Microwave mixers and pre-scalers are designed to translate signals from one frequency to another, but they work in different ways and are used in quite different applications. At present, specifiers for these products have an unprecedented choice of both kinds of components in a variety of technologies and package types. Knowing the basics of how each type of component works and they key specifications to compare them can help when deciding on a particular product for an application.
Much has been written about RF and microwave mixers, with a variety of excellent introductory articles available to better understand these essential components. For example, "Mixer Basics PrimerA Tutorial For RF and Microwave Mixers," written by Ferenc Marki and his son Christopher Marki, is available in PDF form for free download from the Marki Microwave web site at www.markimicrowave.com. Mini-Circuits also offers an excellent nine-page application note, "Understanding MixersTerms Defines, and Measuring Performance." Also, a 20-page introductory article in PDF form written by Liam Devlin of Plextek, simply titled "Mixers" does an excellent job of detailing different types of diode mixers and their key characteristics.
Mixers are aptly named, as their function relies on mixing two signals to produce a third. It is the relationship of the two input signals that determines the characteristics of the third signal. A mixer's three ports are usually labeled as the RF, local oscillator (LO), and intermediate-frequency (IF) ports. The LO port always serves as an input port; the other two ports can swap places as input or output ports, depending on how the mixer is needed. When the desire is to translate a signal of interest to a lower frequency, a process known as frequency downconversion, the signal is injected to a mixer's RF port and mixed with an LO signal to produce an IF signal that is the difference in frequency between the LO and the RF. To translate a signal of interest to a higher frequency, the IF port is used as an input, along with the LO port, and two signals are produced at the RF port: the difference in frequency between the LO and the IF and the sum of the frequencies of the LO and IF signals. This process is typically known as double-sideband upconversion, where two different signals are available at the mixer's RF port. Mixers are also available for single-sideband upconversion, where one of the signals is suppressed or filtered within the mixer resulting in only one signal available at the mixer's RF port.
Mixers rely on the characteristics of a nonlinear device, such as a Schottky diode or GaAs field-effect transistor (FET), to achieve frequency translation. Discrete Schottky diodes are typically used in high-performance passive microwave mixers, whereas GaAs or silicon CMOS transistors are more typically incorporated as the nonlinear devices for active mixers in integrated circuits (ICs). Diode mixers come in various forms, including single-diode, singleended mixers; single-balanced mixers that incorporate two diodes in an anti-parallel pair; and double-balanced mixers that use four diodes in a ring or star configuration. More diodes mean higher isolation between ports. But with an increase in the number of diodes, the amount of LO power required must also increase, since it is the LO power that turns on the diodes and activates the switching or heterodyning function that enables frequency translation.
When particularly high isolation is needed, some manufacturers offer double- double-balanced mixers, essentially two interconnected double-balanced mixers, but the penalty is the higher LO power required with the additional diodes. In addition, for special applications, subharmonic mixers are often used in applications requiring millimeterwave frequencies and image-reject mixers are used where image frequencies produced by frequency translation may be problematic.
Mixers are characterized by a number of parameters, including the frequency ranges of the different ports, the conversion loss through the mixer, isolation between ports, noise figure, 1-dB compression, VSWR, and third-order intercept point. The conversion loss in a passive downconverter mixer is the difference between the input RF power and the output IF power, neglecting the power level of the LO signal, which is essentially used to turn on the mixer's nonlinear devices, such as its Schottky diodes. Mixers designed to handle wider bandwidths tend to have higher conversion loss due to the difficulty of achieving good impedance matches across all the different frequencies.
The conversion-loss relationship usually holds for all levels of RF power within a mixer's specified operating range. For example, a mixer with 5-dB conversion loss working with an applied RF signal of +20 dBm (and the proper LO power) will normally yield an IF level of +15 dBm. The 1-dB compression point is reached at a high-enough RF power level that effectively results in an additional 1 dB of conversion loss over the mixer's specified loss characteristic, and a departure from the normal linear relationship between input RF power and output IF power. It should also be noted that active mixers, which operate with a bias supply, typically provide conversion gain rather than loss, although they are typically limited in the amount of RF power they can handle compared to a diode mixer before compression or distortion degrades performance. It should be noted that, in addition to conversion loss, some manufacturers will also characterize a mixer for noise figure, which is generally slightly less than the mixer's conversion loss.
Mixer isolation is a measure of the amount of power that leaks from one mixer port to another. It is typically measured in three directions: from the LO to the RF port, from the LO to the IF port, and from the RF to the IF port. As mentioned earlier, in a diode mixer, isolation is a function of the number of diodes in the mixer circuit, with greater mixer complexity and number of diodes yielding higher port-to-port isolation. High isolation is important for preventing signal leakage from one port to another, such as high-level LO input signals mixing with IF output signals.
VSWR is a measure of how well a mixer's ports are matched to the characteristic impedance of the system in which the mixer will be used, typically 50 Ω. Ideal impedance matching results in minimal signal reflections from filtering and minimal opportunities for intermodulation distortion from reflected signals. Mixer manufacturers may provide VSWR values or specify the quality of the impedance match in terms of return loss.
PICKING A PRESCALER
Prescalers also provide frequency translation, but in a different manner than frequency mixers. A prescaler is essentially a frequency divider with an integer modulus or division ratio, such as a divide-by-2 prescaler or a divideby- 8 prescaler. Fixed-integer microwave prescalers are commonly used in phaselocked- loop (PLL) frequency synthesizer circuits, to scale down the frequency of a high-frequency source, such as a voltage-controlled oscillator (VCO), for comparison to the phase of a more-stable but lower-frequency reference oscillator, such as a crystal oscillator. Commercial fixed-integer microwave prescalers, such as several models from Centellax, are available at frequencies to 20 GHz and higher, with divide ratios of 2, 4, and 8. A dual-modulus prescaler provides a pair of closely spaced divide ratios, such as 32/33 and 64/65, to allow designers to toggle between divide ratios as needed.
Prescalers are typically available from semiconductor suppliers in the form of ICs, with suppliers including Analog Devices, Avago Technologies, Centellax, Freescale Semiconductor, Hittite Microwave, Maxim Integrated Circuits, National Semiconductor, On Semiconductor, and Renesas Electronics. To assist engineers with the prescaler specifying process, Hittite offers a six-page product application note, "Selecting Prescalers for PLL Synthesizers," which uses its own IC products to teach about prescaler parameters and how to use these components in practical circuits.