Some would say that the most important components in an RF system are the mixer, filter, amplifier, transmitter/receiver, and antenna. However, in addition to these components, there are less glamorous devices that play a pivotal role in the successful design of an RF system. Some of these components are used to either help send the signal in a different direction (or in multiple directions) within the system or change the shape of the RF signal. One of these important, yet sometimes taken for granted, parts is the RF switch.
As noted, the RF switch will change the direction of the RF signal. When a switch operates in both directions, it is referred to as bidirectional. Since the switch requires some sort of power supply, it is considered an active device.
There are two important parameters the designer should look at when selecting an RF switch for his/her design: insertion loss and isolation. When looking at a datasheet, the designer wants a switch with the lowest possible insertion loss and the highest possible isolation.
The “insertion loss” specification of a switch module is a measure of power loss in the signal – which can occur if the length of the transmission line it is made to propagate through is greater than 0.01 of its own wavelength — and signal attenuation. Insertion loss of a switch module at a particular frequency can be used to calculate the power loss or voltage attenuation caused by the switch on a signal at that frequency.
Every switch has some parasitic capacitance, inductance, resistance, and conductance. These parasitic components combine to attenuate and degrade the signal that the switch is being used to route. The power loss and voltage attenuation caused by these components varies with the frequency of the input signal, and can be quantified by the insertion loss specification of the switch module at that frequency. As a result, it is very important to ensure that the insertion loss of a switch is acceptable at the bandwidth requirement of the application.
Power loss can be calculated as follows:
Similarly, to calculate voltage attenuation:
Isolation is defined as the magnitude of a signal that gets coupled across an open circuit. Crosstalk (Figure 1) is defined as the magnitude of a signal that is coupled between circuits (such as separate multiplexer banks on an RF module).
Figure 1: Crosstalk versus isolation. (Courtesy of National Instruments.)
Another key parameter is switching speed – a measure of how long it takes for the part to go from one position to the other. In datasheets we usually see the terms from “on” to “off” and from “off” to “on.” Obviously the fastest possible switching speed is what the designer wants for his/her RF designs.
Other parameters a designer might be concerned with are the frequency range, return loss, settling time, power handling, termination, video leakage, and operating life.
There are basically two types of switches used today: electromechanical and solid-state switches. The basic operation of electromechanical switches is based on the simple theory of electromagnetic induction. They rely on mechanical contacts as their switching mechanism.
In electromechanical switches, the control signal causes the contact to physically change positions during the switching process. These parts can handle high power RF signals since they have low insertion loss and high isolation. Even though the parts benefit from these features, the disadvantage with these parts is that they can be bulky, heavy, and slow (providing a switching speed in the millisecond range). We usually see these types of switches used in Industrial, and test and measurement applications.
As the name implies, a solid-state switch is an electronic switching device based on semiconductor technology, so nothing inside this parts moves (unlike electromechanical switches) - this makes them faster, smaller, and lighter. Switching times for these parts are in the nanosecond range. These parts are inferior with regard to insertion loss, DC power consumption, isolation, and power handling (Table 1) compared to electromechanical switches. A switch of this kind is made from diodes (low insertion loss) or transistors (faster switching time).
||< 15 ms
||< 1 µs
||5 million cycles
||RF power overstress
Table 1: Comparison of electromechanical and solid-state switch parameters. (Courtesy of Wikipedia)
A new type of switch available today is the MEMS RF switch. This type promises better properties over the electromechanical and semiconductor switches. MEMS switches offer the high RF performance and low DC power consumption of electromechanical parts, and the small size, weight, and low-cost features of semiconductor parts.
MEMS switches are surface-micromachined devices which use a mechanical movement to achieve a short circuit or an open circuit in the RF transmission line. RF MEMS switches are designed to operate at RF to mm-wave frequencies (0.1 to 100 GHz). The advantages of MEMS switches over PIN diode or FET switches are:
- Near-Zero Power Consumption: Electrostatic actuation requires 30-80 V, but does not consume any current, leading to very low power dissipation (10 to 100 nJ per switching cycles).
- Very High Isolation: RF MEMS metal-contact switches are fabricated with air gaps, and therefore, have very low off-state capacitances (2 to 4 fF) resulting in excellent isolation at 0.1 to 60 GHz.
- Very Low Insertion Loss: RF MEMS metal-contact and capacitive switches have an insertion loss of 0.1 dB up to 100 GHz.
RF MEMS switches also have their disadvantages, and these include:
- Relatively Low Speed: The switching speed of most electrostatic MEMS switches is 2 to 40 µs, and for thermal/magnetic switches, the switching speed is 200 to 3,000 µs. Certain communication and radar systems require much faster switches.
- High Voltage or High Current Drive: Electrostatic MEMS switches require 30 to 80 V for reliable operation, and this requires a voltage up-converter chip when used in portable telecommunication systems.
RF MEMS types of switches can replace the SP2T and SP3T switches (implemented by semiconductor technology) that are used in today’s dual-band and triple-band cell phones. Here the benefit is improvement in RF performance, which will reduce DC power consumption.
These types of switches would also benefit satellite applications, which not only demand high switching performance, but also mass and volume reduction. Another possible use is in beam forming networks, such as in the design of reconfigurable Butler matrices and phase shifters for multi-beam satellite communication systems. Down the road, MEMS switches should become more advantageous as frequency is increased.
Poles and throws
RF switches are categorized by their number of poles and throws. The number of poles is the number of separate circuits controlled by a switch. The number of throws is the number of separate positions that the switch can adopt. For example, the following are some RF switch definitions:
SPDT (single-pole, double-throw) switch routes RF signals from one input port to two selectable output ports.
SPDT terminated switch is a single-pole, double throw switch that has one open output RF port internally terminated in a 50 Ω resistive load.
A multiposition switch has one input and more than two outputs.
A transfer or DPDT (double-pole-double-throw) switch has two independent paths that operate simultaneously in either of two selected positions.
Bypass switches: These types insert or remove a test component from a signal path.
The ADG918/ADG919 from Analog Devices (Figure 2) are wideband switches using a CMOS process to provide high isolation and low insertion loss to 1 GHz. The ADG918 is an absorptive (matched) switch having 50 Ω terminated shunt legs, whereas the ADG919 is a reflective switch. Housed in an 8-lead MSOP/LFCSP package, the parts offer -43 dB off isolation at 1 GHz and 0.8 dB insertion loss at 1 GHz.
Figure 2: Analog Devices' ADG918.
The SA58643 from NXP Semiconductor (Figure 3) is a wideband RF switch fabricated in BiCMOS technology and incorporating on-chip CMOS/TTL compatible drivers. Its primary function is to switch signals in the frequency range DC to 1 GHz from one 50 W channel to another. The switch is activated by a CMOS/TTL compatible signal applied to the enable channel 1 pin (ENCH1). The extremely low current consumption makes the SA58643 ideal for portable applications. The excellent isolation and low loss makes this a suitable replacement for PIN diodes. It is available in an 8-pin TSSOP package.
Figure 3: NXP’s SA58643.
The Skyworks AS179-92LF is an IC FET SPDT switch in a miniature SC-70 6-lead plastic package. The AS179-92 features low insertion loss and positive voltage operation with very low DC power consumption. This general-purpose switch can be used in a variety of telecommunications applications.
In this article, we discussed the types of RF switches that are available today and the important parameters designers must consider when selecting a switch for his/her design. We examined a relatively new a new category of RF switch, the MEMS switch, which promises the high RF performance and low DC power consumption of electromechanical switches and the small size, weight, and low-cost features of semiconductor designs.
Regardless of the type of switch you choose, selecting the best and most cost-effective RF switch for your application requires a thorough review of the datasheet of the product to determine whether its insertion loss, isolation, and other specifications meet the requirements of your system. Some vendors provide sweep charts to display these specifications for an entire range of frequencies, while others will only provide specifications for a particular frequency. In such cases, it is important to obtain complete specifications to determine if the product is suited to your application.
- Carl J. Weisman, The Essential Guide to RF and Wireless (Upper Saddle River, NJ: Prentice Hall, 2002)
- P.D Grant, R.R. Mansour, and M.W Denfoff, A Comparison between RF MEMS Switches and Semiconductor Switches.