITU Radiocommunication sector and Network Energy IMT-Advanced Mobility (km/h) means “international mobile telecom- Efficiency 105 10 munication system,” with a target date of 106 1 2020. Within that definition, we see how Connection Density (Devices/km2) Latency (ms) IoT will benefit. Parameters like peak data rates, mobility, latency and spectrum ef- (b) ficiency are important, as they help define s Fig. 1 ITU vision for IMT-2020, including usage scenarios (a) what the user experience will be, key to and enhancement of capabilities from IMT-Advanced (b). enhanced mobile broadband (eMBB) and ultra-reliable and low latency communica- tions (URLCC). Figure 1 shows how the ITU envisions 5G. IoT, on the other hand, will need different parameters to operate in a way that needs WWW.MWJOURNAL.COM/ARTICLES/27688 9
in Figure 3). The focus of these ma- TABLE 1 terials has been low cost and ease IoT APPLICATIONS of manufacturing for complex multi- IoT Category Example Applications layer boards, with no real attention on repeatability of electrical proper- Massive IoT Smart Building, Transport Logistics, Fleet Management, Smart Meters, Agriculture ties, since the applications where these are used don’t require that Critical IoT Traffic Safety, Autonomous Vehicles, Industrial Applications, level of performance. The second Remote Manufacturing, Healthcare (including Remote Surgery) set of materials use specialty resins, sometimes blended with epoxy, and Because much of 5G will be data achieve some improvement in loss. intensive, the frequencies around These materials have mainly been 28, 39 and 77 GHz are gaining mo- used in high speed digital applica- Fixed Phone Mobile Phone mentum because of the availability tions up to 10 Gbps. The last group PC/Laptop/Tablet Non-Cellular IoT Cellular IoT of spectrum within those bands. As is defined as high frequency mate- 30 many IoT applications are expected rial (found in tiers 5 and 6), where 25 to be low data rate, most of the IoT the Df is less than 0.005. (Billions) 20 15 activity is centering on the sub-6 Various parameters are consid- 10 GHz spectrum. An exception will ered in the selection process when 5 be IoT for surveillance, where trans- deciding which type of PCB mate- 0 2015 2021 mitting high definition video from rial to use: loss, dielectric constant, remote areas may require the band- thickness, thermal conductivity and, width found in the millimeter wave let’s not forget, cost. In the end, it is s Fig. 2 Connected devices forecast, spectrum. about selecting the appropriate ma- from 2016 Ericsson Mobility Report. terial at the right cost. Much of the minimal to no user action on a day- MATERIALS PERSPECTIVE IoT market today is using traditional to-day basis, after the initial setup. Printed circuit boards (PCB) are FR-4 in the transceiver or antenna Today, we are seeing the beginning a building block in any electronic portions of the radio. However, of the IoT market, an expansion of system. The choice of PCB mate- there is a subset of this market that the M2M market that already ex- rial for RF applications depends requires a higher level of reliability, ists and accounts for 600 million on frequency, power level, circuit including industrial, medical, traffic devices as of 2015.1 The IoT can size and function. Designers have control, automotive and smart me- be divided into two segments.2 The choices from basic epoxy/glass ters. This subset is taking advantage first represents massive IoT connec- materials (FR-4), mid-loss materi- of the higher performance materials tions with high connection volume, als and, ultimately, high end mi- and the increased focus on reliabil- low cost, low energy consumption crowave/millimeter wave materials. ity that tier 5 and 6 materials can and small data traffic. The second The most common PCB material provide. comprises critical IoT connections is FR-4, mainly developed for the So what are the benefits of se- that require ultra-reliability and mechanical properties key to mul- lecting a high performance mate- availability with very low latency. tilayer circuit boards. Variations of rial instead of FR-4? The first ben- Table 1 shows applications for each these materials exist, offering dis- efit a designer will notice is the of the two classifications. Figure 2 sipation factor (Df) or loss tangent impact loss tangent has on the shows a forecast for the connected values ranging from 0.01 to beyond loss of the circuit. Many times, this devices market, based on Ericsson’s 0.02 (see the tier 1 and 2 materials is the primary consideration. This mobility report.3 The traditional connected-device market of fixed, TABLE 2 mobile phones and computer/ LOW POWER IoT STANDARDS4 tablets will increase slightly, while Bluetooth, NB-IoT EC-GPRS SigFox LoRa the overall number of devices as- Wi-fi, RFID, sociated with IoT, both cellular and ZigBee, non-cellular, will grow greater than Z-Wave 20 percent annually. The IoT space Range 10 cm to 200 m < 11 km < 11 km < 9 km < 7 km can also be viewed by how connec- Maximum < 100 164 164 160 157 tivity is achieved, particularly using Coupling low power technology. Various low Loss (dB) power standards are summarized in Spectrum, Unlicensed Licensed IMT, Licensed Unlicensed Unlicensed Table 2. Bandwidth 2.4 GHz 200 kHz 800 to 900 868 MHz, 868 MHz, Frequency band allocations for Shared MHz, Shared 600 Hz 125 kHz 5G focus heavily on available band- width, and they seem to center around three groups: sub 6 GHz, 15 Data Rate < 100 Mbps < 62 kbps UL
tivity and will extend the range of higher than with the more stable the antenna. materials in the figure. FR-4 can In some cases, it is not the loss of change as much as 400 ppm/°C. the material that is the main driver With materials like RO3003™ and for selecting a high performance RO4350B™ laminate, the short- Tier 6 Df < 0.002 material; sometimes it is the variabil- term change is close to 40 ppm/°C. Tier 5 Df = 0.002 to 0.005 ity of the dielectric constant. Most Considering all these factors (i.e., high performance materials have a tolerance, moisture absorption and Tier 4 Df = 0.005 to 0.008 tolerance of less than ±2 percent, temperature variation), selecting a Tier 3 Df = 0.008 to 0.01 even less for materials with low di- high frequency material over FR-4 electric constant, while FR-4 materi- may be the best choice when a Tier 2 Df = 0.01 to 0.02 als can be greater than ±5 percent. more consistent design is needed Df > 0.02 The increased variability of the FR-4 in the field. Tier 1 may require the circuits to be tuned In many cases, the IoT will be to ensure they operate within the about having connectivity with a cir- frequency specified, while materials cuit as small as possible, due to lim- with tighter tolerance will not need ited space. In these cases, reducing such optimization. the size of the antenna or circuit will s Fig. 3 PCB material classification by Environmental changes can af- be desired. By selecting materials loss tangent (Df). fect the dielectric constant of FR-4, with a dielectric constant of 6, 10 or difference can be almost an order yet have minimal impact on high higher, designs using FR-4 with a Dk of magnitude greater with some performance materials. FR-4 has of 4.4 can be reduced in size. Using materials. To keep it simple — and significantly higher moisture ab- a PCB with a Dk of 4.4, the wave- not include in the analysis the im- sorption than high performance length at 1 GHz for 0.020” microstrip pact dielectric constant has as Dk materials. This leads to an increase is about 7”, while a material with a of FR-4 is 4.4 and many high fre- in Dk (also Df). If the circuit needs Dk of 10.2 (e.g., Rogers RO3010™ quency materials options are low- to operate in a high moisture en- laminate), the wavelength is 4.4”, er — consider the simulated in- vironment—tropical areas such close to a 40 percent reduction in sertion loss for a 50 Ω transmission as Malaysia—FR-4 materials have size.5 Higher dielectric constant ma- line on FR-4, with Dk = 4.4 and been known to drift, due to the terials allow designers to shrink the a dielectric thickness of 0.020”. change in dielectric constant. In size of the circuit board, saving area The 50 Ω width is calculated to be comparison, moisture has minimal compared to FR-4. 0.038”.5 Comparing the change in impact on the dielectric constant There will also be IoT applica- insertion loss when Df varies from of the high frequency laminates. tions that operate at higher fre- 0.02 to 0.004 at 2.4 GHz for this Changes in the temperature can quencies, potentially in the 28 to line width, the insertion loss is 0.24 also have a significant impact on 40 GHz range, and not necessarily dB/inch for a Df of 0.02; for a Df the operation of a circuit: with on low power networks where the of 0.004 the insertion loss will only FR-4, notably, dielectric constant use of high frequency materials is a be 0.01 dB/inch. The benefit here changes with temperature (see must. Managing losses is extremely is that if the circuit is an antenna, Figure 4). The change with FR-4 important, so using materials with the lower loss improves the sensi- is close to an order-of-magnitude low dissipation factors is critical, as well as selecting copper foils that are smooth, to reduce the conduc- tor impact on insertion loss. In the case of materials based on PTFE RO3003 RO4350B RO4003C PTFE/WG Dk 3.2 PTFE/WG/Ceramic FR-4 resin, this is often addressed by using rolled copper foil instead of 1.04 traditional electrodeposited cop- 1.03 per foil. However, for materials that use low loss thermoset resins, us- Change from 25°C 1.02 ing smooth foil impacts copper 1.01 peel strength, in many cases lower- 1.00 ing the value to the limit specified by the industry. To address this, 0.99 Rogers introduced LoPro® copper 0.98 foil to go with Rogers RO4000® 0.97 materials, allowing designers to re- –50 –25 0 25 50 75 100 125 150 duce insertion loss while maintain- Temperature (°C) ing copper foil peel strength to the level of standard copper. s Fig. 4 Normalized dielectric constant vs. temperature. 11
References CONCLUSION driving are topics of much interest. 1. Cisco, “Visual Network Index: Global 5G and IoT have moved from if Many of these may benefit by us- Mobile Data Traffic Forecast Update, they happen to when they happen. 2015-2020,” February 2016. ing higher performance materials, 2. International Communications Union, It is likely that the applications that especially for applications that use “IMT Vision–Framework and Overall Ob- will revolutionize our lives may not millimeter wave frequency bands. jectives of the Future Development of yet have been thought of. What we Much is yet to be defined about 5G IMT for 2020 and Beyond,” September do know is that many segments in 2015. and IoT, and we will no doubt see 3. Ericsson, “Ericsson Mobility Report,” the market will be undergoing sig- surprising use cases emerge. When June 2016. nificant change during the next few asked, “Why did we connect a par- 4. Nokia, “From NB-IoT to 5G,” IWPC years. Smart homes, smart cities, ticular thing to the internet,” we Workshop Presentation, October 2016. remote health care monitoring, in- may find ourselves saying “because 5. Simulation Using Rogers’ Microwave dustrial controls and autonomous Impedance Calculator, www.globalcom- we could.”■ mhost.com/rogers/acs/techsupporthub/ en/calculator.php. Microwave Impedance Calculator This software is intended to assist with microwave circuit design in predicting the impedance of a circuit made with Rogers High Fre- quency circuit materials. The software also has some capabilities for predicting transmission line losses as well. The user will select the cir- cuit materials and the circuit construction, after which the software will determine the predicted impedance and other electrical information. The calculator uses well known closed form equations to determine impedance and loss of a given circuit model. The loss calculation is divided into conductor loss and dielectric loss. With specific circuit designs, the calculator also predicts other properties such as wave- length in the circuit, skin depth and thermal rise above ambient. Free (Requires Registration) https://youtu.be/F_AeAUk7DuU 12
Paving the Way for 5G Wireless Networks The ROG Blog is contributed by John Coonrod and various other experts from Rogers Corporation, providing technical advice and information about RF/microwave materials. G rowing demand for mobile wire- search experiments, but 5G represents an less communications services has opportunity to “popularize” millimeter-wave quickly eclipsed the capabilities of frequencies and make them part of everyday Fourth Generation (4G) Long Term life, not just for exotic electronic devices in Evolution (LTE) wireless networks and cre- the limited quantities used in research and ated a need for a next-generation mobile by the military, but for potentially billions of wireless network solution. Fifth Generation electronic devices for people and things, as (5G) wireless networks promise more capac- in how Internet of Things (IoT) devices will ity and capability than 4G LTE systems, us- use 5G networks for Internet access. ing wider channel bandwidths, new antenna Designing circuits at millimeter-wave fre- and modulation technologies, and higher quencies starts with the right PCB material, carrier frequencies even through millimeter- and knowing how different PCB characteris- wave frequencies. But before 5G wireless tics affect circuit performance at millimeter- networks can become a reality, systems and wave frequencies. Variations in certain cir- circuits will be needed for higher frequen- cuit material parameters, such as dielectric cies than current 2.6-GHz 4G LTE wireless constant (Dk), can have greater impact on networks. performance as the operating frequency in- Standards are still being formulated for creases. For example, signal power is a valu- 5G wireless networks, with goals of achiev- able commodity at millimeter-wave frequen- ing data rates of 10 Gb/s and beyond with cies, requiring circuit designers to minimize low latency, using higher frequencies than loss in their circuits as much as possible. in traditional wireless communications sys- This begins with the choice of PCB material, tems. In the United States, for example, last since a PCB material not meant for use at year the Federal Communications Commis- millimeter-wave frequencies can result in ex- sion (FCC) approved the use of frequency cessive signal losses when operated beyond bands at 28, 37, and 39 GHz for 5G. its intended operating frequency range. PCB materials can degrade signal power PCB MATERIALS FOR MILLIMETER in three ways: radiation losses, dielectric loss- WAVES es, and conductor losses. Losses through ra- For circuit designers, one challenge will diation of EM energy largely depend on the be in knowing where to start, which means, circuit architecture, so even the lowest-loss for millimeter-wave frequencies, knowing PCB material may not save a circuit configu- what types of printed-circuit-board (PCB) ration that has a tendency to radiate energy. material characteristics are the most impor- A thoughtful choice of PCB material can tant at higher frequencies. Millimeter-wave help minimize dielectric and conductor loss- frequencies (above 30 GHz) were once used es at millimeter-wave frequencies. A circuit almost exclusively by the military and for re- material’s dielectric loss is closely related 13
to its dissipation factor (Df) or loss effects of copper surface rough- be maintained in climate-controlled tangent, which increases with fre- ness on loss become apparent at environments, circuits may be sub- quency. The Df is also related to a millimeter-wave frequencies. For ex- ject to changing environmental material’s dielectric constant (Dk), ample, two circuits based on 5-mil- conditions, such as high relative with materials that have higher val- thick RT/duroid® 6002 circuit mate- humidity (RH). Water absorption ues of Dk often have higher Df loss, rial from Rogers Corp. but with two can dramatically increase the loss although there are exceptions. At- different types of copper conductor of a PCB material, and the loss of tempts to minimize dielectric losses and surface roughnesses were test- circuit materials with high moisture for millimeter-wave circuits can be ed at 77 GHz. The circuit with rolled absorption will be greatly affected aided by considering circuit materi- copper and root mean square (RMS) under high RH conditions. als with low Df values. conductor surface roughness of 0.3 Testing on 5-mil-thick RO3003™ μm exhibited considerably lower circuit material from Rogers Corp. CONTROLLING CONDUCTOR conductor loss than the same circuit for two different operating environ- LOSS material with electrodeposited (ED) ments showed how loss at millime- Finding a material with low con- copper conductor having 1.8-μm ter-wave frequencies can increase ductor losses at millimeter-wave fre- surface roughness. with RH. One circuit was maintained quencies is not as straightforward, Propagation of the small wave- at room temperature and the other since conductor losses are deter- lengths at millimeter-wave frequen- was subjected to +85ºC and 85% mined by a number of variables, cies can also be affected by the RH for 72 hours. At 79 GHz, the including the surface roughness type of finish used on a PCB’s con- room temperature material had and the type of finish. As the name ductors. Most plated finishes have about 0.1 dB/in. less loss than the suggests, millimeter-wave signals lower conductivity than copper, and material subjected to higher humid- have extremely small wavelengths, their addition to a copper conductor ity and temperature. When testing mechanical variations in a circuit- will increase the loss of the conduc- was performed on a third, thermo- board material can have significant tor, with loss increasing as the fre- set circuit material from a different effects on small-wavelength signals. quency increases. Electroless nickel supplier, the increase in circuit loss Increased copper surface roughness immersion gold (ENIG) is a popular at 79 GHz was even more dramatic. will increase the loss of a conductor, finish for copper conductors; unfor- For those interested in learning such as a microstrip transmission tunately, nickel has about one-third more about the nuances of select- line, and slow the phase velocity of the conductivity of copper. As a re- ing PCB materials and designing signals propagating through it. In sult, ENIG plating will increase the circuits for 5G, in particular at milli- microstrip, signals propagate along loss of a copper conductor, with the meter-wave frequencies, Rogers has the conductor, through the dielec- amount of loss increasing as a func- created a number of tutorial videos tric material, and through the air tion of increasing frequency. in their “The Road to 5G” series. around the circuit material, so the The videos guide viewers on what roughness of the conductor at the ENVIRONMENTAL EFFECTS different circuit material parameters interface with the dielectric mate- Environmental conditions can mean at millimeter-wave frequen- rial will contribute to the conductor also impact the amount of loss ex- cies, and which material character- loss. The amount of loss depends hibited by a PCB material, especial- istics make the most difference at on frequency: the loss is greatest ly at millimeter-wave frequencies. those higher frequencies. The vid- when the skin depth of the propa- Many network scenarios for 5G pre- eos offer quick and easy ways to gating signal is less than the copper dict the need for many smaller wire- learn how to specify PCB materials surface roughness. Such a condition less base stations than used in ear- for 5G, and to get ready for this next also degrades the phase response lier wireless network generations, revolution in wireless communica- of the propagating signal. in part because of an increased tions. The impact of copper surface number of expected users and the Do you have a design or fabrica- roughness on conductor loss de- use of millimeter-wave frequencies tion question? Rogers Corporation’s pends on the thickness of the PCB and their shorter propagation dis- experts are available to help. Log in material: thinner circuits are more tances than lower-frequency carri- to the Rogers Technology Support affected than thicker circuits. The ers. Where 5G base stations cannot Hub and “Ask an Engineer” today. 14
Transformation to 5G: PCB Advantage Tony Mattingly Rogers Corp., Chandler, Ariz. H ard to believe it has been an hour a day of mobile video at 1 over 30 years since 1G Mbps throughput—which is typical was introduced to the of applications supporting stream- world, back when there ing video like YouTube or Netflix— were no official requirements and consumes 13.5 GB per month. The analog technology was being used. demand for data will continue to The 1990s brought 2G, with the grow as the 5G and IoT ecosystems introduction of digital technology are created in the coming years (see and the ability to transmit text mes- Figure 1), and we will all eventu- sages via your cellular device. Ten ally live in a world that depends on years later, in 2000, an upgrade to billions of devices communicating 3G brought IEEE WiMAX standards with one another. The connectivity that were designed to support 30 of these devices will be expected to to 40 megabit per second (Mbps) be instantaneous and without inter- data rates. Then, 4G was intro- ruption. duced a mere seven years ago and, Adding capacity is not cheap, although it brought to the world and space for additional antenna a much higher spectral efficiency, towers is limited. In the U.S., most exponential consumer demand towers are not owned by carriers, for wireless data is now driving a rather companies such as American need for substantially higher mo- Tower, Crown Castle and SBA. Car- bile network capacity and perfor- riers lease space on these towers mance. Let’s take, for example, the and, as strict permitting processes primary driver for data usage on make it difficult to build new sites, smartphones: video streaming. Just carriers are forced to replace anten- nas with newer and more efficient ones. Arguably, the cost for opera- tors to deliver data is proportional to the spectral efficiency of the wireless technology. LTE networks have the highest spectral efficiency of any technology to date, but it is not going to be enough to serve consumer desire to always be con- nected. So how is capacity added in an already saturated wireless infrastructure? One answer is ad- vanced antenna designs employ- ing new technologies that operate within current frequencies. Another answer is adding wireless infra- structure that operates at new and higher frequencies, where plenty of s Fig. 1 Our world is increasingly connected, which is driving spectrum exists. Early in 2015, U.S. the growing demand for wireless data. carriers spent more than $40 billion WWW.MWJOURNAL.COM/ARTICLES/28308 15
EMBB • Personalized Broadcasting Enhanced • Fixed Broadcasting Mobile to Home, GB High User Broadband • Augmented Reality Peak Data Importance Experienced EMBB Rate Data Rate Medium Area Spectrum Traffic Capacity Low Efficiency • Connected Car Network • Home IoT MMTC URLLC • Remote Surgery Energy Efficiency Mobility • Smart City Massive Machine-Type Ultra-Reliable • Remote Control • Smart Grid Communications Low Latency (Drone, Robot) MMTC URLLC Connection Latency Density (a) (b) s Fig. 2 5G is envisioned to support three general use cases: high data rate, low latency and massive numbers of connections (a), and these use cases define the performance requirements of the network (b). on spectrum! Because much of the spectrum below 2.5 high data rates. While millimeter wave frequencies suffer GHz is limited, frequency bands around 3.5, 28, 39 and from characteristics such as higher propagation loss— 77 GHz are gaining interest because of the availability even with line-of-sight conditions and no obstructions— of bandwidth. this challenge can be overcome with beamforming an- tenna arrays or massive MIMO. 5G COMING SOON Regulatory policies around the world are striving to We should start by clarifying the conceptual goals keep pace with these changing technologies. Some for 5G (see Figure 2). Cost to deliver data is crucial to of the complex issues to be addressed: allocating and the success and survivability of the carriers, so the 5G managing the new spectrum, maintaining neutrality infrastructure will need to support a greater number of within the networks and preserving privacy. Spectrum devices at lower average revenue than with 4G systems. is considered a precious commodity within the industry. The infrastructure must also provide peak data rates of This could be seen during the U.S. auction for the 3.5 multi-gigabits per second (Gbps), facilitate a user expe- GHz small cell band last year, spectrum the FCC is now rience that is uniform throughout the coverage area— enabling. As advancements in technology provide a no matter the device density—support numerous fre- path for small cell operation between 6 and 100 GHz, a quencies (including cellular bands and frequencies vast amount of available spectrum is introduced. Wider above 6 GHz), use both licensed and unlicensed bands radio channels operating at higher frequencies enable and follow advanced spectrum sharing rules. much higher data rates. Small cells offload data from The 5G networks of the future will require infrastruc- macro cells, inherently increasing capacity, and can of- ture that has a wide range of capabilities and operates fer improved signal quality in places with a higher con- both below 6 GHz and at millimeter wave frequencies. centration of users or where the signal from a macro These networks will use massive MIMO antenna struc- tower is weak. Millions of small cells will eventually be tures and massive carrier aggregation and will need to deployed, leading to massive increases in capacity. The operate “lightning fast” with very low latency. Massive industry is slowly overcoming challenges that have im- MIMO comprises hundreds of antennas at the base sta- peded small cell deployments. These include govern- tion that enables spatial multiplexing and beamforming. ment regulations, acquiring real estate for the antennas These systems provide three times the spectral efficiency and managing and preventing interference. Another of today’s LTE-Advanced antennas. Carrier aggrega- challenge relates to spectrum sharing. Future wireless tion, which is a key LTE-Advanced feature that operators systems may interface with a planned spectrum access are deploying globally, uses spectrum more effectively, system that manages spectrum among primary users— increases network capacity and user throughput rates government agencies, in some cases—and secondary and provide new ways to integrate unlicensed spec- or tertiary users. This will enable more efficient use of trum. Frequency band allocations for 5G focus heavily spectrum for scenarios in which incumbents use spec- on bandwidth availability and seem to center around trum lightly. three groups: below 6 GHz, 15 to 40 GHz and above 60 Many countries have started 5G trials, focusing on GHz. The combination of lower and higher frequencies various applications and using different frequencies. Ma- is crucial for 5G operation. Lower bands can be devoted jor carriers in the U.S. are conducting trials in 2017 with a to coverage and control, while the higher bands enable focus on fixed broadband at 28 GHz. South Korea is also 16
conducting trials at 28 GHz to pre- of wireless infrastructure. These de- ed using a patented LoPro copper pare for the 2018 Olympic Games signers need to overcome the many foil, and it incorporates low density in Seoul. Japan plans to start trials challenges and understand what microspheres, yielding a laminate in Tokyo during 2017, using both performance properties are needed that is 30 percent lighter than PTFE sub-6 GHz and 28 GHz, and they to meet their design goals. materials. The coefficient of ther- will likely scale up the trials signifi- With the use of additional fre- mal expansion (CTE) is matched to cantly during 2018 and 2019. China quency bands and consumer de- copper in the X and Y directions, has announced ongoing 5G trials at mand for better performance and which minimizes bow and twist and 3.5 GHz, with a major carrier testing lower latency, PCBs have advan- allows for hybrid MLB construction, seven experimental base stations tages compared to competing and the material has a low Z-axis in several cities throughout 2017. technologies, such as bent metal CTE of 30.3 ppm/°C from -55°C to Additionally, the European Com- and cable. Designers at antenna +288°C, for reliable plated through mission recently published their 5G original equipment manufacturers holes (PTH) in multilayer circuit as- action plan, with preliminary trials (OEM) find that PCB-based de- semblies. These laminates offer a starting in 2017 and pre-commercial signs shorten the design iteration practical, cost-effective circuit ma- trials in 2018; they will use 3.4 to 3.8 cycle and enable the development terial for active antenna arrays and and 24.25 to 27.5 GHz. However, of complex, multilayer board (MLB) PCB antennas, whether for current all these trials will be very targeted designs. However, PCB-based de- wireless systems or those on the ho- and limited in scale, constrained by signs have challenges: integration rizon. With the right combination of the pace of 5G standards, technol- of the power amplifier (PA) and materials, these laminates provide ogy development and economic antenna into one structure for ac- the optimum blend of price, perfor- justification. This view has also been tive antennas and higher fabrica- mance and durability. confirmed by the November 2016 tion and assembly costs of MLB Rogers, like other PCB materials forecast from ABI Research, estimat- approaches using PTFE materials. companies, continues to focus on ing 5G subscribers will reach 4 mil- While thermoset high frequency market trends and material needs lion in 2020 and 349 million in 2025, materials are ideal for MLB de- of the future. Under development accounting for only 0.04 percent of signs, there are limited thermoset is a whole family of thermoset lami- all subscribers in 2020 and 3.6 per- materials that have low loss, low nates with multiple thicknesses, low cent in 2025. passive intermodulation (PIM) and dielectric constant and low insertion are flame retardant, meeting the loss for emerging PA and small cell PCB ANTENNAS sought after UL 94 V-0 designation. point-to-point backhaul radio ap- With all that said, there is no To address these challenges, plications operating at millimeter doubt antennas for next-genera- PCB materials companies like Rog- wave. Rogers will soon add a ther- tion wireless systems (e.g., LTE-Ad- ers have launched new antenna moset solution designed to meet vanced and 5G) are becoming more grade laminates. Rogers’ solution the growing RF needs within the complex. With demand for mobile combines a flame retardant, low loss carrier grade Wi-Fi and distributed data projected to grow at approxi- thermoset dielectric with low profile antenna system (DAS) markets, mately 53 percent compound annual copper foil and incorporates a pro- where better loss, improved Dk growth rate through 2020, active an- prietary filler system. The material over frequency and more controlled tennas and small cells will continue has a dielectric constant (Dk) of 3.0, thickness tolerance are desired. to be deployed to handle the data a popular choice for antenna de- These product offerings will only throughput. These trends and the signers, and a dissipation factor (Df) serve to enhance an already com- associated growth are expanding of 0.0023 at 2.5 GHz. The laminate prehensive portfolio of the highest the use of printed circuit board (PCB) possesses a low thermal coefficient reliability and highest quality prod- materials in antenna designs. The of dielectric constant (TCDk), which uct offerings. PCB manufacturers material requirements will be differ- gives consistent circuit performance are meeting the needs of the anten- ent from the traditional cellular net- over a range of temperatures. It na market with new products aimed work. At the same time, 5G designs has low passive modulation (PIM at the performance criteria needed are completely new to the designers < -160 dBc), because it is construct- to meet 5G performance goals.■ Webinar Available On Demand The Impact of Final Plated Finishes on Insertion Loss for High Frequency PCB’s Sponsored by: Rogers Corp. and Sonnet Software Presented by: John Coonrod, Technical Marketing Manager, Rogers Corporation, Advanced Connectivity Solutions and Brian Rautio, Vice President of Operations for Sonnet Software, Inc. http://www.microwavejournal.com/events/1706 17
You can also read