Two strong transmit carriers meet a poor metallic junction, and the junction behaves like a diode. New frequencies appear that nobody generated, and if one lands in the receive band the site desensitises. Nothing has failed in the electronics; the problem is a corroded connector, a loose clamp, or rusty steel on the tower.
Two carriers at f₁ and f₂ pass through a connector. If the joint is perfectly linear, they emerge unchanged. If it is not — oxide between mating surfaces, contamination, insufficient torque, dissimilar metals — the junction is weakly non-linear, and a non-linearity mixes.
The products land at 2f₁−f₂ and 2f₂−f₁ for third order, and 3f₁−2f₂ for fifth. Those frequencies sit close to the originals, which is exactly the problem. In many band pairings the third-order product falls squarely in the receive band, where it cannot be filtered out because it arrives with the wanted signal.
The effect scales brutally with power. A third-order product rises roughly three decibels for every decibel of carrier increase, so a junction that was acceptable at one power level can become the dominant noise source when the site is upgraded. Adding a carrier to an existing antenna is a common way to discover PIM that was always there.
| Source | Why it generates PIM | What resolves it |
|---|---|---|
| Connector torque | Insufficient contact pressure leaves a high-resistance interface that rectifies | Torque to specification with a calibrated wrench; not by feel |
| Corrosion and oxide | Aluminum and copper oxides are semiconducting; the junction becomes a weak diode | Weatherproofing, correct sealing, replacement of degraded connectors |
| Dissimilar metals | Galvanic action at the interface, accelerating with moisture | Matched plating; avoid steel against aluminum in the RF path |
| Loose mechanical joints | Anything metallic in the near field that can vibrate — brackets, clamps, fencing | Tighten, bond or remove; rusty steel near the antenna is a classic cause |
| Cable damage | Crushed or kinked feeder deforms the braid and creates intermittent contact | Replace; a bend that violates the minimum radius does not recover |
| Ferrous material in the path | Nickel and steel are magnetically non-linear even when clean | Specify non-ferrous connectors and hardware for the RF path |
Roughly the order to check in. Torque and weatherproofing account for the majority of field cases, and both are installation practice rather than component quality.
| Term | What it means | Typical practice |
|---|---|---|
| dBc | Product level relative to the carrier | Stated against two carriers at a defined power, commonly 2 × 20 W (43 dBm) |
| −150 dBc | A common component requirement for connectors and antennas | Equivalent to about −107 dBm at 2 × 43 dBm test power |
| Third order (IM3) | 2f₁−f₂ and 2f₂−f₁ — the strongest and the usual offender | The product that most often lands in the receive band |
| Fifth order (IM5) | 3f₁−2f₂ — weaker but further from the carriers | Matters when band spacing puts IM3 outside the receive band |
| Dynamic PIM | Measurement while the assembly is tapped or flexed | Catches intermittent joints that a static test passes |
A site that passes a static test and fails in wind has dynamic PIM. Testing only when still is how that gets missed.
A weakly non-linear metallic junction carrying two or more strong transmit carriers. Insufficient connector torque, oxide or corrosion between mating surfaces, dissimilar metals, loose brackets or fencing in the near field, and damaged feeder are the common causes. No active component is involved, which is why it is called passive.
Because the intermodulation product can land inside the receive band. There it arrives with the wanted signal and cannot be filtered out, so the site desensitises: coverage shrinks, uplink throughput falls, and handsets transmit harder. The electronics report no fault.
The products at 2f₁−f₂ and 2f₂−f₁, where f₁ and f₂ are the transmit carriers. They are the strongest products and sit closest to the originals, which is why in many band pairings they fall directly in the receive band.
Sharply. A third-order product rises roughly three decibels for every decibel of carrier increase, so a junction that was acceptable becomes dominant when a site is upgraded or a carrier is added. That is a common way to discover PIM that was always present.
Typically 10 to 20 dB against a stable source. It predicts the products from the known transmit signal and subtracts them, so it needs the non-linearity to hold still — PIM that varies with wind or vibration moves faster than the canceller adapts. It refines a good site; it does not repair a bad one.
The intermodulation product is 150 dB below the carrier, measured at a stated test power — usually two carriers at 43 dBm. It is a common requirement for connectors and antennas, and equates to roughly −107 dBm of product at that test condition.
Because the test was static. An intermittent joint — a loose clamp, a cracked weather seal, corroded fencing — only misbehaves when it moves. Dynamic PIM testing taps or flexes the assembly during measurement and catches what a still test does not.
Written for engineers. Share it with one.
Where linearization happens.
KNOWLEDGEDPD, CFR and the amplifier.
TECHNOLOGYLink budget and the RF chain.
PRODUCTThe radio at the top of the mast.
KNOWLEDGEHow signal quality is measured.
INDUSTRYWhere this is deployed.
Tell us the specification, the constraint and the deadline. Programs that cross silicon, radio, embedded and AI are where Faststream is strongest.