ROG Blog

The Rog Blog is contributed by John Coonrod and various other experts from Rogers Corporation, providing technical advice and information about RF/microwave materials.

Ten Years of Blogging on Microwave Journal’s Website

10 Years of the ROG Blog and It May Just Be the Start

August 3, 2020

For 10 years, the ROG blog has been serving the RF/microwave community with the aid of the Microwave Journal and we are happy to celebrate this milestone.  We have written blogs on many different circuit-material-related topics and, admittedly, have covered the same topic a few times. This can pose a challenge to give more information on the more detailed topics, but we hope we have helped some readers through our efforts.   Not only did our blogs address RF/microwave issues, but also many blogs were related to Printed Circuit Board (PCB) processing and how constructing a circuit can affect RF/microwave performance.   All of these blogs for the past 10 years are archived on the Microwave Journal website and can be read at your leisure.

Some of our first blogs were related to how to choose materials for different RF/microwave applications.  We also discussed the transition from using FR-4 materials to working with the higher performance provided by higher quality, high frequency materials. Over the past 10 years, one factor that has changed remarkably is the number of millimeter-wave (mmWave) applications.  Ten years ago, there were very few PCB applications operating at mmWave frequencies. At present, it is not unusual to see PCB manufacturers producing circuits in high volume quantities for mmWave applications.   The high volume circuit demands for mmWave applications may be for 77 GHz automotive radar as part of advanced driver assistance system (ADAS) equipment, 60 GHz backhaul communications links, and for several mmWave applications for 5G systems.

The dramatically increased interest in mmWaves is because of all the bandwidth those frequencies offer. One of the major hurdles that allowed more mmWave applications to be supported was the breakthrough in semiconductor technology at the higher frequencies needed to support these applications.  After the chips were defined, the PCB industry set about to make materials and circuits for mmWave applications. Circuits with the small dimensions needed for mmWave frequencies are incredibly sensitive to small anomalies which could go undetected in the past for lower frequency circuits.   PCB manufacturers had a steep learning curve for mmWave requirements and they responded very well. 

Along with the achievements of chip and PCB manufacturers reaching mmWave frequencies were the many efforts of high frequency circuit material suppliers.  Even though these material suppliers have dabbled with mmWave circuitry over the years, many of these applications were very specialized and low in volume. Evolving mmWave applications have far more demanding specifications and a drive for data at these high frequencies, which has been previously unseen. Not to mention the need to adjust technical support to deal with these complex applications in high volume manufacturing.

Rogers Corporation has dealt with mmWave applications for a few decades prior to the recent boom in PCB based mmWave applications.  Despite our strong technology background, however, we have had to make many adjustments to properly serve these emerging mmWave applications.  One major adjustment was acquiring test equipment capable of testing circuits and PCBs at mmWave frequencies.  Obtaining the equipment was only the tip of the iceberg in understanding how to make accurate measurements and generate meaningful data at mmWave frequencies.  It was not as if those frequencies were new territories for us: we previously had multiple vector network analyzers (VNAs) capable of testing to 50 GHz, and we had experience and knowledge of measurement issues through that frequency. But when trying to test to 77 GHz and higher, there were many more things to consider than we expected.

Another part of our mmWave learning curve was adapting and adjusting our technical support.  Our technical support has always been split into two categories: one for support of the RF designer and one for support of the PCB fabricator.  RF design support obviously had to change to address the many concerns related to mmWave applications. But there was a good amount of learning that had to take place on the PCB fabrication support.  Since mmWave applications operate at much smaller wavelengths than lower frequency applications, they are typically much more sensitive to the normal fabrication variations which take place in manufacturing a PCB.  Understanding etching conditions to have tighter control of circuit features is mandatory; however, there are many other things to consider.  The normal variations of circuit dimensions such as hole-to-pad and layer-to-layer alignment and spacing are much more critical to understand at mmWave frequencies than at RF/microwave frequencies.  Not to mention the fact that grounding viaholes must be on a much tighter pitch to account for the mmWave performance concerns; with that tighter pitch comes some drilling concerns when working with different high frequency materials.   Additionally, it was found that some final plated finishes that have been acceptable in the past for RF/microwave applications have been found inadequate for mmWave circuits where the final plated finish can play a more significant role in achieving final performance target goals.

Copper surface roughness is another issue that Rogers Corporation has explored for decades; the material characteristic is far more important at mmWave frequencies than for lower frequency applications.  The copper surface roughness of concern is at the substrate-copper interface of a laminate.  We have known for many years the impact that the copper surface roughness has on transmission-line insertion loss and phase response; however, we were less mindful of the impact of normal surface roughness variations.  All copper types exhibit some variation of surface roughness.  A typical electro-deposited (ED) copper may have an average surface roughness of 2 μm RMS, but if many sheets of the copper are measured it is not uncommon to see the roughness vary from 1.7 to 2.3 μm RMS.   This roughness variation can cause much more difference in electrical performance at mmWave frequencies than at lower frequencies and Rogers Corporation has put a lot of effort into understanding this issue and minimizing any adverse effects on high frequency performance.

Ten years ago, when we started the ROG Blog with the help of Microwave Journal’s dedicated editorial staff, it might have seemed impossible to keep a regularly scheduled blog on circuit materials going for even a few years, let alone for 10, without undue repetition. But the continued interest in the ROG Blog from our readers—thank you, folks—and the ever-improving quality of the circuit materials that we are writing about has “fueled the fire” and given us much to write about. Knowing readers are interested in knowing more about circuit materials has made that 10 years seem to fly right by. Whether the same is true for the next 10 years, time will tell. But with so many applications reaching mmWave frequencies, from ADAS to 5G, don’t be surprised if we are still at it 10 years from now. High quality circuit materials are vital for these high frequencies, and we at Rogers Corp. will be with you as we reach further into the future.

Do you have a design or fabrication question? Rogers Corporation’s experts are available to help. Log in to the Rogers Technology Support Hub and “Ask an Engineer” today.