Connector Frequency Range
How the Moding Limit Sets a Connector's Ceiling
A precision coaxial connector is, electrically, a short section of air-line that must present a single dominant mode across its rated band. That mode is the TEM mode, which has no low-frequency cutoff and supports a well-defined 50 Ω reference plane. The first parasitic mode able to propagate in coaxial geometry is the circular TE11 mode, and its cutoff is fixed by the physical dimensions of the inner and outer conductors. As long as the highest operating frequency stays below TE11 cutoff, the connector is single-mode and its electrical behavior is predictable. Cross that line and the interface becomes multi-moded, so any bead, gap, or conductor offset begins to radiate energy into the TE11 field.
This is why connector families come in a ladder of shrinking sizes. To raise the moding frequency you must shrink both conductors together, holding the 50 Ω ratio constant. The result is a clean inverse relationship: halving the line dimensions roughly doubles the mode-free ceiling. The 7mm APC-7 standard modes near 19 GHz; the 3.5mm interface lifts that to about 38 GHz; the 2.92mm, 1.85mm, and 1.0mm interfaces carry their air-line onset to roughly 45, 70, and 120 GHz, which vendors derate to the familiar 40, 67, and 110 GHz ratings. Each step down trades power handling, mechanical robustness, and cost for additional usable bandwidth.
The published "frequency range" on a datasheet is therefore a derated value, not the raw mode onset. A vendor specifies a few GHz of margin below TE11 cutoff so that bead resonances, plating roughness, and connector wear do not push spurious responses into the operating band. RF Essentials selects connector interfaces for millimeter-wave assemblies on exactly this basis: choose the largest interface whose derated ceiling clears the top of the operating band, because it maximizes power handling and durability without inviting moding.
Governing Cutoff Relations
λc ≈ π × (D + d) / 2
Mode-free upper frequency:
fmode ≈ c / λc = 2c / [π × (D + d)]
Derated rating with margin:
frated ≈ 0.90 × fmode
Where D = outer-conductor inner diameter, d = inner-conductor outer diameter (both in mm), c = 2.998 × 108 m/s. Example: a 2.92mm line with D ≈ 2.92 mm and d ≈ 1.27 mm gives λc ≈ 6.6 mm → fmode ≈ 45 GHz, derated to a 40 GHz rating.
Interface Frequency-Range Comparison
| Interface | Outer / Inner (mm) | Rated Range | Mode Onset | Impedance | Typical Use |
|---|---|---|---|---|---|
| Type-N | 7.00 / 3.04 | DC to 18 GHz | ~19 GHz | 50 Ω | Test gear, base stations |
| SMA | 4.13 / 1.27 | DC to 18 GHz | ~25 GHz (PTFE-loaded) | 50 Ω | General microwave |
| 3.5mm | 3.50 / 1.52 | DC to 26.5 GHz | ~38 GHz | 50 Ω | Precision, mates SMA |
| 2.92mm (K) | 2.92 / 1.27 | DC to 40 GHz | ~45 GHz | 50 Ω | Ka-band, mates SMA |
| 1.85mm (V) | 1.85 / 0.80 | DC to 67 GHz | ~70 GHz | 50 Ω | E/V-band |
| 1.0mm (W) | 1.00 / 0.43 | DC to 110 GHz | ~120 GHz | 50 Ω | W-band metrology |
Frequently Asked Questions
Why does a smaller coaxial connector reach a higher frequency?
The ceiling is the cutoff of the first higher-order mode, circular TE11, in the air line. Its cutoff wavelength is roughly π times the mean conductor radius, so shrinking both conductors proportionally (holding the 50 Ω ratio) raises the cutoff. A 7mm line modes near 19 GHz, 3.5mm near 38 GHz, 2.92mm near 45 GHz, 1.85mm near 70 GHz, and 1.0mm near 120 GHz. Vendors rate the connector a few GHz below onset to preserve return loss and repeatability.
Is an SMA connector really usable to 18 GHz?
SMA uses a 4.13mm / 1.27mm PTFE-supported interface. The PTFE bead pulls its first higher-order mode down to about 25 GHz, below the ~35 GHz an air line of the same dimensions would reach, and performance degrades earlier still. Production SMA is specified to 18 GHz, yet VSWR rises and repeatability falls above roughly 12 to 15 GHz due to the dielectric bead, tolerances, and wear. For clean metrology to 26.5 GHz use a 3.5mm connector; to 40 GHz use 2.92mm (K), both of which mate mechanically with SMA.
What happens to a connector above its rated frequency?
Below moding cutoff only TEM propagates, so the line stays a clean 50 Ω interconnect. Above TE11 cutoff, energy couples into the higher-order mode at every discontinuity (the bead, the mating plane, a conductor offset), producing resonant suckouts in S21, VSWR spikes, and non-monotonic return loss. Signal may still pass, but the reference plane is undefined, so the part is unfit for calibrated measurement or low-reflection use.