[This knowledge is not too cold] Exploring RF filter technology (Part 1)

okhxyyo

[This knowledge is not too cold] Exploring RF filter technology (Part 1) [Copy link]

In the past decade, the rapid growth of mobile wireless data has made operators more and more urgent to use new frequency bands and technologies to meet users' demand for wireless data capacity. This demand has not only promoted the development of wireless technology, but also increased the demand for enhanced radio frequency (RF) filter technology to help reduce system interference, expand RF coverage, enhance receiver performance, and improve coexistence characteristics.

This article will explain how RF filters work and the various versions of the technology used in applications today. It will begin with some basic facts about filters and the benefits they bring.

Basic knowledge about RF filters

Let's first understand these important filter terms and concepts.

Attenuation: The loss in amplitude of a signal after it passes through an RF filter, usually measured in decibels (dB). It is also called rejection when referenced outside the filter passband.

Cutoff: The point at which the filter response decreases by 3dB.

Insertion Loss: The loss of signal power in the filter’s passband of interest.

Isolation: Separating signals from each other to prevent unwanted interactions between them. For example, you might isolate a transmit signal from a receive signal to prevent them from interacting.

Passband: The region where signals pass with little attenuation. Q-factor: Short for quality factor, it is the ratio between the energy stored and the energy lost per oscillation cycle. It is used to measure the selectivity of a resonant circuit.

Ripple: The difference in insertion loss in the passband.

Selectivity: A measure of a filter's ability to pass or reject specific frequencies relative to its center frequency. Selectivity is generally defined as the loss that occurs at nodes that differ by a specified frequency from the filter's center frequency.

Stopband: The frequency band that a filter operates within to achieve the desired out-of-band rejection (expressed in decibels required).

Figure 1: Typical RF filter response

Filters remove unwanted frequency components from a signal while maintaining the desired frequency bandwidth. Figure 2 shows four basic filters that accept or reject signals in different ways:

Low-pass filter: allows all frequencies below a certain frequency to pass, and blocks all other frequencies (opposite of high-pass)

High pass filter: allows all frequencies above a certain frequency to pass, and blocks all other frequencies (opposite of low pass)

Bandpass filter: allows all frequencies between two frequencies to pass and blocks all other frequencies (as opposed to bandstop)

Band Rejection Filter (also called Band Stop or Notch Filter): Blocks all frequencies between two frequencies and allows all other frequencies to pass (as opposed to Band Pass)

Figure 2: Basic RF filter and response

Comparison of filter technologies

The filter structure varies depending on the application. The most common differences in RF filter technology are size, cost, and performance. The structure of the filter is the main factor that causes these differences. Here are some examples of RF filter structures:

Acoustic Filter: A filter that can cater to both low and high frequencies (up to 9 GHz) and in some special cases up to 12 GHz. It is small and offers an excellent combination of performance and cost to meet complex filter requirements. Acoustic filters are the most common filter structure in commercial RF microwave applications such as cell phones, WiFi, and Global Positioning System GPS.

Cavity filter: A type of filter used only in infrastructure applications. It provides good performance at a reasonable cost but is larger than an acoustic filter.

Discrete Inductor-Capacitor (LC) Filter: A low-cost filter with medium performance and size. LC components are sometimes integrated on a substrate in the form of a printed structure, called an integrated passive device (IPD). LC filters can also be realized with discrete surface-mount device (SMD) components.

Monolithic Ceramic Filter: A filter that has higher cost and performance than multilayer ceramic filters. It is also larger and not suitable for mobile applications.

Multilayer Ceramic Filter: A medium to low cost filter with performance comparable to LC filters. They usually have a more reasonable footprint, but are thicker, making them unusable in some applications.

Filters can be designed to meet a variety of requirements. Although they use the same basic circuit configuration, the circuit values will be different when the circuit is designed to meet different criteria. For example, different circuit values will result when the criteria such as in-band ripple, fastest transition to final roll-off, highest out-of-band rejection, etc. need to be met.

Understanding Piezoelectric Acoustic Filters

For many applications today, the filter technology of choice is piezoelectric filters. These RF filters are compact, cost-effective solutions used in many commercial, military, and scientific applications.

The piezoelectric effect is a reversible physical phenomenon. A crystalline substance generates an electric current when subjected to mechanical stress and vice versa. When an electric field or voltage is applied, the crystal stretches slightly. Piezoelectric materials convert applied mechanical stress into electrical energy and vice versa.

There are two types of acoustic filters available on the market: surface acoustic wave (SAW) and bulk acoustic wave (BAW).

As shown in Figure 3, SAW and BAW filters can be divided into two categories: ladder and grid. Ladder filters have high rejection near the passband, but poor out-of-band rejection. Grid filters provide good out-of-band rejection and lower rejection near the passband. The hybrid ladder-grid configuration offers a compromise between rejection and passband rejection performance.

Figure 3: SAW and BAW configuration design

>>>>>>>>Click here to view the full original text

okhxyyo

SAW Filters

SAW filters are widely used and very mature. They propagate high frequency waves laterally. They convert electrical input signals into acoustic waves using interdigitated metal interdigital transducers (IDTs) created on a piezoelectric substrate such as quartz, lithium tantalate (LiTaO3), or lithium niobate (LiNbO3), as shown in Figure 4. SAW filters are slow and can be used to accommodate many wavelengths passing through the IDT in a very small device.

Figure 4: Basic SAW RF filter

A key advantage of SAW is that it can be optimized for standard filter applications in the 600-1900MHz range. It not only meets the low 5G band of 600MHz, but also performs well and cost-effectively in the 1, 5, 8, 13 low mobile phone bands and GPS bands.

To reach above 1900 MHz, we use layered SAW technologies such as low-loss resonator technology (LRT), thin-film SAW (TF-SAW), ultra-SAW, or extremely high performance (IHP) SAW. These technologies enable multilayer SAW to reach 2.7 GHz, which can be used for mid- and high-band applications. Above 2.7 GHz, the selectivity of SAW filters decreases. However, at the end of the usefulness of SAW filters, BAW filters (described later in this chapter) come into play.

Temperature Compensated SAW

Among the many SAW technology innovations is temperature compensated SAW (TC-SAW). TC-SAW uses a thermal compensation layer to avoid thermal drift, as shown in Figure 5. Its performance is better than standard SAW due to its higher temperature stability.

Compared with traditional SAW-based filters and duplexers, TC-SAW devices are integrated on single-crystal lithium niobate and the IDT electrodes are buried in the silicon dioxide (SiO2) layer, which improves the TCF and quality factor.

Figure 5: Comparison between SAW and TC-SAW

To use a temperature coefficient on a SAW, the interdigital transducer (IDT) needs to be isolated from any ambient temperature changes. To achieve this isolation, the SAW structure needs to be covered with a SiO2 layer and sometimes an additional silicon nitride layer, as shown on the right side of Figure 5.

TC-SAW improves filter performance to cope with the stringent frequency band allocations for LTE and 5G, as well as carrier aggregation (CA) for uplink and high-power user equipment (HPUE). These improvements include lower insertion loss, steeper skirt characteristics, temperature stability, higher power durability, wider bandwidth frequency range, higher linearity, smaller size and integration. To achieve these characteristics, a higher Q factor and a smaller temperature coefficient of frequency (TCF) are required.

>>>>>>>>Click here to view the full original text

okhxyyo

Multilayer SAW

In a conventional SAW substrate structure, the acoustic wave propagates along the surface, reaching several wavelengths under the electrode, while the bulk acoustic wave radiation mode appears deep in the substrate. This causes bulk acoustic wave leakage, which reduces the Q factor, TCF and coupling.

One way to reduce this leakage is to create a new multilayer structure or a variation of TC-SAW, as shown in Figure 6. Adding this extra SiO2 layer below the surface of the piezoelectric and functional layers provides an important benefit: it draws the SAW propagating at the surface into the piezoelectric and functional layers. This confines the acoustic energy near the surface, improving TCF and Q factor.

Figure 6: Comparison between multilayer SAW and SAW

There are several types of SAW filters on the market. They are called SAW, TC-SAW and multilayer SAW, also known as guided SAW, LRT-SAW, TF-SAW, ultra-SAW and IHP-SAW. They all belong to the SAW variety, with an additional layer such as SiO2 between the substrate and the functional layer, as shown in Figure 6. In this book, we will refer to this type as SAW.

Multilayer SAW filters offer higher performance than standard SAW filters. In some frequency bands, multilayer SAW filters can perform as well as BAW filters, as we will discuss next.

Multilayer SAW technology is based on piezoelectric material distributed on an isolator. In multilayer SAW, the thermal compensation mechanism comes directly from the substrate and not from the electrode surface.

The table below compares the different features of various acoustic SAW RF filter technologies.

>>>>>>>>Click here to view the full original text

okhxyyo

BAW Filters

BAW can be used to replace SAW. BAW filters support the frequency range of 1.4 to 9 GHz.

Besides the structure, the main difference between BAW and SAW is the piezoelectric material the filter is based on. BAW filters use polycrystalline aluminum nitride, which has a higher coupling coefficient and higher wave speed.

BAW filters have very high acoustic energy density and their structure captures acoustic waves very well, so they can achieve very low losses. In addition, they can achieve higher Q factors than other types of appropriately sized filters used at microwave frequencies. This allows for excellent rejection and insertion loss performance, also at the critical passband edges.

BAW filters offer lower insertion loss, helping to extend battery life in smartphone applications. BAW is also suitable for applications with short separation between uplink and downlink, and for applications with high attenuation requirements in closely spaced adjacent frequency bands.

Unlike SAW filters, the acoustic waves in BAW filters propagate vertically. As shown in Figure 7, the resonator uses a quartz crystal as a substrate. Metal paths on both the top and bottom sides of the quartz excite the acoustic waves, which bounce off the top to the bottom surface, forming standing waves. The thickness of the plate and the quality of the electrodes determine the frequency at which resonance occurs. At the high frequencies at which BAW filters are effective, the piezoelectric layer must be only a few microns thick, which requires the resonator structure to be based on a carrier substrate and constructed using thin film deposition and precision machining.

Figure 7: Basic BAW RF filter

In this type of filter, to prevent the sound waves from escaping into the substrate, thin layers of alternating stiffness and density are stacked to create a Bragg acoustic reflector. The resulting device is called a solid-mounted resonator (SMR) BAW, as shown in Figure 7. The Bragg reflector is made up of alternating layers of materials that have different refractive indices.

There are two main types of BAW filters on the market, as shown in Figure 8: solid mounted resonator (SMR) BAW and polycrystalline film bulk acoustic resonator (FBAR).

Figure 8: Cross-sectional comparison of BAW filter technologies

>>>>>>>>Click here to view the full original text