Connectors & Interconnects

Contact Flange

Q: When should I use a contact flange instead of a choke flange?

Use a contact flange when both mating faces are flat, clean, and can be clamped with enough torque to maintain reliable metal-to-metal contact across the full RF current path. Contact flanges are the better choice for fixed, low-VSWR connections in test setups and assemblies that are rarely demated, because a properly seated flat-face joint introduces very little reflection and almost no extra insertion loss. Choke flanges are preferred when the joint must be made and broken many times, when slight misalignment or contamination is likely, or when a small air gap may exist between faces. The quarter-wave choke groove on a choke flange transforms the gap impedance so that even an imperfect contact still presents a near short circuit to leakage current, tolerating gaps up to a few thousandths of an inch. The trade-off is that a choke flange is band-limited to roughly the design waveguide band, while a flat contact flange works across the full waveguide bandwidth.

Q: How much torque should I apply to a WR-28 contact flange?

For a WR-28 UG-style flange using 4-40 stainless steel screws, a typical specification is 5 to 8 inch-pounds per screw, applied in a balanced cross pattern so the faces seat evenly. Under-torquing leaves micro-gaps that raise contact resistance and RF leakage, while over-torquing can warp the thin flange face or strip the threads, which actually degrades flatness and contact. Larger waveguide bands such as WR-90 commonly use 6-32 screws at 8 to 14 inch-pounds, and WR-15 and smaller higher-frequency millimeter-wave bands may use precision alignment pins plus 0-80 or 2-56 screws at 2 to 4 inch-pounds. Always follow the connector or component manufacturer's torque table, and use a calibrated torque driver rather than feel, because consistent, repeatable clamping force is what keeps a flat-face joint within its specified insertion loss and return loss limits.

Q: What surface flatness and finish does a contact flange need?

A good contact flange face is typically held flat to about 0.0005 inch (12.7 micrometers) or better over the mating area, with a surface finish of 32 microinches RMS or smoother. Flatness matters because any local deviation creates an air gap that behaves like a small series impedance and a radiating slot, raising both insertion loss and leakage. Surface finish controls the true contact area at the microscopic level, since RF current flows in a skin depth that is only a fraction of a micrometer at millimeter-wave frequencies. Faces are usually gold plated over nickel to lower contact resistance, resist corrosion, and keep the joint repeatable over many mating cycles. Even tiny scratches, burrs, or particles across the waveguide aperture can cause measurable VSWR and leakage degradation, so contact faces are kept clean and protected with caps when not in use.

/KON-takt flanj/

A flat-face waveguide flange that joins two waveguide sections through direct metal-to-metal contact, carrying the surface current across the joint without a choke groove. Because the entire RF return path runs through the mated faces, performance depends almost entirely on face flatness, surface finish, plating, and clamping torque rather than on any tuned cavity. A clean, well-seated contact flange adds well under 0.05 dB of insertion loss and works across the full waveguide band, but it is intolerant of gaps, burrs, and contamination, which raise contact resistance and RF leakage. Contact flanges are common in fixed, low-VSWR assemblies and precision test setups from WR-90 down to millimeter-wave bands.

Category: Connectors & Interconnects

Face flatness: ≤ 0.0005 in

Added loss: < 0.05 dB

How a Flat-Face Waveguide Joint Carries RF Current

In a rectangular waveguide, RF energy travels as wall currents flowing on the inner conducting surfaces. When two waveguide sections are joined, those currents must cross the seam between the mated parts. A contact flange (often a UG-style flat-face or "cover" flange) makes that crossing through direct metal-to-metal contact: the two precision-machined faces are clamped together with screws, and the surface current flows continuously from one face into the other. There is no resonant cavity involved, so the joint behaves essentially as a continuous waveguide wall as long as the contact is intimate and uniform around the full aperture.

The quality of that contact is everything. At microwave and millimeter-wave frequencies the current is confined to a skin depth that is only a fraction of a micrometer, so the actual conduction happens at the microscopic high spots where the two faces truly touch. Any local separation acts as a small series impedance and a radiating slot, which shows up as increased insertion loss, degraded return loss, and measurable leakage outside the joint. This is why contact flanges specify tight face flatness (commonly 0.0005 inch or better), a smooth finish near 32 microinches RMS, and gold plating over nickel to lower resistance and keep the joint repeatable across many mating cycles.

Because the contact flange relies on physical pressure rather than a tuned structure, clamping torque and screw pattern matter as much as the machining. Screws are tightened in a balanced cross sequence so the faces seat evenly and the thin flange does not warp. Alignment pins or dowels keep the two waveguide apertures registered to a few thousandths of an inch, since lateral offset between the bores creates its own reflection. When assembled correctly, a contact flange is the lowest-loss, widest-bandwidth way to join two waveguide sections.

Estimating Joint Loss and Leakage

Skin depth (current penetration):
δ = 1 / √(π f μ σ)

Joint loss from contact resistance:
IL_joint ≈ 20 log₁₀(1 + R_c / 2Z₀) dB

Gap as a series reactance (small air gap g):
X_gap ≈ 2π f μ₀ g

Where f = frequency, μ = permeability, σ = conductivity, δ = skin depth, R_c = series contact resistance of the seam, Z₀ = waveguide wave impedance, μ₀ = free-space permeability, g = face gap. At 35 GHz in copper, δ ≈ 0.35 μm, so even a sub-micron gap intercepts most of the current path. Example: a marginal joint with R_c ≈ 1 Ω against Z₀ ≈ 500 Ω gives IL_joint ≈ 0.009 dB, and a 12.5 μm gap presents X_gap ≈ 3.5 Ω (a small but measurable series reactance). A clean, well-seated joint keeps R_c in the milliohm range, so its added loss is negligible.

Contact, Choke, and Cover Flange Comparison

Flange type	Mating mechanism	Gap tolerance	Bandwidth	Best for
Contact (flat-face)	Direct metal-to-metal	Very low (< 0.0002 in)	Full waveguide band	Fixed low-VSWR joints, precision test
Choke	Quarter-wave groove transforms gap	High (a few thousandths)	~Single waveguide band	Repeated demating, field assembly
Cover	Flat face, no groove (mates a choke)	Low	Full band	The flat half of a choke/cover pair
Gasketed contact	Conductive gasket fills micro-gaps	Moderate	Full band	Pressurized or high-leakage-sensitive

Common Questions

Frequently Asked Questions

When should I use a contact flange instead of a choke flange?

Use a contact flange for fixed, rarely demated joints where both faces are flat, clean, and clamped with adequate torque; a well-seated flat-face joint adds very little reflection or loss and works across the full waveguide band. Choose a choke flange when the joint is made and broken often, when small misalignment or a slight air gap is likely, or for field assembly, because its quarter-wave groove tolerates gaps up to a few thousandths of an inch but is band-limited.

How much torque should I apply to a WR-28 contact flange?

A WR-28 UG-style flange with 4-40 stainless screws is typically torqued to 5 to 8 inch-pounds per screw in a balanced cross pattern. WR-90 commonly uses 6-32 screws at 8 to 14 inch-pounds, while WR-15 and smaller higher-frequency bands use alignment pins and 0-80 or 2-56 screws at 2 to 4 inch-pounds. Under-torquing leaves micro-gaps that raise leakage and contact resistance; over-torquing warps the face. Always use a calibrated torque driver and the manufacturer's table. See torque specification.

What surface flatness and finish does a contact flange need?

Mating faces are usually held flat to about 0.0005 inch (12.7 micrometers) or better with a 32 microinch RMS or smoother finish, and gold plated over nickel. Flatness limits air gaps that act as radiating slots and series impedance; finish controls the true microscopic contact area where skin-depth current actually flows. Burrs, scratches, or particles across the aperture cause measurable VSWR and leakage degradation, so faces are kept clean and capped when not in use.

Related Terms

← Contact Discharge Contact Resistance →