EM Applications of ElectromagneticsBuilt using ChatGPT · Connected engineering reference
Home › RF & Microwave Engineering
On this page

RF & Microwave Engineering

Smith chart relating normalized impedance and reflection coefficient.
The Smith chart connects transmission-line impedance to the complex reflection coefficient and remains one of RF engineering's most useful graphical tools. Image source: Wikimedia Commons.
30-second intuitionAt RF and microwave frequencies, interconnects are no longer “just wires.” Their physical length becomes comparable with wavelength, so voltage and current propagate as waves, reflections matter, and geometry becomes part of the circuit.

1. Transmission lines

A distributed line is described by per-unit-length $R,L,G,C$:

\[Z_0=\sqrt{\frac{R+j\omega L}{G+j\omega C}},\qquad \gamma=\sqrt{(R+j\omega L)(G+j\omega C)}.\]

For an ideal lossless line,

\[Z_0=\sqrt{L/C},\qquad \gamma=j\beta.\]

2. Reflection and mismatch

At a load $Z_L$,

\[\Gamma_L=\frac{Z_L-Z_0}{Z_L+Z_0}.\]

Return loss and VSWR are

\[RL=-20\log_{10}|\Gamma|,\] \[VSWR=\frac{1+|\Gamma|}{1-|\Gamma|}.\]

Mismatch loss is

\[ML=-10\log_{10}(1-|\Gamma|^2).\]
Common misconceptionA low $S_{11}$ is not automatically a “good antenna/device.” It only says little power is reflected at that reference plane; power may still be dissipated rather than usefully radiated/transmitted.

3. Input impedance of a line

For a lossless line of length $l$,

\[Z_{in}=Z_0\frac{Z_L+jZ_0\tan\beta l}{Z_0+jZ_L\tan\beta l}.\]

At $l=\lambda/4$, a quarter-wave transformer can match two real impedances using

\[Z_t=\sqrt{Z_0Z_L}.\]

4. Smith chart

The Smith chart maps normalized impedance/admittance to complex reflection coefficient. It is especially useful for:

5. S-parameters

For a two-port network,

\[\mathbf b=\mathbf S\mathbf a.\]

At high frequency, S-parameters are practical because incident/reflected traveling waves are easier to measure than terminal open/short currents and voltages.

6. Rectangular waveguide

For a rectangular guide with broad dimension $a$ and height $b$, cutoff is

\[f_{c,mn}=\frac{c}{2\sqrt{\epsilon_r}}\sqrt{\left(\frac{m}{a}\right)^2+\left(\frac{n}{b}\right)^2}.\]

The dominant mode is usually TE$_{10}$:

\[f_{c,10}=\frac{c}{2a\sqrt{\epsilon_r}}.\]

Waveguides carry modes, not simple TEM waves.

7. Resonators and quality factor

A resonator stores electric and magnetic energy. A useful relation is

\[Q=2\pi\frac{\text{energy stored}}{\text{energy lost per cycle}} \approx\frac{f_0}{\Delta f_{3\text{dB}}}.\]

Loaded $Q$ includes internal loss and external coupling.

8. Filters and coupled resonators

Microwave filters are implemented with transmission-line sections, cavities, microstrip resonators, dielectric resonators, waveguide irises and lumped/distributed hybrids. Key specifications include insertion loss, return loss, fractional bandwidth, group delay, rejection, power handling and temperature stability.

9. Couplers, dividers and hybrids

Directional couplers sample forward/reverse waves. Wilkinson dividers provide matched splitting and isolation. 90°/180° hybrids enable balanced amplifiers, mixers, beamforming and measurement networks.

Worked example — 50 Ω to 100 Ω quarter-wave match

\[Z_t=\sqrt{50\times100}=70.7\ \Omega.\]

At 2 GHz in a medium with effective dielectric constant $\epsilon_{eff}=2.25$,

\[\lambda_g\approx\frac{c}{f\sqrt{\epsilon_{eff}}}\approx0.10\ \text{m},\]

so a quarter-wave section is roughly 25 mm before discontinuity/end-effect corrections.

10. Measurement workflow

A VNA is the central RF network instrument. Good practice:

  1. calibrate at the actual reference plane;
  2. choose source power that does not compress the DUT;
  3. use appropriate IF bandwidth/averaging;
  4. control cable movement;
  5. de-embed fixtures only with a defensible model;
  6. inspect phase/group delay, not only magnitude.

See Measurements & Instruments.

11. Engineering reality

What changes real hardwareConductor/dielectric loss, connector launches, dispersion, radiation, surface roughness, via inductance, package modes, parasitics, thermal drift, nonlinear devices and calibration error all become increasingly important with frequency.

References

Microwave and space component reference

For a hardware-oriented map of the components that make up real microwave and satellite RF chains—including LNAs, PAs/TWTAs, mixers, filters, switches, circulators, limiters, PLL/VCOs, beamformers, ADC/DACs, waveguide hardware, MMIC technologies, and space qualification—see Microwave & Space Electronics Components.

Radio astronomy as an RF receiver problem

For cryogenic LNAs, SIS mixers, low-noise feeds/OMTs, phase-stable LO distribution, IF chains, ADC/channelizer backends, RFI linearity and interferometric correlation, see Radio Astronomy — RF Circuits & Receiver Systems.

RF transceiver architectures

For an interview-focused comparison of superheterodyne, zero-IF, low-IF, image-reject, IF-sampling and direct-RF-sampling receivers—plus direct-conversion/direct-RF transmitters, TDD/FDD front ends, mixers, synthesizers, I/Q impairments and calibration—see RF Transceiver Architectures — Interview Ready.

Modern RF signal generation

For the internal architecture of a modern RF/vector signal generator—including reference clocks, PLLs, baseband I/Q, interpolation, DDS/NCO, DACs, I/Q modulation, direct-RF generation, ALC, phase noise and multi-channel coherence—see Modern Signal Generator.

Spectrum analyzer signal processing

For the internal RF/DSP chain of a modern real-time analyzer—using the Signal Hound BB60D as a concrete example—see Spectrum Analyzer Signal Processing.

RF band quick reference

For a compact table of GSM/LTE/NR operating bands plus L/S/C/X/Ku/K/Ka/V/W microwave bands, see RF Bands — Compact Reference.

Cellular RF standards

For NR/LTE transmitter, receiver, RRM, conformance and test-environment specifications, see 3GPP Standards — What an RF & Wireless Engineer Should Know.

Related: Antennas · Signal Integrity & Power Integrity · Paul Traps · EMC · Calculators · Worked Examples


Advanced study — network stability, power waves and nonlinear RF design

Cascading linear networks correctly

For a two-port network, (S)-parameters are referenced to port impedances. Power transfer depends on source/load match, so ( S_{21} ^2) is not generally the transducer power gain. When cascading stages, use an ABCD or chain-scattering representation with consistent port conventions, then convert the combined network back to (S)-parameters. Avoid multiplying individual (S_{21}) terms when substantial interstage reflections are present.

Stability and matching

For a two-port transistor network define (\Delta=S_{11}S_{22}-S_{12}S_{21}) and

\[K=\frac{1-|S_{11}|^2-|S_{22}|^2+|\Delta|^2}{2|S_{12}S_{21}|}.\]
For the usual two-port small-signal stability test, (K>1) and ( \Delta <1) indicate unconditional stability for passive source and load terminations at the frequency examined. This is a linear frequency-domain statement; check bias-dependent stability, out-of-band modes, packaging resonances and large-signal behavior separately.

Noise and nonlinear tradeoffs

The Friis noise-factor cascade is

\[F_{\mathrm{tot}}=F_1+\frac{F_2-1}{G_1}+\frac{F_3-1}{G_1G_2}+\cdots,\]

with linear available gains under the assumptions of the cascade model. Early attenuation can dominate noise figure, while early high gain may compromise blocker tolerance. Power amplifiers require large-signal load-pull, AM-AM/AM-PM and efficiency characterization instead of small-signal (S)-parameters alone.

Advanced exercise: design an LNA chain against simultaneous NF, IIP3 and stability constraints. Document impedance reference planes and a sensitivity analysis to component tolerances.

Further reading: D. M. Pozar, Microwave Engineering; G. Gonzalez, Microwave Transistor Amplifiers; NIST VNA uncertainty analysis.


Advanced topic — stability and cascade noise

The usual two-port small-signal stability quantities are

\[\Delta=S_{11}S_{22}-S_{12}S_{21},\qquad K=\frac{1-|S_{11}|^2-|S_{22}|^2+|\Delta|^2}{2|S_{12}S_{21}|}.\]
For a linear two-port, (K>1) and ( \Delta <1) indicate unconditional stability for passive source and load terminations at the frequency considered. Inspect out-of-band frequencies, bias dependence and large-signal operating points as well. Forward gain ( S_{21} ^2) is not generally the transducer gain if the source and load are mismatched.

For a noise-factor cascade with linear available gains,

\[F_{\mathrm{total}}=F_1+\frac{F_2-1}{G_1}+\frac{F_3-1}{G_1G_2}+\cdots.\]

Loss preceding the first LNA often dominates receiver noise figure; excessive early gain may instead compromise blocker tolerance.

Advanced experiment: design a two-stage receiver to meet NF, compression and stability simultaneously, then propagate VNA calibration and bias uncertainty into the predicted gain.

Further reading: D. M. Pozar, Microwave Engineering; NIST covariance-based VNA uncertainty.

Advanced RF applications

For systems that turn RF theory into calibrated observables, see RF Direction Finding, Precision Timing & Synchronization, Advanced Antenna Metrology, Radio Science & Tracking, and the Advanced Engineering Learning Map.