Electronics Guide

Microscopic Contact Theory

When two metal surfaces meet, they do not make continuous contact across their entire area. Instead, contact occurs at discrete points where surface asperities (microscopic peaks and valleys) meet. The real contact area is typically only a small fraction of the apparent contact area, often less than 1% even under substantial pressure.

At these micro-contacts, several phenomena contribute to nonlinear behavior:

• Constriction resistance: Current crowding into a spot of radius a adds a constriction resistance that Holm's theory gives as ρ/(2a) for a single circular spot, typically a few milliohms. Constriction resistance is not nonlinear in itself. It becomes nonlinear because the same current heats a volume only micrometers across, and the temperature rise changes the local resistivity, so the resistance follows the signal instead of staying fixed.
• Tunneling: When contact gaps are extremely small, on the order of nanometers, quantum mechanical tunneling allows current flow even without metallic contact. Tunneling current falls exponentially with gap distance, which makes the junction strongly nonlinear; classical contact theory places the practical limit for measurable tunneling near 10 nm, or 100 angstroms, of film.
• Thermionic emission: At elevated temperatures, electrons can be thermally excited over the potential barrier at a contact interface rather than tunneling through it. The applied field lowers that barrier through the image force, and the resulting field-dependent emission is the Schottky effect.

The total contact resistance can be modeled as a combination of constriction resistance and film resistance (due to oxide or contamination layers), with each component exhibiting its own nonlinear characteristics.

Contact Force Effects

The force pressing two contact surfaces together profoundly affects PIM performance. Higher contact forces increase the real contact area by elastically and plastically deforming asperities, reducing both contact resistance and nonlinearity.

The relationship between contact force and resistance follows a power law:

R_contact ∝ F⁻ⁿ

The exponent follows from Ragnar Holm's classical contact theory, in which the constriction resistance of a single circular spot is ρ/(2a). For purely elastic, Hertzian deformation the spot radius grows as the cube root of the load, so n is approximately 0.33. For fully plastic deformation the real contact area is the load divided by the indentation hardness, the radius grows as the square root of the load, and n approaches 0.5. Real connector interfaces deform partly elastically and partly plastically and sit between those values, and the effective exponent can exceed 0.5 while the force is still breaking through a surface film.

This force dependence is not itself the PIM mechanism. A resistance that depends on a static force generates no intermodulation while the force stays static. What the relationship does is set the operating point: a joint held at low force starts with a small real contact area, a high current density in each spot, and an intact surface film, and every current-driven mechanism described below is far stronger under those conditions. Force that varies during operation, from vibration, thermal expansion, or a fastener working loose, converts the same relationship into a directly time-varying resistance.

For threaded RF connectors, proper torque specifications ensure adequate contact force. Under-torqued connections exhibit higher PIM due to insufficient contact pressure, while over-torqued connections can damage contact surfaces or cause permanent deformation that degrades long-term performance.

Multi-Point Contact Dynamics

Real contact surfaces involve many simultaneous contact points, each with its own resistance and nonlinear characteristic. The aggregate behavior depends on how these contacts are distributed and how they respond to current flow and mechanical stress.

Under RF excitation, the current distribution among contact points varies with the instantaneous voltage and current, as each contact has a slightly different voltage-current characteristic. This current redistribution occurs on the timescale of the RF signal, creating the mixing action that produces intermodulation products.

Mechanical vibration causes contact points to make and break cyclically, with some contacts opening while others close. This dynamic contact behavior can create severe PIM, particularly at frequencies corresponding to the mechanical resonances of the contact assembly.

Electro-Thermal Modulation

Every contact mechanism shares one amplifier: temperature. A micro-contact dissipates the whole constriction loss in a volume micrometers across, and the resistivity of the metal rises with the temperature that results. The resistance therefore tracks the instantaneous power in the joint, and a resistance that varies with the signal is by definition a mixer. This electro-thermal mechanism explains why a contact whose resistance is perfectly reproducible under direct current can still generate intermodulation under RF drive.

The mechanism carries a distinctive frequency signature, because heat diffuses far more slowly than an RF cycle passes. A contact spot of radius a settles in a time near a²/α, where α is the thermal diffusivity of the metal, about 1.2 × 10⁻⁴ m²/s for copper. A spot 1 µm in radius therefore settles in roughly 9 ns and one 10 µm in radius in roughly 0.9 µs, giving corner frequencies near 18 MHz and 185 kHz respectively. Against a 2 GHz carrier, whose period is 0.5 ns, the spot cannot respond at all. Against the beat between two carriers, whose period is set by their separation, it responds whenever that separation falls below the corner.

Electro-thermal PIM consequently behaves as a low-pass response in tone separation: it is strongest for closely spaced carriers and weakens as the carriers move apart. Two practical consequences follow. A measurement taken at one tone separation does not predict the result at another while this mechanism is active, which is one reason swept and multi-carrier tests disagree with fixed two-tone tests. And the mechanism requires no junction at all, since any conductor whose resistivity varies with temperature will serve, which is why printed antennas and thin-film transmission lines show a PIM floor even when every joint in the assembly is sound.

Ferromagnetic Materials

Ferromagnetic materials (iron, nickel, cobalt, and their alloys) are among the most problematic for PIM because their magnetic permeability varies with magnetic field strength. This nonlinear permeability causes nonlinear inductance in any component containing ferromagnetic material in the RF current path.

The physical mechanism involves magnetic domain behavior. In ferromagnetic materials, atomic magnetic moments are organized into domains with uniform magnetization. An applied magnetic field causes domain walls to move and domains to rotate, processes that are inherently nonlinear and hysteretic. The B-H curve (magnetic flux density versus magnetic field intensity) of ferromagnetic materials is fundamentally nonlinear, with the slope (permeability) varying with field strength.

Even trace amounts of ferromagnetic material can cause significant PIM. Steel screws, nickel plating, or ferromagnetic contamination on contact surfaces can introduce nonlinearity into otherwise low-PIM assemblies. For this reason, low-PIM designs specify non-ferromagnetic materials throughout the RF path, including:

• Brass, bronze, or beryllium copper for connector bodies and contacts
• Austenitic stainless steel of low magnetic permeability, such as annealed 316L, and only for hardware that carries no RF current. Cold working from machining, bending, or thread rolling transforms austenite into martensite and raises the permeability, so a nominally non-magnetic fastener may still test magnetic after fabrication
• Silver, white bronze, or gold plating applied without a nickel underplate
• Aluminum or copper for cable shields and structural elements

The magnitudes explain why this is treated as an absolute rule rather than a preference. Field experience with cellular hardware places the PIM contribution of stainless steel in the RF path at roughly 10 dB to 20 dB, and that of nickel plating, including gold plated over a nickel underplate, at roughly 20 dB to 40 dB. No amount of care with torque or cleanliness recovers 30 dB, which is why plating specifications for low-PIM connectors exclude nickel even as an intermediate layer.

Ferrites are a separate case, since they are ferrimagnetic rather than ferromagnetic and their magnetic behavior is the reason they are present at all. Isolators, circulators, resonators, and phase shifters cannot avoid it. Such parts can be optimized for low PIM, but they commonly produce more PIM than any other passive component in an antenna system, and nominally identical units often differ from one another. The PIM test is accordingly used on them as a production screen, principally for cracked ferrite.

Semiconductor Effects

Semiconductor junctions are intentionally nonlinear and are used in mixers and detectors to produce intermodulation. However, unintentional semiconductor junctions can form in RF systems and cause PIM.

Metal-oxide-metal (MOM) structures can behave as semiconductor junctions when the oxide layer is thin enough for tunneling or when it contains metallic inclusions. These structures exhibit a voltage-current characteristic similar to a diode, with exponential or power-law dependence that produces strong intermodulation.

Contamination with semiconductor materials (carbon, silicon compounds, or metal oxides with semiconductor properties) can also create PIM sources. Proper cleaning and handling procedures are essential to prevent such contamination.

Dielectric Nonlinearity

Most dielectric materials used in RF applications exhibit some degree of nonlinearity in their permittivity, though this effect is typically much weaker than contact or ferromagnetic nonlinearity. The electric field dependence of permittivity arises from the nonlinear response of molecular dipoles to applied fields.

Ferroelectric materials show particularly strong dielectric nonlinearity and should be avoided in low-PIM applications. Some ceramic dielectrics, particularly those with high permittivity, may exhibit measurable nonlinearity at high power levels.

For low-PIM applications, stable dielectric materials such as PTFE (Teflon), polyethylene, and air are preferred. These materials have low permittivity and exhibit minimal nonlinearity over typical operating conditions.

Magnetic Hysteresis

Ferromagnetic materials exhibit hysteresis in their magnetization curves. When an alternating magnetic field is applied, the B-H relationship traces a hysteresis loop rather than a single-valued curve. The area of the hysteresis loop represents energy dissipated as heat during each cycle.

More importantly for PIM, the hysteretic behavior means that the material's response to a magnetic field depends on its magnetization history. When multiple RF signals are present, the magnetization state varies with the combined instantaneous field, causing the permeability (and hence inductance) to be modulated by all signals present. This modulation produces intermodulation products.

The severity of hysteretic PIM increases with signal amplitude, because larger field excursions trace larger portions of the nonlinear B-H curve. The rate of that increase deserves a precise statement, because the familiar figure is quoted far more often than the condition it assumes.

Take a cubic nonlinearity driven by two carriers of amplitude A₁ and A₂. Cubing the sum produces a term 3A₁²A₂cos²(ω₁t)cos(ω₂t), and that is the only term capable of yielding 2f₁ − f₂. Reducing the product to a sum leaves a coefficient of (3/4)A₁²A₂ on cos((2ω₁ − ω₂)t), so the product amplitude carries the square of one carrier and the first power of the other, and the product power goes as P₁²P₂. The exponents are what set the slope:

• Raise both carriers together by 1 dB and the product rises by 2 + 1 = 3 dB. This is the familiar 3 dB per dB figure, and raising both carriers together while holding them equal is the condition every standard PIM test applies.
• Raise f₁ alone by 1 dB and the 2f₁ − f₂ product rises by only 2 dB.
• Raise f₂ alone by 1 dB and it rises by only 1 dB.

Quoting 3 dB per dB without naming the condition is therefore a real ambiguity, and a single-carrier power change produces a result that the figure does not describe. One further conversion is worth keeping straight: expressed in dBc rather than dBm, the same product degrades by 2 dB for every 1 dB of equal carrier increase, because the reference carrier rises along with it.

Measured slopes seldom match the ideal. Published field experience with cellular hardware places most real PIM sources between roughly 2.2 dB and 2.8 dB per dB of equal carrier increase, below the theoretical 3, and the value also drifts with frequency and with the mechanical history of the joint. A slope steeper than 3 dB per dB is the exception, and the documented cause is not ferromagnetic material but micro-arcing across a very small gap or between lightly touching surfaces, a threshold mechanism that switches on abruptly once the field is high enough and then climbs far faster than any polynomial. The practical value of a power sweep is therefore less in naming the mechanism than in warning the tester that extrapolating a low-power result to a higher operating power is unreliable.

Magnetic Domain Dynamics

At the microscopic level, hysteresis arises from the behavior of magnetic domains. Domain walls (boundaries between regions of different magnetization direction) move under applied fields, but their motion is impeded by defects, grain boundaries, and stress in the material. This pinning and sudden release of domain walls (Barkhausen effect) creates discrete jumps in magnetization that contribute to nonlinearity.

The frequency dependence of ferromagnetic nonlinearity is complex. At low frequencies, domain walls can follow the applied field relatively well. At higher frequencies, eddy currents in the conductive magnetic material limit domain wall motion, and eventually the material cannot respond at all. However, even at frequencies where bulk magnetization cannot change, local magnetic fluctuations at grain boundaries and defects can still cause nonlinearity.

Temperature also affects domain dynamics. Near the Curie temperature (where ferromagnetism disappears), domain behavior becomes highly nonlinear. Even at normal operating temperatures, thermal fluctuations contribute to the stochastic nature of domain wall motion.

Identifying Ferromagnetic Contamination

Ferromagnetic contamination can be difficult to detect visually. Small particles of steel from cutting tools, iron-containing solder flux residue, or nickel from plating processes can introduce PIM even when invisible to casual inspection.

Techniques for identifying ferromagnetic contamination include:

• Magnetic inspection: A small magnet or magnetic indicator paper can detect ferromagnetic particles on surfaces and in assemblies.
• Permeability measurement: Instruments that measure relative permeability can identify materials with magnetic properties.
• X-ray fluorescence (XRF): Elemental analysis can detect iron, nickel, or cobalt contamination.
• PIM power sweep: Sweeping the carriers, raised together and held equal, which is the condition the 3 dB per dB figure assumes, separates a threshold mechanism such as micro-arcing, whose slope exceeds 3 dB per dB, from ordinary contact and material nonlinearity, whose measured slope usually falls between 2.2 dB and 2.8 dB per dB. Ferromagnetic contamination announces itself by the sheer magnitude of the PIM rather than by an anomalous slope, so a power sweep supports the magnetic and elemental tests above rather than replacing them.

Oxide Formation and Properties

All metals oxidize to some degree when exposed to oxygen and moisture, though oxidation rates and oxide properties vary widely. Oxide layers range from a few nanometers (self-limiting oxides like aluminum oxide) to much thicker films that continue growing over time.

The electrical properties of oxide layers depend on the metal and oxide thickness:

• Films thinner than about 2 nm: electrons tunnel straight through, and the junction conducts with a resistance that falls exponentially as the film thins
• Films of about 2 nm to 10 nm: tunneling continues but the tunnel resistance climbs steeply with thickness. Classical contact theory places the practical limit for measurable tunneling near 10 nm, or 100 angstroms, and semiconducting oxides in this range can rectify as well
• Films thicker than about 10 nm: the film is effectively insulating, and conduction resumes only when the voltage breaks it down or when mechanical force punches through it

For PIM, the most problematic regime is the one in which the oxide thickness still allows some conduction through tunneling or semiconductor mechanisms, creating a nonlinear junction that behaves much like a diode. A film thick enough to shut conduction down is a better neighbor than one just thin enough to permit it.

Thick films have a failure mode of their own. A film too thick to tunnel through can still break down under a modest applied voltage, a process Holm named fritting: A-fritting opens a fresh conducting channel through the film, while B-fritting widens a channel that already exists. Fritting is abrupt, partly irreversible, and accompanied by metal transfer between the surfaces, so a joint that frits under RF drive produces bursts of wideband interference rather than a steady intermodulation product, sometimes recurring every few seconds. Micro-arcing across a small gap behaves in much the same way, and the two together account for PIM that appears and disappears on a timescale of seconds and that rises faster than 3 dB per dB on a power sweep.

Metal-Oxide-Metal Junctions

When two oxidized metal surfaces are pressed together, the resulting structure is a metal-oxide-metal (MOM) junction. Its current-voltage characteristic is nonlinear, but several distinct conduction mechanisms compete inside it, and which one dominates depends on the barrier thickness, on the voltage across it, and on temperature. Each mechanism has its own law and its own diagnostic plot, and naming them separately matters because they differ enormously in how much intermodulation they produce.

Below roughly 3 nm the electrons pass through the barrier directly. The current is close to proportional to voltage at small bias and rises superlinearly above it, and it falls exponentially with thickness, so a film growing from 1 nm to 3 nm changes the junction resistance by orders of magnitude. Once the voltage across the barrier exceeds the barrier height expressed in volts, roughly 1 V to 3 V for the oxides found on connector metals, the barrier becomes triangular and Fowler-Nordheim tunneling takes over:

I = CV²e^−D/V

The signature of this regime is that log₁₀(I/V²) plotted against 1/V is a straight line. Fowler-Nordheim tunneling matters for PIM because an exponential in 1/V is a far stronger nonlinearity than any polynomial, and because a connector carrying two 20 watt carriers can readily develop several volts across a film a few nanometers thick.

When electrons instead acquire enough thermal energy to pass over the interface barrier rather than through it, the mechanism is Schottky emission, and lowering of the barrier by the image force gives:

I = Ae^B√V

A closely related law applies when the carriers are released not at the interface but from traps in the bulk of the film, which is the Poole-Frenkel mechanism:

I = AVe^B√V

The two are easy to confuse and are routinely reported under the wrong name. They are told apart by which quantity is linear against √V: log₁₀(I) for Schottky emission, log₁₀(I/V) for Poole-Frenkel. The Poole-Frenkel coefficient is also twice the Schottky one for the same permittivity, because the trap is not screened by an image charge. Both are thermally activated, so both strengthen as the contact self-heats, which couples them to the electro-thermal mechanism described earlier.

In thicker films, where injected carriers accumulate and limit their own flow, the Mott-Gurney law for space-charge-limited conduction applies and the current becomes a square law with no exponential at all:

I = KV²

A pure square law is worth singling out, because it generates only even-order products. On its own it produces f₁ + f₂ and 2f₁ but not the 2f₁ − f₂ product that PIM testing looks for; the third-order products come from odd-order terms, which the exponential mechanisms supply in abundance. The practical conclusion is that tunneling through a thin film, whether direct or Fowler-Nordheim, is the regime that produces severe PIM, and that the severity depends as much on how large the oxide resistance is compared with the surrounding circuit impedance as on the shape of the law itself.

Aluminum is particularly problematic because its oxide (Al₂O₃) forms immediately upon exposure to air and is extremely stable and hard. Aluminum contacts require special treatment to achieve low PIM, including:

• Plating with silver, gold, or tin to prevent oxide formation
• Using gas-tight connections that exclude oxygen
• Applying sufficient force to mechanically penetrate oxide layers

Preventing Oxide-Related PIM

Several strategies minimize oxide-related PIM:

• Noble metal plating: Gold and silver resist oxidation and maintain low-resistance contacts over time. Gold is preferred for critical applications because silver can tarnish (form silver sulfide) in certain environments.
• Hermetic sealing: Excluding oxygen prevents oxide growth after assembly. This approach is used for some high-reliability RF connectors and components.
• Wiping action: Some connector designs make the contacts slide across each other during mating, mechanically breaking through oxide layers and exposing fresh metal. The term for this is wiping, and it should not be confused with fretting, which is the opposite: a degradation mechanism in which small oscillatory motion under vibration grinds oxide debris into the interface and raises contact resistance over time. Wiping must be balanced against wear for connectors that undergo many mating cycles.
• High contact force: Sufficient force can break through thin oxide layers, establishing metal-to-metal contact through the oxide film. This requires careful mechanical design to ensure adequate force over the component's lifetime.

Organic Contamination

Organic materials on contact surfaces can cause PIM through several mechanisms:

• Insulating films: Oils, greases, and organic residues can form thin insulating layers that create MOM-like junctions when contacted
• Carbon formation: Some organic materials carbonize under high current density or elevated temperature, creating semiconductor-like deposits
• Moisture absorption: Organic contamination can absorb moisture, creating ionic conduction paths with nonlinear characteristics

Sources of organic contamination include fingerprints (skin oils), lubricants from manufacturing processes, adhesive residues, and atmospheric pollutants. Even brief handling without gloves can deposit enough contamination to degrade PIM performance.

Cleaning with appropriate solvents (isopropyl alcohol, acetone, or specialized RF contact cleaners) removes organic contamination. The cleaning process must avoid leaving residues and should be followed by low-PIM assembly procedures.

Metallic Contamination

Metallic particles on contact surfaces can cause PIM even in small quantities:

• Ferromagnetic particles: Iron, nickel, or steel particles from cutting tools, filing, or environmental sources introduce magnetic nonlinearity
• Loose metallic debris: Conductive particles can create intermittent contacts that make and break under vibration, causing severe PIM
• Dissimilar metal particles: Particles of different metals than the contact surfaces can create galvanic cells that promote corrosion, or can form unstable junctions with nonlinear characteristics

Prevention of metallic contamination requires clean manufacturing environments, proper tool selection (avoiding steel tools for low-PIM assembly), and cleaning procedures that include inspection and particle removal.

Environmental Contamination

Outdoor installations are exposed to various environmental contaminants that can cause or worsen PIM:

• Salt spray: Chloride ions promote corrosion of most metals, degrading contact surfaces
• Industrial pollutants: Sulfur compounds cause silver tarnishing; acidic pollutants attack contact surfaces
• Biological growth: Fungi and bacteria can colonize connector surfaces in humid environments
• Sand and dust: Abrasive particles can damage contact surfaces and introduce contamination

Weatherproof connector designs with appropriate sealing and corrosion-resistant materials mitigate environmental contamination. Regular inspection and maintenance are essential for installations in harsh environments.

Stress and Strain Effects

Mechanical stress in metals can affect their electrical properties in several ways:

• Magnetomechanical effects: In ferromagnetic materials, mechanical stress changes magnetic domain structure and hence magnetic properties (magnetostriction and the Villari effect). Even non-ferromagnetic metals may contain ferromagnetic inclusions that exhibit these effects.
• Piezoelectric effects: Some materials generate electric fields under mechanical stress. While pure metals are not piezoelectric, certain oxides and compounds that might be present as contamination or corrosion products can exhibit piezoelectric behavior.
• Contact stress: The stress distribution at contact surfaces affects the contact area and hence the nonlinearity of the contact resistance.

Residual stress from manufacturing processes (cold working, heat treatment, welding) can create latent PIM problems that may not be apparent in initial testing but manifest later as stress relaxation occurs.

Vibration and Motion

Mechanical vibration is a significant cause of PIM, particularly in mobile or outdoor installations:

• Loose hardware: Screws, nuts, and other fasteners that are not properly torqued can move under vibration, creating intermittent contacts
• Cable movement: Cables that can flex near connectors may cause connector contacts to shift, changing contact resistance
• Structural resonances: At frequencies matching mechanical resonances, vibration amplitudes can become large enough to cause severe PIM

PIM from vibration often appears as modulation sidebands around the intermodulation products, with the sideband spacing corresponding to the vibration frequency. This characteristic can help identify vibration-induced PIM during troubleshooting.

Mitigation approaches include proper torquing of all hardware, strain relief for cables, vibration damping, and design changes to shift mechanical resonances away from likely excitation frequencies.

Mechanical Damage

Physical damage to RF components typically degrades PIM performance:

• Scratches and gouges: Damage to contact surfaces increases contact resistance and may expose reactive base metals
• Deformation: Bent or deformed connectors may not make proper contact
• Thread damage: Damaged threads prevent proper mating and torquing of threaded connectors
• Cracked solder joints: Fractures in solder connections create nonlinear junctions

Surface condition matters at a scale that is easy to underestimate. In one published experiment, abrading a 1 mm copper center conductor across its axis raised PIM by 13 dB to 22 dB at approximately 1.5 GHz and 44 dBm, and simply cleaning solder flux from the end of the same wire reduced PIM by 10 dB. Neither change would register in a return-loss or insertion-loss measurement, which is the central practical point: PIM is sensitive to surface and junction detail that every other routine RF test ignores.

Inspection for mechanical damage should be part of any PIM troubleshooting procedure. Damaged components should be replaced rather than repaired, as repair processes often fail to restore original PIM performance.

Thermal Expansion

Different materials expand at different rates with temperature. When materials with different coefficients of thermal expansion (CTE) are joined, temperature changes create stress at the interface:

• Connector interfaces: If the center conductor and outer conductor expand at different rates, contact pressure can vary with temperature
• Solder joints: CTE mismatch between solder, copper, and substrates can stress solder joints
• Cable-connector junctions: Differential expansion between cable dielectric, conductor, and connector materials can affect contact quality

Temperature cycling can cause cumulative degradation as repeated stress cycles lead to fatigue, work hardening, or progressive loosening of mechanical connections.

Temperature-Dependent Material Properties

The intrinsic properties of materials vary with temperature in ways that affect PIM:

• Oxide layer properties: The conductivity of metal oxides is often strongly temperature dependent, changing the characteristics of MOM junctions
• Ferromagnetic properties: Magnetic permeability and coercivity vary with temperature; ferromagnetic materials become less nonlinear as temperature approaches the Curie point
• Contact resistance: Both the bulk resistance of contact materials and the physics of micro-contacts vary with temperature

These temperature dependencies make it important to test PIM performance over the full operating temperature range, not just at room temperature.

Self-Heating

High RF power can cause significant self-heating in RF components, particularly at high-resistance contacts or in lossy materials. This self-heating can change PIM performance during operation:

• Transient behavior: PIM may be different immediately after power-up compared to thermal equilibrium
• Power cycling effects: Repeated heating and cooling from power cycling can cause thermal fatigue
• Hot spots: Localized heating at high-resistance contacts can cause material damage or change contact characteristics

In high-power applications, thermal design must ensure that component temperatures remain within acceptable limits to prevent thermally induced PIM degradation.

Short-Term Variations

PIM can vary on short time scales due to:

• Thermal transients: As components warm up under power, PIM may change over seconds to minutes
• Vibration: Mechanical vibration causes PIM fluctuation at the vibration frequency
• Environmental changes: Rapid temperature or humidity changes can cause short-term PIM variations
• Intermittent contacts: Marginal connections may make and break irregularly, causing random PIM bursts

Short-term variations can make PIM difficult to measure accurately. Test procedures typically specify stabilization times and averaging periods to obtain consistent results.

Aging and Degradation

PIM performance typically degrades over time due to:

• Oxide growth: Continued oxidation of contact surfaces increases oxide-related PIM
• Corrosion: Environmental exposure causes progressive degradation of contact surfaces
• Wear: Connectors that undergo repeated mating cycles experience contact wear
• Stress relaxation: Springs and pressure contacts may lose force over time, reducing contact pressure
• Contamination accumulation: Environmental contaminants accumulate on exposed surfaces

Periodic PIM testing can track aging trends and identify components that need maintenance or replacement before they cause system problems.

Latent PIM

Some PIM sources may not be apparent initially but develop over time:

• Hidden corrosion: Corrosion may develop in concealed areas (inside connectors, under cable jackets) before becoming visible
• Fatigue cracks: Mechanical fatigue from vibration or thermal cycling can create cracks that develop into PIM sources
• Contamination migration: Contaminants may migrate to critical contact areas over time
• Galvanic corrosion: Dissimilar metal contacts may undergo accelerating corrosion that worsens PIM over time

Latent PIM makes initial qualification testing insufficient to guarantee long-term performance. Environmental stress screening and accelerated life testing help reveal potential latent PIM sources.

Cable PIM

Coaxial cables can generate PIM through several distributed mechanisms:

• Shield construction: Braided shields have numerous contact points between wire strands that can be nonlinear; solid or semi-rigid shields have lower distributed PIM
• Center conductor joints: Where center conductor sections are joined, nonlinear contacts may exist
• Dielectric variations: Inhomogeneities in cable dielectric can create localized field concentrations
• Cable flexing: Bending cables causes mechanical stress that can vary contact conditions in the shield

Low-PIM cables use solid or corrugated outer conductors, carefully controlled manufacturing, and materials selected for linearity. Even low-PIM cables can exhibit elevated PIM if bent too sharply or damaged during installation.

Antenna PIM

Antennas are potential sources of distributed PIM because they contain extended metallic structures and multiple junctions:

• Feed networks: Antenna feed systems contain numerous connections and junctions that can cause PIM
• Radiating elements: The antenna elements themselves may have joints or contacts that contribute PIM
• Environmental exposure: Outdoor antennas are subject to contamination, corrosion, and mechanical stress
• Metal objects in the near field: Reflectors, mounting hardware, and nearby structures can contribute to PIM

Low-PIM antenna designs minimize joints in high-current areas, use appropriate materials, and incorporate protective measures against environmental degradation.

Structural PIM

The physical structure around an RF system can contribute distributed PIM:

• Metal contacts in the near field: Any metallic junction illuminated by significant RF fields can produce PIM that couples back into the system
• Rusty or corroded structures: Corroded steel illuminated by a transmitter produces the effect known in the literature as the rusty-bolt effect. Rust is a mixture of iron oxides and oxyhydroxides that commonly includes magnetite, Fe₃O₄, which is ferrimagnetic and semiconducting. Experimental work cautions against assuming the corroded oxide is the whole story: the dominant contributor is often the loose, small-area contacts that corrosion leaves behind, which is why bonding or tightening a rusted joint sometimes helps more than cleaning it
• Non-metallic conductors: Carbon fiber, conductive paints, and other materials with marginal conductivity can exhibit nonlinear behavior

Identifying structural PIM sources requires systematic investigation, often using distance-to-PIM measurements to locate the responsible structure. Remediation may involve improving or eliminating metallic junctions, or reorienting antennas to avoid illuminating problematic structures.