Embedded digital-capacitor ICs enable antenna tuning

Femto/picofarad antenna-tuning ICs enable smartphones to work over several bands and scenarios, thus solving the challenging impedance-matching dilemma.

One of the indicators of innovative engineering is to take an attribute that is normally considered to be a drawback and not only negate that characteristic but actually use it to advantage in another setting.

Consider this example of noise. Additive white Gaussian noise (AWGN) – is an ever-present concern in many circuits, especially those for low-level analog signal chains. Yet that same noise can be used to increase the apparent resolution of an analog/digital converter beyond its physical rating. This is done by taking repeated samples (oversampling) of a static or slowly changing signal with noise deliberately added (sometimes called dithering) and then averaging the results.

The tradeoff is the need to take more samples over a period, thus slowing the speed at which new, fresh data can be acquired. Similarly, mechanical systems often add a small amount of mechanical dither or jitter to “keep things moving” and ensure that stiction (static friction) does not impede the start-up motion of a valve or mechanical readout.

There’s another circuit attribute that often causes engineers headaches: parasitics. These are the tiny amounts of unavoidable stray capacitance and inductance due to lead length, packaging, and other physical factors. While they are generally of no concern at lower frequencies (there are some definite exceptions) they become significant at high megahertz and gigahertz frequencies. Modeling them is difficult but essential to properly characterize components or circuits as they negatively affect performance.

Now, digitally controlled and calibrated parasitic capacitances are being used for antenna tuning. The need for such tuning – either by adjusting the antenna itself or via a matching circuit between the power amplifier output and the antenna – is almost as old as wireless itself. Even in those early days, electromagnetic theory and hands-on practical experimentation showed that effective power transfer and optimal antenna performance require that the source impedance and the load impedance be complex conjugates. The voltage standing-wave ratio (VSWR) should ideally be at unity (1:1), but in most applications a VSWR between 2:1 and 1:1 is acceptable.

Figure 1. This variable L-network random wire antenna tuner is designed to allow manual matching of the low output impedance of a transmitter (up to 200 watts) to the high impedance of a random wire over the operating range of 2 to 30 MHz. (Image source: MFJ Enterprises)

The problem has not gone away but instead has morphed into a new and more challenging form. Traditionally, antenna-matching circuits were built into the radio of smaller, lower-power designs; in other cases, they were and still are offered as commercial units in external enclosures. This was due to the high-power ratings needed spanning tens and hundreds. or even thousands of watts, along with the physically larger values needed at the lower frequencies of tens or several hundred megahertz.

Some of these external tuners were designed for a single band, while others for multiband use such as in amateur (ham) radio had front-panel switchers to enable adjustment settings for the different bands in use (Figure 1).

Many of them are one-off, hand-crafted works combining artful form and needed function (Figure 2).

Figure 2. Many amateur radio enthusiasts prefer to design and fabricate their own antenna-matching units for their bands of interest and power levels, such as this one covering 3 to 30 MHz and handling up to 150 watts; note the toroidal transformer with multiple windings (Image: PD7MMA, “Ham Stuff Blog”).

Newer antenna tuners incorporate autonomous self-controlled autotuning using an internal processor or allow an external PC to do so via a USB port.

In other cases, designers and radio amateurs would use a short length of twisted wires, often just a centimeter or two long, to create what was called a gimmick capacitor. This zero-cost tuning device could provide a few picofarads of capacitance, with the actual amount being a function of the wire gauge, insulation thickness, twist tightness, and length. The advantage of a gimmick capacitor is that you can start with a longer length and snip it down until you reach the optimum value.

New applications, new approaches
Of course, design requirements have changed dramatically. Now the tuning battle is over 5G and even older 4G phones supporting multiple bands and embedded antennas such as the widely used planar inverted-F antenna (PIFA). These smartphones are relatively low-power devices operating above a gigahertz, with multiple bands that must be supported via seamless band transitions and handoffs. The associated LC antenna-tuning values are small, which simplifies the challenge in some ways but also makes it harder in other ways.

Complicating the situation, the matching values are not static but are dynamic in normal use, as the user’s hand changes location and the angle and the phone’s position move with respect to the head and body. Certainly, expecting the user to tune and optimize the antenna-match circuit in use is simply not an option.

Fortunately, there are now solutions to this dilemma via antenna-tuner ICs. These address the issue of allowing digital setting of up to 16 capacitance values, thus changing the electrical characteristics just enough to optimize the match (or get close enough). Among the vendors are Peregrine Semiconductor (PE64909), Qorvo (QM13025), Skyworks Solutions (SKY59272-707LF), and Infineon (BGSC2341ML10).

Unlike lower-frequency matching circuits with capacitance in the tens of picofarads and even extending to the microfarads range, these ICs allow tweaking of very small capacitance shifts. For example, the Peregrine Semiconductor PE64909 is a digitally tunable capacitor for 100-3000 MHz (Figure 3).

Figure 3. The PE64909 antenna-tuner IC has a simple function and schematic (left) but its equivalent-circuit model is more complicated (right) (Image: Peregrine Semiconductor).

In operation, a system processor can use a four-bit code to select one of 16 capacitance values via its 3-wire (SPI compatible) serial interface, from 0.6 pF to 2.35 pF (that’s a 3.9:1 tuning ratio) in discrete steps of 117 femtofarad (fF). That’s clearly a modest dynamic range and a small step size, but that’s enough for the application.

Qorvo notes that there are two ways to use capacitance to adjust the antenna appearance (Figure 4):

Figure 4. Antenna tuning can be accomplished via aperture tuning or impedance tuning topologies, each with distinct tradeoffs in attributes and capabilities (Image: Qorvo)

Aperture tuning optimizes the total antenna efficiency from the antenna terminal’s free space, and it can do so across multiple bands. It can provide advantages with respect to antenna efficiency for both transmit and receive communications, improving total radiated power (TRP) and total isotropic sensitivity (TIS) by 3 dB or more in some situations.

Impedance tuning maximizes power transfer between the RF front end and the antenna, and it increases the TRP and TIS by minimizing mismatch loss between the antenna and the antenna front end. It also helps to compensate for environmental effects such as a person’s hand position on a smartphone.

According to Qorvo, “Today, aperture tuning is the primary method used in handsets to overcome reduced antenna area and efficiency. Mid-tier and higher-end smartphones use a combination of aperture and impedance tuning to support the ever-broadening range of frequency bands — especially for 5G.”

These ICs are somewhat analogous to the widely used digital potentiometers (digipots), except those components typically have 256 or more steps spread over a fairly wide kilohm range, along with a much larger relative step size. Perhaps commercially available digi-inductors will be coming soon as well.

Beyond antenna tuning, it’s likely that creative engineers are already looking at these parts and finding unforeseen uses for them. Historically, that’s been the reality as components that originally targeted one class of situations are soon adopted and adapted to address other problems.

Perhaps these tunable picofarad capacitors will be used to compensate for or cancel circuit parasitics in a balanced or differential topology. Or perhaps they will be useful for precise calibration and measurement in some Wheatstone-bridge type of arrangement? After all, you never know what path innovation in circuit topology will take.