Phase shifters are critical elements for electronically scanned phased array antennas, and typically represent a significant amount of the cost of producing an antenna array. Phase shifters are the devices in an electronically scanned array that allow the antenna beam to be steered in the desired direction without physically re-positioning the antenna.

It is not uncommon for the cost of the phase shifters to represent nearly half of the cost of the entire electronically scanned array. This excessive cost has limited the deployment of electronically scanned antennas and restricted their use to expensive and specialized systems, such as fighter aircraft radar and certain commercial systems such as cellular telephone base stations.

There is significant demand in the wireless and microwave industries for affordable phase shifters that can reduce the cost of an electronically scanned antenna system and allow them to be deployed more widely.

Additionally, phase shifters provide an elegant way of linearizing amplifiers for such applications as cellular base stations. This paper will provide further detail on why this is so and explore avenues for lowering the cost of phase shifters.

Electronically scanned antennas have broad applicability for both commercial and military applications, including advanced military radars, cellular base stations, satellite communications, and automotive anti-collision radar.

Allowing an antenna to electronically scan has many benefits, including fast scanning, the ability to host multiple antenna beams on the same array, eliminating mechanical complexity and reliability issues, the ability to angle the antenna in such a way that it reduces radar cross section for stealth reasons and the ability to operate over a wider frequency range.

Phased array antennas typically consist of individual elements that are essentially each little antenna on their own as shown in Figure 1.

Numbers of these elements are arrayed in patterns depending on the desired performance characteristics needed by the application, such as operating frequencies, antenna gain, sensitivity, and power requirements.

Each element typically requires a phase shifter that applies a phase shift to that element based upon the desired beam scanning angle commanded by the system. Figure 2 depicts how different phase shifts are applied to elements across an array to electronically steer the beam.

In Figure 2, the phase shifters apply a different phase shift to each element to steer the beam. The shifted phase of each element causes the waves to interfere constructively, giving the maximum gain in the desired direction. By simply commanding different phase shifts to the various elements across an array, the beam can be steered very rapidly as opposed to mechanically scanned antennas. Additionally, depending on the antenna design, multiple beams can be formed and controlled simultaneously from a single array.

Phased array antenna systems have potentially many applications that could be realized if it is possible to reduce the cost of manufacture. With the high cost of the phase shifters themselves being such a significant portion of the overall antenna cost, any substantial reduction in the cost of phase shifters will allow much wider deployment of phased array antenna systems.

When a signal is output from an amplifier, there are certain undesired mixing products that are generated by the amplification process that introduce noise into the system. It is highly desirable to filter the unwanted signals shown in Figure 3 out of the system and leave a cleanly amplified signal.

Phase shifters provide an interesting and elegant way to perform this function. Because phase shifters possess the ability to change the phase of a signal, it is possible to sample the undesired harmonics and change their phase by 180 degrees.

Signals that are 180 degrees out of phase cancel each other out, as shown in Figure 4.

By using a feed-forward technique, the out of phase signal is coupled into the system, causing the undesired harmonics to be easily cancelled out of the circuit. This technique has been utilized for years, but found limited use due to the high expense of the phase shifters required to perform the task.

A lower cost phase shifter has the potential to allow many more systems to take advantage of this technique and dramatically and cost-effectively improve the performance of their systems.

There are a number of approaches utilized to design and produce phase shifters depending upon the application. This article will focus on solutions that enable arrays to be developed that can employ standard commercial semiconductor manufacturing processes. By using standard commercial practices a manufacturer stands the best chance of using the economies of scale to produce an affordable product.

There are basically two general methods of producing phase shifters: mechanical shifting and electronic shifting. Larger mechanical devices are expensive and can't take advantage of the economies of scale needed to make affordable phase shifters. Therefore, the focus of this discussion will be on electronic phase shifters.

Electronic phase shifters manufactured using industry standard semiconductor manufacturing processes come in basically three flavors:

• Gallium-Arsenide-based digital phase shifters;
• Micro-electrical mechanical system (MEMS) devices; and
• Ferroelectric-based analog phase shifters.

Each of these technologies has their strengths and weaknesses. This article explores the trade-offs of each technology and how to arrive at the most cost effective solution to get the best performing phase shifters for the money.

Gallium-Arsenide (GaAs)-based devices use an approach that employs a number of switches with differing delay values to achieve an aggregate phase shift.

Each of these switches is biased either on or off, giving it digital nature. A typical GaAs based phase shifter will employ a five-bit control of the phase shift, which yields discrete control of the phase shift in 11.75 degree increments. A six-bit control would yield 5.875 degree phase control. Figure 5 shows how a generalized digitally controlled phase shifter works.

Based on which bits are activated, the total phase shift is additive. For example, if the 180 degree, 90 degree and 22.5 degree bits were turned on, the device would supply a signal with roughly a 292.5 degree phase shift.

Depending on the application, digital phase shifting may present undesirable performance limitations if the resolution is inadequate. These GaAs phase shifters are typically very small, on the order of a few square millimeters, which make them a good candidate for thin-film semiconductor manufacturing processes. GaAs, on the other hand, is one of the most expensive semiconductor technologies to manufacture and generally requires special packaging during manufacture. Table 1 details the key strengths and weaknesses of the GaAs phase shifter.

MEMS technology has received a lot of attention in recent years due to the promise of extremely low loss components and lack of undesirable intermodulation distortion.

MEMS are essentially extremely small mechanical devices that can be manufactured using semiconductor manufacturing techniques. In the case of a phase shifter, these devices would physically reshape the circuit to provide the desired phase change.

In MEMS-based phase shifters, the phase shift is typically provided by using a delay line technique. The amount of phase shift is controlled by the amount of delay provided in a circuit. This delay is created by using MEMS capacitors, which change their capacitance value based upon the physical movement of the miniature parallel plates that form the capacitors.

Capacitance is produced by charging parallel plates, and the amount of capacitance is a function of the parallel plate's area and distance apart. By providing more capacitance in a delay line, the amount of delay is increased, and, hence the phase shift, as shown in:

Δø=ω√L(C+C_d)

Where Δø is the phase change, ω is the frequency, L is the inductance, C is the fixed capacitance and C_d is the variable capacitance provided by the MEMS capacitors. By this equation, it is seen that the phase change is controlled by variable capacitance. Figure 6 shows a depiction of a MEMS phase shifter.

One of the great attractions of MEMS devices is they are expected to be very low loss devices since the moving plates in the capacitors should deliver very high quality capacitors.

In practice, they do demonstrate some excellent loss characteristics, but the reliability of these miniature mechanical devices has proven to be very low. Significant effort is being expended to improve the reliability of MEMS devices in general and it is possible that they will mature in the years ahead, but currently there are not any commercially viable devices on the market. Table 2 summarizes the strengths and weaknesses of current MEMS phase shifters.

Ferroelectric materials are ceramic compounds with unique dielectric properties and a very high capacitance density. In other words, large value capacitors can be constructed in a very small physical area.

Bulk ferroelectric, or thick-film, materials have been used for many years for such things as large industrial capacitors. Some work has been performed using thick-film ferroelectrics for phase shifters in a ferroelectric lens configuration, but that has yet to yield a commercially viable product and they have the unfortunate property of requiring very large control voltages, on the order of thousands of volts, to achieve large phase changes.

Thick-film ferroelectrics also have highly non-linear behavior with temperature change that makes them difficult to precisely control. Figure 7 compares thick and thin-film ferroelectric materials relative to temperature change.

From Figure 7, it is seen that thin-film ferroelectric material has very different properties from bulk or thick film material, including a very flat temperature response profile, giving thin-film ferroelectric material good controllability over wide temperature ranges.

Thin-film ferroelectric material does maintain a very high capacitance density, which makes it a good material for many applications. The difficult challenge of thin-film ferroelectrics comes in controlling the growth or deposition of the material during manufacture. If good material is grown on wafers, it is possible to construct high performance phase shifters.

Unlike MEMS, which requires the physical mechanical movement of the capacitor plates, thin-film ferroelectrics require only a voltage change to adjust the capacitance. The capacitance changes simply by the movement of atoms within the material making a capacitance change nearly instantaneous. Figure 8 shows how the capacitance changes with voltage in the thin-film ferroelectric material Barium Strontium Titanate (BST).

This unique property of changing the capacitance with applied voltage gives the material the ability to tune the capacitance over a fairly wide range, at least a 2:1 change.

Because the material has an extremely high capacitance density, very small tunable capacitors can be made, which allows many thousands of them to be constructed on a single wafer.

These tiny tunable capacitors can also be used in a voltage controlled delay line to make phase shifters in the same manner as MEMS capacitors. By simply applying a control voltage to the phase shifter, the amount of phase shift is controlled.

BST also has an interesting property that results in increasing capacitor Q with increasing control voltage. Capacitor Q is a measure of a capacitors quality and is essentially a ratio of energy stored to energy wasted.

Since the Q increases with tuning voltage, the loss through the phase shifter decreases with increasing phase shift. The result is the loss at 360 degrees of phase shift is less than zero degrees of phase shift, generally the loss decreases by about half.

This property results in improved power density of the scanned antenna beam at large scan angles off boresight, and improved power density results in improved sensitivity of the antenna at large scan angles. Figure 9 shows a phase shifter that was constructed using BST capacitors.

Figure 9 is an actual phase shifter using a delay line configuration with a number of BST capacitors providing the delay. The size of this phase shifter is only a few square millimeters. Table 3 shows the strengths and weaknesses of thin-film ferroelectric phase shifters.

The electronically scanned phased array antenna market is in desperate need of affordable phase shifters, as is the cellular base station amplifier market. Nearly half the cost of an array can easily be consumed by phase shifters.

Using commercial semiconductor processes is the key to reducing the cost, since many thousands of phase shifters can be manufactured on a single wafer, which brings their cost down significantly.

Gallium Arsenide is the most mature technology, yet is more expensive. Newer technologies, such as MEMS and thin-film ferroelectrics, have the potential to dramatically change the market as they mature. Significant advances in thin-film-ferroelectric-based phase shifters have been made in recent years, and thin-film ferroelectrics is very simple by nature and, thus, a less risky proposition than MEMS.

Tom Watson is the director of business development for Agile Materials and Technologies Inc. (www.agilematerials.com). He has over 20 years of military and commercial management experience, including nearly 3,000 operational and flight test hours in F-14 and F/A-18 tactical jet aircraft. Watson holds a bachelor of science from the U.S. Naval Academy, as well as a master of science in aeronautical engineering and a master of science in systems management. He is also a graduate of the U.S. Naval Test Pilot School. Watson can be reached at twatson@agilematerials.com.

RSS    Save to Del.icio.us  Digg This