Addressing five common RF filtering challenges for mmWave applications

Selecting the right millimeter-wave (mmWave) filter technology is vital for RF designers developing applications ranging from mainstream 5G wireless communication equipment to mission-critical military devices. Because mmWave applications operate at much higher frequencies than their predecessors, the physics is quite different. Therefore, a lot of filtering options available at lower frequencies, such as lumped element filters and metal wave guides, are not feasible for higher frequencies.

While filtering at mmWave frequencies is just as important as it is at lower frequencies, it can be much more difficult. There are new challenges to consider, ranging from the availability of applicable filtering technologies to the unique characteristics of mmWave systems including physical size, manufacturing tolerances, and temperature stability. Let’s explore how to tackle five of the most common filtering challenges for mmWave applications.

1. Finding filter technology that works across all FR2 bands

Innovation in 5G encompasses a wide variety of technologies and use cases, which requires a range of frequencies to be deployed. According to the 3GPP TS 38.101 table in Fig. 1, mmWave frequencies from 24.25 GHz to 52.6 GHz operate in bands that are part of Frequency Range 2 (FR2). While there is not a single filter that can operate across all these frequencies, it is ideal to identify one filtering technology that can be used across this entire spectrum.

Fig. 1: The different bands used at various mmWave frequencies.

To determine potential FR2 filter options, one of the first factors to evaluate is if the filter should be on-chip or off-chip. Because of the smaller dimensions of mmWave devices, high isolation and low loss are challenges for on-chip filtering. Therefore, to avoid a reduction in Q when decreasing size, an off-chip bandpass filter is best. Through extensive research comparing a variety of filtering technologies, Knowles Precision Devices found that planar thin-film implementations are the most desirable approach for mmWave applications from the standpoint of size, cost, and performance.

2. Determining how “good” your filter really needs to be

If the majority of your driving is commuting a short distance to and from the office, it likely doesn’t make sense to spend the money for a high-performance vehicle like a Ferrari when a Ford Focus would get the job done at a fraction of the cost. The same is true when it comes to selecting an RF filter. It’s best to think about the level of rejection really required given the application, or, in other words, how “good” the filter’s performance needs to be.

For example, let’s look at selectivity, or the loss through a filter that occurs at some specified distance from the center frequency. Selectivity is crucial in environments where adjacent channels are close together, as high selectivity enables designers to make good use of available bandwidth. Achieving highly selective filters is simplified if the filter technology is inherently high Q — this can have the benefit of reducing the complexity of the filter design required to reach a selectivity target.

3. Working with strict size constraints

mmWave systems are smaller than their lower-frequency counterparts. Thus, component size is a key enabling or restricting factor. In a traditional antenna-array system, an inter-element spacing of less than half the wavelength (λ/2) is required to avoid the generation of grating lobes. This principle holds true in a 5G beamforming antenna in which, for example, a 28-GHz-band antenna needs approximately 5 mm of inter-element spacing. Therefore, the resulting compact arrays need to have a way to integrate the necessary filtering.

Depending on where RF filtering is deployed in the architecture, space will come at a premium. After comparing the size of a variety of common bandpass filters, Knowles Precision Devices engineers determined surface-mount technology (SMT) to be a great choice for working with the space constraints presented by mmWave applications. Furthermore, among SMT options, a microstrip approach can reduce filter space while maintaining the required levels of performance in terms of bandwidth, rejection, and insertion loss. However, because not all microstrip filters are created equal, considerations such as choice of substrate, plating technology, and topology can also dramatically change the size of the filter. 

4. Impact of tolerance on overall costs and performance

Considering the importance of miniaturizing mmWave filters, especially in applications like small cells, manufacturing tolerance plays a crucial role. Tolerance not only affects filter specifications such as planned versus realized performance and potential loss of bandwidth, it also impacts the cost of implementation, especially if one considers the cost of rejecting out-of-compliance boards.

Plus, as discussed, filter size must be closely monitored. Poor tolerance encroaches on potential board space or layers that could be used for adding other devices or functionality. Using a fully integrated design featuring thin-film technology and a high-permittivity dielectric lets RF designers shrink the overall size and integrate the resistor, reducing variation from resistor tolerances and improving overall RF performance. Additionally, SMT devices generally provide better shipped tolerances compared with a PCB approach if the impact of the SMT filter’s temperature stability on bandwidth capacity is considered.

In short, at mmWave, tolerance impacts can become significant and potentially alter the total cost of an implementation. If tolerance isn’t considered in manufacture, it can impact the yields of the overall system and further increase the need for guard banding, taking up useful spectrum space.

5. Temperature stability issues

With mmWave applications, there can be a variety of issues related to fluctuating operating temperatures. This ranges from needing mmWave antenna arrays to operate reliably in extremely cold or hot outdoor conditions to thermal complexities in smartphones to general heat dissipation issues from miniaturized, crowded circuitry. In these densely packed systems, there is no way to control temperature, which means frequent variations may occur and these systems will run hot. Therefore, filters must have the ability to perform within specification over a wide range of temperatures, with a good temperature stability of approximately 3 ppm/°C.

A key component of temperature stability in microstrip filters is the selection of substrate material. Take, for example, the comparison in Fig. 2 between two 18-GHz bandpass filter designs, one manufactured on Knowles Precision Devices’ “CF” substrate and the other on an alumina board with the filter response measured from –55°C to 125°C.

Fig. 2: Graph A (left) shows the response of microstrip bandpass filters built on alumina, and Graph B (right) shows the response of microstrip bandpass filters built on a CF dielectric. (Click chart to open larger version.)

At the 35-dB rejection point, the alumina-based filter shifts by 140 MHz, whereas the CF dielectric filter shifts only 17 MHz. By designing with the right dielectric material and filter topology, temperature-stable SMT filters with high rejection and low loss can be produced.

Effectively evaluating your RF filter needs

As radio architectures evolve, the need for filters is also evolving. And as the industry simultaneously miniaturizes mmWave devices and works to minimize costs, designers need filter solutions that will keep prices manageable. Therefore, filter companies are helping RF designers overcome the unique challenges of mmWave by developing filters that are smaller are more temperature-stable, and allow for more design flexibility.