What is a Kumar choke for radio frequencies?

A Kumar choke (often called the VE4MA or Kumar feed/choke) is a choke-ring / cavity around a horn/feed that suppresses unwanted surface and coaxial currents and “traps” or reflects energy that would otherwise spill past the feed. The result: lower spillover, smaller sidelobes, reduced cross-polarization and lower antenna noise temperature — especially useful on dish feeds around 1–10 GHz. (om6aa.eu)

How it works (step-by-step, plain language)

1. Problem it solves
   • A simple horn or probe feeding a reflector lets some energy leak sideways and down the structure (surface currents on the horn flange, currents on the coax shield, etc.). That leakage becomes spillover and sidelobes, and can couple to warm ground or structure — raising the antenna temperature and distorting the pattern. (lso.fe.uni-lj.si)
2. The choke geometry
   • The Kumar solution surrounds the horn aperture/ flange with a circular cavity or set of choke rings (a “choke flange” / low-Q trap). The cavity dimensions are chosen on the order of ~λ/4 (depth and/or ring width around the aperture) so that it presents a high impedance to the unwanted surface currents at the operating band. (w1ghz.org)
3. What the choke does electromagnetically
   • At the design frequency the choke/cavity is resonant so currents trying to flow into the flange or down the outside of the feed see a high impedance and are reflected back into the horn instead of radiating outward. In transmission-line language: the choke puts a “stop” on common-mode / surface current paths at RF. That reduces radiation from the feed edges and the feed support structure. (ResearchGate)
4. Net effects on the antenna pattern
   • Less spillover → lower sidelobes and lower pickup of ground noise (lower system/antenna temperature).
   • Improved symmetry and reduced cross-polarization in many designs, because stray currents that distort the polarization are suppressed.
   • Often broader usable bandwidth than a single choke flange because the cavity/ring approach behaves like a low-Q trap (more forgiving to frequency changes). (astropeiler.de)

Practical notes / design tips

• Typical dimensions are related to wavelength (λ): choke depth and ring widths around ~λ/4 are common starting points. Exact dimensions depend on horn geometry and f/D of the reflector. (w1ghz.org)
• There are variations (single choke flange, multiple choke rings, cavity around the horn, etc.). Different implementations trade bandwidth vs. peak suppression. (pa0ply.nl)
• If you’re building or simulating one: model the feed + choke together (EM solvers or measured testbeds) — real performance depends on the whole assembly, not just the ring dimensions. (ResearchGate)