Contact Resistance (EMC)
Why a Few Milliohms Decide Shield and PIM Performance
Two pieces of metal pressed together never touch over their full nominal area. Surface roughness limits real contact to a scattering of microscopic asperity spots, often a small percentage of the apparent area, so current is forced to constrict through these tiny conducting bridges. That constriction is the dominant component of contact resistance, and it falls as contact force rises and as the spots deform plastically. Surface films make it worse: a thin oxide, sulfide, or fluoride layer adds a tunneling or insulating series term that can swamp the metallic constriction resistance, which is why aluminum (with its tenacious self-limiting oxide) and corroded tin are poor EMC bonding surfaces while silver and gold plating are preferred.
For an EMC engineer the number itself matters less than its stability. A static milliohm of clean, well-pressured contact is harmless; the same milliohm in an oxide-laden, fretting, or under-torqued joint becomes current dependent and time varying. That nonlinearity is the seed of passive intermodulation, and it is also the mechanism behind intermittent shield leakage that defeats an otherwise well-designed enclosure. Because high contact pressure both lowers resistance and breaks through thin films, bonding hardware is specified by torque, gasket compression force, and finger-stock deflection rather than by resistance alone.
At RF, the relevant measure also shifts. A joint that reads a fraction of a milliohm on a DC bond tester can still leak through the seam at higher frequencies because current crowds onto the skin of the conductors and onto the same asperity bridges, so transfer impedance and measured PIM, not DC ohms, are the true acceptance criteria for high-performance hardware.
Governing Models
Rc ≈ ρ / (2a) Ω
Multi-spot cluster (n spots, two metals):
Rc ≈ ρ1 / (4na) + ρ2 / (4na) Ω
Spot radius vs. contact force:
a ≈ √( F / (π ξ H) ) (H = material hardness)
Third-order PIM from nonlinear contact:
V(I) = R0I + R2I2 + R3I3 + … → V3rd ∝ R3 I12 I2
(third-order product power rises ~3 dB per 1 dB of carrier power)
Where ρ = resistivity, a = effective contact-spot radius, n = number of conducting spots, F = contact force, ξ ≈ 0.3 to 1 (deformation factor), H = surface hardness, and R0, R2, R3 are the Taylor coefficients of the current-dependent contact resistance. Example: a single copper spot (ρCu ≈ 1.7 × 10-8 Ω·m) of 50 μm radius gives Rc = ρ/(2a) ≈ 0.17 mΩ.
Joint and Interface Comparison
| Joint / interface | Typical Rc | PIM risk | Primary mitigation | Where used |
|---|---|---|---|---|
| Silver-plated bolted flange | < 0.5 mΩ | Low | Spec torque, plating | Waveguide, antenna feeds |
| Beryllium-copper finger stock | 1 to 10 mΩ | Low to moderate | Set deflection range | Shielded enclosure seams |
| Conductive EMI gasket | 1 to 50 mΩ | Low to moderate | Controlled compression | Panel and door seams |
| Tin-plated connector | 1 to 10 mΩ | Moderate | Gold flash, retorque | General interconnect |
| Corroded / fretted joint | 10 to 1000 mΩ | Very high | Clean or replace | Field failure mode |
Frequently Asked Questions
What level of contact resistance is acceptable for an RF bonding joint?
MIL-STD-464 and aerospace bonding practice target a DC bond resistance under 2.5 mΩ between mating structures and under 1 mΩ for radio-frequency class bonds. A clean silver-plated bolted flange typically reads below 0.5 mΩ, while a tin-plated or anodized interface sits between 1 and 10 mΩ. Once a joint corrodes or loosens it climbs into the tens or hundreds of milliohms, at which point shield seams leak and ground loops form. Gasket seams are usually graded by transfer impedance or resistance per unit length rather than one lumped value.
How does contact resistance cause passive intermodulation?
A junction that behaves as a nonlinear, current-dependent resistor instead of a clean ohmic contact mixes two carriers at f1 and f2 and generates products at 2f1 − f2, 3f1 − 2f2, and so on. Because the third-order product power rises about 3 dB for every 1 dB of carrier power, even a fraction of a milliohm of unstable, oxide-laden contact can lift third-order PIM above the −150 dBc class limit and into a sensitive receive band. Stable, high-pressure, silver or gold plated metal-to-metal contact is the primary defense.
How is EMC contact resistance measured?
DC bond resistance is measured with a four-wire Kelvin (Kelvin clip) method on a milliohmmeter or DLRO, forcing 100 mA to 10 A through one lead pair while sensing voltage on a separate pair to cancel lead and probe resistance. Microohm-range bonds need a dedicated micro-ohmmeter. At RF the acceptance criteria shift to transfer impedance, measured with a vector network analyzer, and to PIM level, measured by a two-carrier PIM analyzer reporting reverse and forward intermodulation products.