How Do You Convert Linear to Circular Polarization?

To convert linear to circular polarization, produce two perpendicular electric-field components with equal amplitudes and a phase difference of +90° or −90°. For an existing beam, place a quarter-wave plate with its fast and slow axes at 45° to the input polarization. In a microwave feed, use a waveguide polarizer; for a dual-feed antenna, use a 90° hybrid feeding orthogonal radiating modes. The result is circular only where both the amplitude and phase conditions are met. [1] [2]

At a fixed point, a linearly polarized electric field oscillates along a fixed line. A circularly polarized field rotates while keeping a constant magnitude. Simply rotating a linear antenna changes its polarization angle; it does not create circular polarization.

The practical choice depends on your starting point: an optical or microwave beam, a waveguide port, or an antenna feed. A circularly polarized radiator such as an axial-mode helix is another option when replacing the antenna is acceptable; it is not a converter placed in an existing beam.

Axial ratio describes the polarization ellipse: 0 dB, or a linear ratio of 1.0, is a perfect circle. An ideal circularly polarized wave received by a linear antenna incurs 3.01 dB of polarization mismatch loss. That is a receiving-basis mismatch, not an unavoidable loss in the converter. [3] Which phase sign represents RHCP or LHCP depends on the coordinate and time conventions. State the handedness convention, viewing direction and propagation direction on the drawing. [4]

Conversion Tolerance Budget

Set an axial-ratio target before choosing hardware. These calculated values use two coherent orthogonal fields at one frequency and observation direction, with the amplitude imbalance and phase error shown in each row. They exclude additional multipath, coupling and measurement errors. Amplitude imbalance means 20 log10 of the larger field amplitude divided by the smaller.

With equal amplitudes, a 10° phase error gives 1.52 dB axial ratio. With perfect quadrature, a 0.5 dB amplitude imbalance gives 0.50 dB. A combined 0.4 dB imbalance and 6° phase error calculates to approximately 1.00 dB. These are total field-error budgets, not separate allowances for every component.

A Calculated Assembly Tolerance Check

Suppose a differential-phase converter has 0.4 dB intrinsic amplitude imbalance and 6° phase error. An additional 1° input misalignment from 45° contributes 0.30 dB of amplitude imbalance. If the errors reinforce, the total imbalance becomes about 0.70 dB and the calculated axial ratio rises to 1.15 dB, missing a 1 dB target. At 5° misalignment, even an otherwise ideal converter has 1.52 dB axial ratio. Allocate alignment within the total budget and leave room for measurement uncertainty. These are ideal-model calculations, not measurements.

Convert in Free Space With Wave Plates and Sheets

Use this conversion route when the linear wave is already in a beam and you need circular output without changing the feed.

Quarter-Wave Plate

A quarter-wave plate delays one field component relative to the other. Orient the linear input at 45° to its fast and slow axes so the incident components are equal. At the design wavelength, a 90° retardance produces circular output if differential transmission loss is negligible. Input along either principal axis stays linear. [5] [6]

For a simple ideal plate with constant birefringence, retardance scales as 90° × λ0/λ. At 1.05λ0, the calculated axial ratio is 0.65 dB; at 0.95λ0, it is 0.72 dB. At 1.10λ0 and 0.90λ0, the values are 1.24 dB and 1.52 dB. These assume equal transmitted amplitudes and no other errors; dispersive, multi-order or achromatic plates need their own retardance data. Specify wavelength, temperature and incidence angle rather than relying on the nominal plate label.

Meander-Line Sheet

At microwave frequencies, a meander-line sheet provides differential phase delay across a beam. NASA’s Juno toroidal low-gain antenna used a four-layer meander-line polarizer. The published design addressed the DSN 7 GHz transmit and 8 GHz receive bands, approximately 6 dBic antenna gain and a ±10° edge-of-coverage requirement. Those are complete-antenna requirements, not gain supplied by the sheet. [7] Match the polarizer’s clear aperture, spacing and incidence range to the beam; a small sheet covering only part of a horn aperture will leave mixed polarization across the field.

Convert Inside a Waveguide Polarizer

A differential-phase polarizer loads the two orthogonal TE11 polarizations of a circular guide differently, using features such as truncations, irises or dielectric loading. In a uniform loaded section, the accumulated phase difference is Δφ = (β∥ − β⊥)L. With equal transmitted amplitudes and 90° differential phase, the output is circular. A 45° linear launch supplies equal incident components; transitions and unequal losses still need checking.

A NASA study of a 30 GHz truncated circular-guide phase shifter illustrates the machining scale. Its tables give a parent diameter of 0.597λc and a phase-section length of 4.955λc, with λc = 11.38 mm: approximately 6.79 mm and 56.4 mm. The truncation is specified as t/d = 0.049, and the design table lists reflection of −31 dB at the 90° point. The authors tested an electroformed copper prototype using a vector network analyzer with thru-reflect-line calibration; differences from simulation were consistent with approximately 25 µm fabrication tolerances. The practical lesson is to control flange alignment and calibrate the test interfaces, not just machine the nominal length. [8]

A septum polarizer instead converts a rectangular-port TE10 excitation into the required orthogonal components at its common output. In a dual-input design, selecting the other rectangular input selects the opposite circular sense; follow the port drawing rather than applying the 45° launch rule. An orthomode transducer (OMT) alone separates orthogonal modes and does not necessarily supply quadrature. If an OMT and polarizer are combined, verify the assembled feed.

For high power, an all-metal design avoids dielectric heating but still needs breakdown and thermal analysis. A SLAC conceptual design at 11.424 GHz reported reflection below −20 dB, equivalent to return loss above 20 dB. Its thermal study considered 1 MW pulses lasting 100 µs at 100 Hz; that is not a 1 MW continuous-wave qualification. The results are design calculations, not a general polarizer power rating. [9]

Convert With a Hybrid and Two Radiators

A 90° hybrid divides one RF input into equal-amplitude signals in quadrature. [10] Feed orthogonal antenna modes whose radiated fields have equal amplitudes in the desired direction. Equal hybrid outputs alone are insufficient if the antenna paths differ. Both output branches must contribute: terminating one branch of an ideal equal-split hybrid discards half the input power, or 3.01 dB, and does not convert a single linear radiator to circular polarization.

Two perpendicular half-wave dipoles fed 90° apart form a turnstile, producing circular polarization along the axis normal to the dipole plane, rather than in every direction. [11] NASA also described a 2×2 patch subarray with orthogonally arranged patches and phase-shifted feeds. [12] Include the hybrid, feedlines, antenna imbalance and coupling in the assembled error budget. Check amplitude and phase at the band edges as well as the center.

Convert in the Radiator Itself

An axial-mode helix can radiate circular polarization from a single RF feed. A circumference near one wavelength helps excite this mode, but pitch, turn count and ground plane also matter. A NASA-sponsored study describes an approximately 1.7:1 frequency range with relatively constant circular polarization for the cylindrical-helix configuration discussed. That is a design example, not a guaranteed bandwidth for any helix. [13] Check winding sense, axial ratio and coverage angle for the actual geometry.

Which Conversion Method Fits Your Starting Point

If you need to switch RHCP and LHCP, specify how. A dual-input septum can change sense by input-port selection; a hybrid-fed antenna can change the relative feed phase. A quarter-wave plate can also reverse the output sense by rotating its axes by 90° relative to a fixed linear input. None of these passive parts switches itself: the system needs a port switch, phase control or mechanical rotation.

Check the receiving antenna before converting. An ideal matched circular-to-circular link has no polarization mismatch loss, whereas an ideal circular-to-linear link loses 3.01 dB; real links also have finite axial ratio and propagation effects. [3] Circular polarization can reduce sensitivity to rotation of a linear polarization plane, including ionospheric Faraday rotation, a motivation discussed in NASA’s L-band aircraft-terminal study. It does not remove blockage or multipath. [14]

Worked Conversion Example in a Waveguide

Consider an ideal, air-filled circular guide with a 20 mm internal diameter. Using the circular-guide cutoff equations, TE11 cuts off at 8.79 GHz and the next mode, TM01, at 11.47 GHz. Thus 10 GHz lies in the TE11-only mode-family window, which still supports two orthogonal polarizations. [15] The unloaded guide wavelength is approximately 62.8 mm and its propagation constant is 100.1 rad/m.

A perfectly round, unloaded guide gives both polarizations the same propagation constant: making it longer will not create quadrature. If a loaded section supplies an assumed differential propagation constant Δβ = 10 rad/m, then L = (π/2)/Δβ gives a 157 mm quarter-wave section. At 5 rad/m it becomes 314 mm; at 20 rad/m, 78.5 mm. These Δβ values are illustrative, not properties inferred from diameter alone. Loading changes the modes and cutoffs, so validate the loaded geometry and transitions with an electromagnetic model before fabrication.

How to Verify the Conversion

Test the assembled converter and antenna over frequency and coverage angle, not just at one boresight point. NIST’s three-antenna extrapolation example reported 1.34 dB axial ratio with an estimated ±0.05 dB uncertainty at 7.25 GHz. That result illustrates a calibrated measurement, not the uncertainty available from every test range. [16]

For a coherent, fully polarized wave, rotating a calibrated linear receive probe gives AR(dB) = 10 log10(Pmax/Pmin), after accounting for probe and range errors. This amplitude-only check cannot distinguish RHCP from LHCP. Verify sense with a known circular reference or calibrated phase-resolved orthogonal measurements. Also record return loss and insertion loss at defined reference planes; good matching alone does not prove circular polarization.

As a diagnostic, with equal components, changing the relative phase from +90° to −90° reverses the sense; changing it to 0° or 180° makes the field linear. Test both commanded senses if switching is required. High-power qualification is separate from these small-signal polarization checks.

Conversion Checklist

The following is an illustrative specification for a 1 dB axial-ratio objective, not a universal standard. Apply the limits to the assembled field and include uncertainty and production margin.

Against ideal same-sense and opposite-sense circular reference components, XPD = 20 log10[(r + 1)/(r − 1)], where r is the linear axial ratio. Calculated values are 24.8 dB at 1 dB AR, 21.3 dB at 1.5 dB AR and 15.3 dB at 3 dB AR. These describe one polarization ellipse, not guaranteed end-to-end link discrimination: the receiving antenna and path can add errors. [4]

Conversion Mistakes

Check these common failure modes when the measured output is not circular.

1. Launching along a principal axis. At 0° or 90°, only one component is excited. Other departures from 45° give unequal components, rather than making one disappear.
2. Using a 90° hybrid with one radiator. The hybrid splits the signal; it doesn’t make two orthogonal fields by itself.
3. Budgeting every part separately. Alignment, feed loss and intrinsic imbalance share one amplitude budget; phase errors also accumulate through the feed.
4. Measuring only at boresight. Off-axis fields and band-edge phase response can fail even when the center point passes.
5. Driving high power through a sharp edge. Peak field, pressure and surface finish matter; ask for a high-power design, not a scaled low-power one.

Choose the converter for the field or feed you actually have, establish equal output components in quadrature, and verify axial ratio and sense across the required band and beam. A nominal 90° component is only the starting point; the assembled field is what must meet the specification.

References

1. IEEE Standards Association. IEEE 145-2025 — IEEE Standard for Definitions of Terms for Antennas. IEEE, approved 10 September 2025; published 31 March 2026. https://standards.ieee.org/ieee/145/7742/
2. NASA Science, NISAR mission. Polarimetry. NASA, last updated 2 June 2026. https://science.nasa.gov/mission/nisar/polarimetry/
3. International Telecommunication Union. Recommendation ITU-R F.1245-3 — Mathematical model of average and related radiation patterns for point-to-point fixed wireless system antennas for use in interference assessment in the frequency range from 1 GHz to 86 GHz. ITU, approved January 2019. https://www.itu.int/rec/R-REC-F.1245/en
4. NASA/Jet Propulsion Laboratory, DESCANSO Radiowave Propagation Handbook. Depolarization on Earth-Space Paths, Chapter 4. NASA/JPL, undated PDF; references through 1986; accessed 24 October 2026. https://descanso.jpl.nasa.gov/propagation/1082/1082ch4.pdf
5. Richard P. Feynman, Robert B. Leighton, Matthew Sands. The Feynman Lectures on Physics, Volume I, Chapter 33: Polarization. California Institute of Technology, 1963; online edition accessed 24 October 2026. https://www.feynmanlectures.caltech.edu/I_33.html
6. Harvard Natural Sciences Lecture Demonstrations. Circular Polarization. Harvard University, accessed 24 October 2026. https://sciencedemonstrations.fas.harvard.edu/presentations/circular-polarization
7. Luis R. Amaro, Ronald C. Kruid, Joseph D. Vacchione, Aluizio Prata. Low-Gain Circularly Polarized Antenna With Torus-Shaped Pattern. NASA Tech Briefs; JPL NPO-48320, October 2012. https://ntrs.nasa.gov/api/citations/20120016278/downloads/20120016278.pdf
8. N. P. Lourie, D. T. Chuss, R. M. Henry, E. J. Wollack. Investigation of Truncated Waveguides. NASA Technical Reports Server, document 20130011724, 29 January 2013. https://ntrs.nasa.gov/api/citations/20130011724/downloads/20130011724.pdf
9. Chen Xu, Sami Tantawi, Juwen Wang. Conceptual design of X band waveguide dual circular polarizer. Physical Review Accelerators and Beams 19, 062003; SLAC National Accelerator Laboratory, 17 June 2016. https://doi.org/10.1103/PhysRevAccelBeams.19.062003
10. Jim Stiles. The Quadrature Hybrid Coupler. University of Kansas, Department of Electrical Engineering and Computer Science, 17 April 2009. https://www.ittc.ku.edu/~jstiles/723/handouts/The%20Quadrature%20Hybrid%20Coupler%20723.pdf
11. R. B. Jackson. The Canted Turnstile as an Omnidirectional Spacecraft Antenna System. NASA Goddard Space Flight Center; NASA X-712-67-441, September 1967. https://ntrs.nasa.gov/api/citations/19670027843/downloads/19670027843.pdf
12. John Huang. Wide-Band, Wide-Scan Antenna For Circular Polarization. NASA Tech Briefs, Vol. 12, No. 1; NPO-16831, 1 January 1988. https://ntrs.nasa.gov/citations/19880000006
13. J. M. Loomis III and E. R. Graf. Receiver Antennas for Application in a Television Broadcast Relay System. Auburn University Engineering Experiment Station for NASA Marshall Space Flight Center; NTRS document 19710018952, 30 January 1970. https://ntrs.nasa.gov/api/citations/19710018952/downloads/19710018952.pdf
14. W. V. Kiskaddon and Duncan M. Carman; Commercial Airplane Division, Boeing. Experimental L-Band SST Satellite Communications/Surveillance Terminal Study, Volume IV: Aircraft Antenna Studies. NASA Electronics Research Center; NTRS document 19690011734, November 1968. https://ntrs.nasa.gov/api/citations/19690011734/downloads/19690011734.pdf
15. W. C. Chew. More on Hollow Waveguides. Purdue University, ECE 604 Lecture 19, 2020. https://engineering.purdue.edu/wcchew/ece604s20/Lecture%20Notes/Lect19.pdf
16. Andrew G. Repjar, Allen C. Newell, Douglas T. Tamura. Extrapolation Range Measurements for Determining Antenna Gain and Polarization. National Bureau of Standards (now NIST), Technical Note 1311, August 1987. https://nvlpubs.nist.gov/nistpubs/Legacy/TN/nbstechnicalnote1311.pdf