Single-, double-, and triple-balanced mixer architectures are designed to optimize the electrical performance, cost, complexity, and technology limitations of a fabrication process. Hence, mixer architectures can be discussed in terms of their benefits and trade-offs compared to other mixer architectures and value in specific applications.
A triple-balanced mixer is composed of two diode quads, with a total of eight junctions. Power splitter at the RF and LO microwave baluns feed the mixer structure, which enables both of the diode quads to be coupled, with matching RF/LO isolation. This allows for the IF signal to be available at two separate isolated terminals, that typically exhibit very large bandwidths compared to other mixer architectures. Though, a DC IF is not available in this architecture. Practical, and available, triple-balanced mixers generally demonstrate better spur suppression than other mixer architectures, with the exception of some high port isolation double-balanced mixer designs.
Triple-balanced mixers enable low intermodulation distortion (IMD) upconversion and downconversion over very wide bandwidths, even into the high microwave and millimeter-wave frequencies. Double-balanced mixers, on the other hand, are less complex and lower cost circuits that are fit for applications where moderate LO power is available and there are no concerns over overlapping RF and IF frequencies. The above, and the DC IF capability of some double-balanced mixers make them suited for demodulator, I/Q modulators and phase detector circuits for narrow or wide bandwidths. Generally, triple-balanced mixers also require about 3 dB more LO power, as the LO power is divided between the two diode quads.