Choke Flange
Understanding Choke Flange
When two waveguide sections are bolted together, the quality of the RF connection depends entirely on electrical continuity at the junction. Any gap, surface roughness, or oxide layer creates a discontinuity that reflects energy and increases loss. Cover (flat) flanges rely on precise machining and high bolt torque to achieve intimate metal contact, and even then, performance degrades with repeated assembly cycles as surfaces wear. The choke flange solves this by engineering the discontinuity out of the problem: a concentric groove machined into the flange face creates a transmission-line transformer that presents a virtual short circuit at the waveguide aperture edge, regardless of the physical contact quality.
The mechanism uses two cascaded quarter-wave sections. The groove bottom is a physical short circuit. The first quarter-wave section (groove depth) transforms this into an open circuit at the groove opening. The second quarter-wave section (from groove to waveguide wall) transforms this open circuit back into a short circuit at the critical junction. The total path length from groove bottom to waveguide wall is therefore a half wavelength, creating a resonant structure that enforces current continuity. In practice, the standard mating configuration is one choke flange paired with one cover (flat) flange. Never mate two choke flanges together, as the facing grooves create a resonant cavity that produces spurious responses. For frequencies from 8 to 110 GHz, MIL-DTL-3922 and IEEE 1785 define standard choke flange dimensions for each waveguide size.
Choke Flange Design Parameters
d = λg / 4 [mm, at center frequency]
Guide Wavelength:
λg = λ0 / √(1 - (λ0/(2a))²) [mm]
Groove-to-Wall Distance:
rgroove - rwall ≈ λg / 4 [mm]
Where λg = guide wavelength, λ0 = free-space wavelength, a = broad wall dimension, rgroove = groove centerline radius, rwall = waveguide aperture edge radius. Total transformer length (groove bottom to wall) = λg/2.
Flange Type Comparison
| Parameter | Choke + Cover | Cover + Cover | Precision Flat (IEEE 1785) |
|---|---|---|---|
| Return Loss | > 50 dB (center) | 35 to 45 dB | > 55 dB |
| Insertion Loss | < 0.03 dB | 0.02 to 0.05 dB | < 0.02 dB |
| Gap Tolerance | Up to 50 μm | < 5 μm | < 2 μm |
| Repeatability | Excellent | Good (degrades) | Excellent |
| Typical Use | Field, rotating joints | Lab bench setups | Calibration, metrology |
Frequently Asked Questions
How does the quarter-wave choke groove work?
The groove depth is λg/4. Its bottom is a short circuit (metal wall), transformed to an open circuit at the groove opening by the quarter-wave section. A second λg/4 from groove to waveguide wall transforms this open back to a short circuit. The total half-wave path ensures current flows continuously across the junction as if the flanges were welded, even with a small physical gap.
When should a choke flange be used instead of a cover flange?
Use choke-to-cover for rotating joints, field connections, thermal cycling, or frequent disconnect/reconnect. Never mate two choke flanges (double grooves create a resonant cavity). For precision lab work with careful torquing, cover-to-cover with gaskets works, but choke-to-cover provides superior repeatability. Above 60 GHz, alignment pins become critical even with choke flanges.
What is the bandwidth limitation of a choke flange?
The groove is optimized for center frequency, so performance degrades off-center. Typical choke flanges maintain better than -40 dB return loss over 15 to 20% fractional bandwidth, matching standard waveguide single-mode range. Broadband designs with stepped or tapered grooves extend to 25 to 30%. At band edges, return loss is 30 to 35 dB versus 50+ dB at center.