FAQ on RF directional couplers: part 1

This essential measurement accessory for microwaves has morphed from waveguides down to tiny surface-mount components.

Engineers who work with circuits typically use a multimeter as their primary test and measurement tool to check voltages and currents. They may also need to assess power with a separate instrument, but this is often secondary. However, voltage and current are far behind on the list of parameters of interest for RF designers working at hundreds of megahertz and above (in the microwave and millimeter wave ranges).

Instead, it’s usually about RF power, which is usually measured over the defined frequency range. The challenge is measuring power with little or at least known impact on the measured signal. Unlike voltage measurement, which can be done between two circuit points using a high-impedance meter, which does not load the circuit itself, what is needed is a way to “tap off” or divert a small fraction of the incident RF power for measurement purposes while leaving the remaining much larger fraction for use in the circuit and system.

That’s what a directional coupler does. This article will look at this device’s basics and how its physical construction has advanced to produce smaller, surface-mount devices that are a good fit for lower-power applications. Advanced stripline and other processes on a ceramic substrate make these directional couplers far smaller and less costly. However, those other implementations are still vital for situations where power levels are higher, discrete cable connections are needed, or the application mandates certain levels of performance and capability.

The first part of this article reviews the principles and parameters of directional couplers, and the second part examines some of the many older and newer constructions.

Q: What is the objective and role of the directional coupler?
A: Its basic function is to sample RF signals with a predetermined degree of sampler coupling, with high isolation between the signal and sampled ports. In this context, ” isolation ” does not refer to galvanic (ohmic) isolation but rather interaction between the various ports.

Q: Is the coupler active or passive?
A: It is completely passive, meaning it provides no gain or needs operating power. Its ability to function is due to its “mechanical” construction and materials.

Q: What does the schematic diagram of a coupler look like?
A: Figure 1 shows a complete basic coupler called a 4-port coupler. It includes both coupled (forward) and isolated (reverse or reflected) ports. This basic version is bidirectional but is simplified for directional performance.

Figure 1. The basic 4-port coupler is a symmetrical, bidirectional device for RF power paths. (Image: Mini-Circuits)

Q: So where does the “directional” aspect come into the picture?
A: A directional coupler (for obvious reasons, this name is always spelled out and not abbreviated as “DC,” which would cause terminology confusion!) is a simplified version of the 4-port coupler. It eliminates the isolated port and is called a 3-port coupler.

This coupler has three accessible ports, as shown in Figure 2 (left). The isolated fourth port is internally terminated in the line’s characteristic impedance (usually 50 Ω or 75 Ω, if needed) to provide maximum directivity and minimize reflections, as seen in Figure 2 (right).

Figure 2. (left) In the 3-port coupler, Port 4 is inactive while Port 3 represents diverted (reflected) power; (right) Port 4 is terminated in the lines’ characteristic impedance to minimize undesired reflections. (Image: Mini-Circuits)

The basic function of a directional coupler is to sample the isolated (reverse-power) signal. The coupled port samples the reverse wave in the forward direction (as shown). If needed, that sampled power can then be directed to a power meter or other part of the circuit for a control loop or other function. As the ratio between input-port power and the sampled power is known, the actual amount of power can be quantified.

Top performance parameters

Q: What is the relationship between input power and sample power, and how is it measured?
A: This is called the coupling factor and indicates the fraction of the input power (at P1) delivered to the coupled port, P3. It is measured in dB:

Coupling factor = 10 log (P1/P3)

Q: What are typical coupling factors available in standard directional couplers?
A: The most common coupling factors are 10, 20, 30, and 40 dB, corresponding to coupling ratios of 10:1, 100:1, 1000:1, and 10,000:1. Other values, such as 3, 6, 13, and 35 dB, are also available. Nearly all vendors will make a custom coupler with a non-standard coupling factor.

Q: What are some of the other top considerations that define the performance of a directional coupler?
A: Besides the coupling factor, three other parameters are expressed in dB. For convenience, we’ll put all four and their respective and simple equations here:

1) Coupling Factor (C) is the fraction of the input power (at P1) delivered to the coupled port, P3.

Coupling factor (C) = 10 log (P1/P3).

2) Directivity (D): a measure of the coupler’s ability to separate waves propagating in forward and reverse directions, as observed at the coupled (P3) and isolated (P4) ports.

Directivity = 10 log (P3/P4)

3) Isolation: Indicates the power delivered to the uncoupled load (P4).

Isolation = 10 log (P1/P4)

4) Insertion Loss (L): Measures the input power (P1) delivered to the transmitted (P2) port, which is reduced by power delivered to the coupled and isolated ports.

Insertion Loss = 10 log (P1/P2)

Q: What do the performance curves look like for these parameters?
A: It depends on the frequency range, size, vendor, technology used, packaging, and many other factors; one representative quartet is shown in Figure 3.

Figure 3. The four key performance graphs for the Mini-Circuits ZCDC20-E18653+, a 20-dB, 50 Ω directional coupler for 18 GHz to 65 GHz, capable of handling 12 watts and fitted with 1.85mm female RF connectors. (Image: Mini-Circuits)

Q: Are those the only parameters of interest?
A: Absolutely not. In the RF world, nothing is ever simple or defined by just a few factors. Other factors for directional couplers include the frequency range and flatness over which the coupler operates and meets its specifications, temperature-related performance, and maximum power-handling rating.

A coupler’s frequency sensitivity (flatness) characterizes how coupling varies over a given frequency range. Typical coupling flatness for a coupler operating over an octave band (2:1 frequency range) is within ± 0.75 dB of the nominal value. High coupling factors (3, 6, and 10 dB) generally have better flatness than weaker coupling factors (20 through 50 dB).

Power handling is a function of the coupler’s technology, design, and construction. A coupler for a few watts differs greatly from one for hundreds or thousands of watts. The physical elements must not only handle the power but also accommodate and dissipate the heat that results from insertion loss.

Q: What are the frequency ranges of directional coupler operation?
A: Whatever you want, within limits. Narrowband couplers are available with consistent specifications over a narrow range, such as 1-to-2 GHz. Wideband ones handle 20-to-65 GHz; there are also couplers that function to the tens of GHz and even to 100 GHz and beyond.

The next part of the article looks at directional couplers’ technologies and physical design.