Microwave mixers are widely used in commercial and military electronic systems for frequency conversion. They are essential for shifting the frequency of a signal downward (downconversion) or upward (upconversion) as might be needed in a receiver and transmitter, respectively. By understanding some fundamental concepts about mixer operation and performance, specifying a mixer for an application can be greatly simplified.

A frequency mixer is a three-port component, with two ports accepting input signals and one providing an output signal that is the sum and difference of the input signals (see figure). This mixing process can be represented mathematically by

fout = fin1 fin2 (1)

The three ports in a microwave mixer are the radio-frequency (RF), local oscillator (LO), and intermediate-frequency (IF) ports. The LO port, which is used exclusively for input signals, is typically driven with either a sinusoidal continuous wave (CW) signal or a square-wave signal. The choice depends on the application and the mixer. The LO signal acts as a gate, with the mixer "on" when the LO is a large enough voltage and "off" when the LO is a small voltage. The RF and IF ports can be interchanged as either the second input or the output. When the desired output frequency is lower than the second input frequency, then the process is called downconversion and the RF is the input and the IF is the output:

fIF = |fLO - fRF| (2)

When the desired output frequency is higher than the second input frequency, then the process is called upconversion and the IF port is used for an input signal and the RF port provides the output signal. In the upconversion case, both the sum and difference frequencies (fRF1 and fRF2) are available at the RF output port. This type of upconversion is known as double-sideband (DSB) upconversion, in contrast to single-sideband (SSB) conversion where either the sum or the difference frequency is intentionally canceled inside the mixer.

IF/RF signals tend to be information-bearing signals (as denoted by the broadened spectra surrounding the RF and IF center frequencies). During frequency conversion, the information carried by the RF (IF) signal is frequency translated to the IF (RF) output. Therefore, mixers perform the critical function of translating in the frequency domain.

Any nonlinear or switching device can be used to make a mixer circuit, but only a handful provide the type of electrical performance preferred for many applications. The devices of choice for modern mixer designers are Schottky diodes, GaAs FETs and CMOS transistors. FET and CMOS mixers are typically used in higher-volume, cost-sensitive applications, while Schottky diode mixers are used almost exclusively in applications requiring the highest performance.

The simplest mixer consists of a single diode. An RF signal and higher-level LO signals are combined at the diode's anode. The LO is assumed to be a much larger signal such that only it affects the transconductance of the diode. The diode is assumed to switch instantaneously with the injection of the LO signal, in the manner of an ideal commutator. Mixing occurs due to the switching response of the diode's current-voltage (I-V) curve to the strong LO signal. As the diode is forced open and closed by the LO, the small signal RF is "chopped," creating IF signal products according to

fIF = nfLO fRF (n is odd only) (3)

In an ideal switching commutator, only odd harmonics of the LO will mix with the fundamental RF tone. In reality, some amount of "turn-on" transition is needed by the diode. Moreover, even a low-level RF signal will modulate the diode transconductance. The combination causes additional spurious mixing products, so that diodes produce even and odd harmonic mixing products:

fIF = nfLO mfRF (m, n = integers) (4)

High-performance mixers are designed using four and sometimes eight diodes, with additional circuitry to route signals to and from the diodes and provide filtering and port-to-port isolation. In single diode mixers, the extra circuitry typically comes in the form of some kind of passive coupling, power division, and/or filtering. More complex mixer architectures can be created by combining a pair of single-balanced mixers (each with two diodes) to form a double-balanced mixer, or two double-balanced mixers to form a triple-balanced mixer.

Mixer performance can be characterized by a number of key parameters, including conversion loss, isolation, and 1-dB compression. Conversion loss is defined as the difference in power between the input RF power level and the desired output IF frequency power level. If the input RF is -10 dBm and the downconverted IF output signal -17 dBm, then the conversion loss is 7 dB. The theoretically optimum (minimum) conversion loss for a passive diode mixer is 3.9 dB, although typical values range between about 4.5 to 9 dB. Additional losses are caused by transmission-line losses, balun mismatch, diode series resistance, and mixer balance. Double-balanced mixers tend to have lower conversion loss than triple-balanced mixers. Wider-bandwidth mixers tend to have higher conversion loss than narrower-bandwidth models in part due to the difficulty in maintaining circuit balance over the entire bandwidth.

Conversion loss is a benchmark metric for mixers since it is a good indicator of mixer quality and correlates closely with other mixer parameters, such as isolation and 1-dB compression. A mixer with good conversion-loss performance tends to have good isolation and 1-dB compression performance as well. The converse is not necessarily true however; it is possible to have good isolation and poor conversion loss, for example.

Isolation is a measure of the amount of power that leaks from one mixer port to another. The three types of isolation commonly quoted in mixer data sheets are isolation between the LO and RF ports (L-R isolation), between the LO and IF ports (L-I isolation), and between the RF and IF ports (R-I isolation).

A mixer will undergo conversion-loss compression for large RF signal levels. Under normal linear operation, the conversion loss will be constant, regardless of input RF power. If the input RF power increases by 1 dB, the output IF power will also increase by 1 dB. As the RF power level increases, this linear relationship will not hold, and the mixer will reach a compression level known as the 1-dB compression point. It is defined as the input RF power required to increase the conversion loss by an extra 1 dB from ideal.

Under linear operation, the LO power is so much stronger than the RF power that the diode switching action is dominated by the LO signal power. In compression, the RF power is commensurate with the LO power such that the diode switching action is compromised. When the RF power is too high, the mixer balance degrades and the mixer circuit behaves like a single diode mixer. Operating a mixer in compression can cause increased levels of intermodulation distortion, higher conversion loss, and degraded isolation performance.

Mixer compression can be improved by using higher turn-on diodes. In this way, a larger RF power can be applied to the mixer without challenging the diode turn-on voltage. The trade-off is that a larger LO drive must also be applied to switch the diode on and off. As a general rule of thumb, the 1 dB compression point will be anywhere from 4 to 7 dB below the minimum recommended LO drive level of the mixer.

Editor's Note: This RF Primer is excerpted from a longer white paper, "Mixer Basics Primer," which is available for free download from the Marki Microwave web site at www.markimicrowave.com.