Antenna Gain Is Not a PIM Cure—But Lower Conducted Power Can Help

Gain is a directional antenna quantity, not a measure of connector or material linearity. Trading gain for lower conducted power can reduce some PIM products, but it does not guarantee lower installed-site interference. Keep the ERP/EIRP budget, the parent-tone powers, the PIM source location and the receiver acceptance criterion separate.

Start with the Right Antenna and Power Quantities

IEEE 145-2025 is the active standard for antenna terminology. Antenna gain is direction-dependent and includes radiation efficiency; realized gain additionally accounts for input mismatch. Directivity describes pattern concentration without the same loss accounting. A higher number in one direction says nothing by itself about PIM generation.

ITU terminology also keeps ERP and EIRP directional:

Use dBi with EIRP and dBd with ERP. For the same system and direction, EIRP is 2.15 dB above ERP because the references differ. Here Gi and Gd are paired with power accepted by the antenna, while realized gains Gri and Grd include mismatch and are paired with incident port power. Choose one convention; do not omit mismatch or count it twice. Always declare the power and antenna reference planes.

The 50 W to 12.6 W calculation

Fifty watts is 46.99 dBm. If system gain in the direction of interest rises by exactly 6.00 dB while feed loss and every other term remain unchanged, transmitter power can fall by 6.00 dB while preserving ERP or EIRP in that direction:

A change from 6 dBd to 12 dBd preserves a chosen directional ERP with a 6 dB conducted-power reduction only if those are comparable gains at the operating frequency, loss and mismatch accounting is consistent, and the transmitter remains linear and stable at the new setting.

Equal Directional ERP Is Not Automatically Equal Coverage

Gain is produced by redistributing radiation, commonly by narrowing or tilting a pattern. Matching the old ERP in one elevation and azimuth does not preserve:

• close-in coverage beneath a narrower vertical lobe;
• coverage in sidelobes, nulls or other azimuths;
• polarization match, downtilt or pattern stability on the actual structure;
• near-field illumination of tower steel and rooftop hardware;
• uplink carrier-to-interference-plus-noise performance.

For a reciprocal passive antenna, the receive pattern follows the transmit pattern at the same frequency and state. That can strengthen desired uplink signals in some directions, but it can also strengthen noise or interference from those directions and weaken other users. “Same coverage” requires an agreed service-area metric and before/after field or link-budget evidence, not one peak-gain number.

What PIM Is—and Where It Can Be Generated

Passive intermodulation is generated when two or more strong signals encounter nonlinear behaviour in a nominally passive path or illuminated object. Contact interfaces, weak or contaminated connector joints, cable assemblies, filter interfaces, antenna feed networks and exposed metalwork can all be candidates. Receiver front-end intermodulation is a separate active-device mechanism and should not be labelled PIM.

For two parent frequencies, the familiar third-order products are:

A product matters when its spectrum and coupled power degrade the receiver at the actual receive frequency. It can be a stable line, a modulation-shaped spectrum or a transient noise-like burst. Calling every observation a “PIM noise floor” obscures the distinction among receiver random noise, the test system's residual intermodulation, and a PIM product from the device or site.

Absolute product level and receiver susceptibility are separate

Report a PIM product in dBm at a declared component output, antenna port or receiver-input plane. For a discrete product, record detector and resolution bandwidth; for a modulation-shaped or noise-like product, record the channel or integration bandwidth. Changing analyzer resolution bandwidth can change the displayed level of distributed energy without changing the physical source.

Receiver impact then depends on the product frequency and spectrum, duplexer and preselector rejection, receiver bandwidth, wanted-signal level, blocking and intermodulation behaviour, and the permitted degradation in SINAD, BER or another service metric. A product below one receiver’s acceptance limit can still desensitize another receiver with a different bandwidth or front-end state. Absolute dBm at the receiver input and measured wanted-signal degradation are therefore the final acceptance quantities; dBc at a component test plane is supporting evidence.

Component and site PIM are different boundaries

This is why “gain is a power discount coupon” is useful only as a planning metaphor. It can reduce conducted excitation upstream of the antenna. It does not establish the linearity of a replacement antenna, the behaviour of other transmitters or the illumination of external objects.

Read a −153 dBc PIM Claim Correctly

IEC 62037-1:2025 defines general PIM measurements using two transmitting signals. It requires the test to be fully specified, notes that PIM is usually frequency-dependent, requires the reported result to include the maximum measured over the test duration, and does not turn a test into a long-term reliability guarantee.

In the common equal-tone convention, dBc is relative to one test tone, not the sum of both tones. Kaelus gives the standard conversion example:

The two carriers together total +46.01 dBm, but that is not the reference for the −153 dBc figure. With unequal tones, the dBc convention must be declared. Absolute dBm is often clearer when comparing the product at the receiver with an interference allowance.

A defensible PIM specification includes at least the two frequencies, power per tone, IM order and product frequency, dBm or dBc reference, forward or reverse direction, port terminations, fixture residual level, sweep or fixed-tone plan, duration and maximum detector, static or dynamic stress, environmental state, return loss or VSWR, and reference planes. “Low PIM” without those fields is not a reproducible limit.

The 18 dB Improvement Is a Conditional Model Result

For an ideal memoryless cubic nonlinearity, the two IM3 powers scale as:

Keysight's two-tone polynomial derivation gives the familiar 3 dB change in absolute IM3 for every 1 dB change applied simultaneously to both equal parent tones. Under that model, lowering both tones by 6 dB lowers either IM3 product by 18 dB.

The data-sheet number does not “stay −153 dBc” in this cubic extrapolation. If one merely held dBc constant while lowering a carrier by 6 dB, absolute product power would fall only 6 dB.

What if only one co-site transmitter is reduced?

An 18 dB reduction does not apply when only one parent changes. In the same ideal model, lowering f1 by 6 dB lowers 2f1−f2 by 12 dB but lowers 2f2−f1 by 6 dB. Lowering only f2 swaps those results. A shared site must identify which transmitters create the product that lands in the receive channel.

Real passive junctions can show other slopes, hysteresis, bursts, temperature dependence and frequency dependence. IEC 62037 does not promise a universal cubic scaling law. Measure a power sweep—both tones together and, where useful, one at a time—without exceeding any device or test-system rating.

Specify the Component and the Installed Budget

The IEC 62037 family separates device classes because their fixtures and stresses differ. Current editions provide dedicated methods for coaxial connectors, cables, filters and antennas; Part 7 covers reverse field measurements of deployed systems; Part 8 covers objects exposed to radiated RF. A passing antenna certificate does not cover every jumper, connector, duplexer interface or nearby object.

Set a receiver-centred requirement rather than copying a generic −153 dBc threshold:

1. Define receiver performance. State the wanted-signal criterion and permitted degradation, such as SINAD or BER change, with bandwidth and receiver configuration.
2. Allocate an interference budget. Convert the permitted product at the receiver input into component or path limits using measured losses, coupling, combining and uncertainty.
3. List every parent combination. Include the repeater transmitter and relevant co-site transmitters; calculate IM3, IM5 and higher products that can fall in the receive band.
4. Specify each test. Include tone powers and frequencies, direction, maximum-over-time result, sweep, dynamic stress, residual floor, termination and environment.
5. Retest the assembled system. Installation introduces interfaces and illuminated objects absent from a component fixture.

Multiple PIM sources can combine by phase as well as power, so a simple sum of individual catalogue maxima is not necessarily the installed result. Conversely, a system can pass at one tone pair and fail elsewhere. IEC 62037-1:2025 explicitly highlights frequency dependence and the value of swept or multiple fixed-frequency testing.

A Repeater Test That Separates Cause from Correlation

1. Lock the reference configuration. Record transmitter powers, feed losses, antenna model, installed pattern assumptions, receiver settings, duplexer state, weather and active co-site carriers.
2. Measure the receiver baseline. With local transmitters off, measure the wanted-signal threshold using a defined modulation and quality criterion.
3. Measure operational desense. Key each local transmitter alone and in relevant combinations. Re-measure the wanted threshold rather than relying only on a spectrum display.
4. Identify candidate products. Calculate product frequencies and distinguish stable PIM, transient PIM, transmitter noise or spurs, carrier blocking and receiver-generated IM.
5. Characterize PIM safely. Use rated low-PIM test equipment and terminations, establish receiver noise and test-system residual IM, sweep relevant frequencies, capture the maximum over time and apply the specified dynamic stress.
6. Vary power deliberately. Reduce both parent tones together when possible, then one at a time, and record the observed slope. Do not infer a cubic law from a single before/after point.
7. Evaluate the antenna trade. Compare directional coverage and receiver performance, not only transmitter watts. Check close-in service, sidelobes and external-object illumination.
8. Localize by controlled substitution. Change one jumper, connector, load, antenna path or mechanical condition at a time and preserve calibrated before/after results.

Bottom Line

Antenna gain does not make passive hardware more linear. A 6 dB increase in comparable directional system gain can allow a 6 dB transmitter-power reduction while preserving ERP or EIRP in that direction. That can reduce conducted PIM excitation, heating, receiver blocking and some transmitter-related interference.

It does not guarantee the same coverage, an 18 dB reduction when only one parent tone changes, better PIM from the replacement antenna, or lower external PIM where illumination is unchanged. The useful engineering claim is conditional: trade verified directional gain for lower power only after identifying the parent signals and PIM source, then prove the result at the receiver and across the required service area.

Primary Standards and Measurement Sources

• IEEE 145-2025, Standard for Definitions of Terms for Antennas, active edition.
• ITU-R V.573-6, Radiocommunication Vocabulary, in-force source for directional ERP and EIRP terminology.
• IEC 62037-1:2025, general PIM requirements and measuring methods.
• IEC 62037-3:2025 for connectors; IEC 62037-4:2012+A1:2025 for cables; and IEC 62037-5:2021 for filters.
• IEC 62037-6:2021+A1:2025 for antennas; IEC 62037-7:2022 for field measurements; and IEC 62037-8:2025 for exposed objects.
• Kaelus PIM measurement guidance for dBm/dBc reference conventions, residual IM and frequency dependence.
• Keysight Spectrum Analysis Basics for the ideal two-tone cubic IM3 derivation. That derivation is a model boundary, not a universal PIM guarantee.

Questions, antenna-factor records or height trials to share? Contact RF.Guru.