Intermodulation is what happens when two or more signals pass through a system that isn’t perfectly linear, and the system creates new frequencies that weren’t in the original signals. These new frequencies are combinations of the originals: their sums, their differences, and multiples of those. It’s a fundamental concept in electronics, audio, and telecommunications, and it shows up in places you might not expect, including the human ear.
How New Frequencies Get Created
Every electronic component, from an amplifier to a cable connector, is supposed to pass signals through unchanged. In reality, no component is perfectly linear. When a system has even slight nonlinearity, it doesn’t just amplify or transmit the signals it receives. It also multiplies them against each other, producing new frequencies at mathematical combinations of the inputs.
If you feed two tones into a nonlinear system, one at frequency f1 and another at f2, the system generates new tones called intermodulation products. You can predict exactly where these products will appear using a simple formula: M × f1 ± N × f2, where M and N are whole numbers (0, 1, 2, 3, and so on). The “order” of each product is M + N.
Second-order products land at f1 + f2, f2 − f1, 2×f1, and 2×f2. Third-order products appear at 2f1 − f2, 2f2 − f1, 2f1 + f2, and 2f2 + f1. Higher orders exist too, but they’re progressively weaker. The third-order products at 2f1 − f2 and 2f2 − f1 cause the most trouble because they fall very close to the original frequencies. If f1 and f2 are near each other, those third-order products land right in the same frequency band, making them nearly impossible to filter out.
Why Third-Order Products Matter Most
Second-order products (sums and differences) often land far from the original signals. A system operating in a narrow frequency range can usually filter them away. But the third-order difference products sit just above and below the original tones. Imagine two radio signals at 100 MHz and 101 MHz: the third-order products show up at 99 MHz and 102 MHz, right in the middle of the operating band. No filter can separate them from the desired signals without also removing the signals themselves.
The power of third-order intermodulation products grows rapidly with input power. For every 1 dB increase in input signal strength, third-order products increase by 3 dB. This cubic relationship means the problem gets worse fast as you push a system harder. Engineers use a metric called IP3 (third-order intercept point) to rate how well a component resists intermodulation. IP3 is the theoretical power level where the unwanted third-order products would equal the strength of the desired signal. A higher IP3 number, measured in dBm, means better linearity and less distortion. It’s one of the most important specifications for amplifiers, mixers, and receivers in radio systems.
How It Differs From Harmonic Distortion
Harmonic distortion and intermodulation distortion both come from nonlinearity, but they behave differently. Harmonic distortion happens with a single input signal: the system creates copies at exact multiples of the original frequency (twice, three times, four times the frequency, and so on). You measure it by feeding in one pure tone and looking at what extra tones appear at the output. Total harmonic distortion, or THD, combines all these unwanted harmonics into a single percentage.
Intermodulation distortion requires at least two input signals. The system creates new tones at combinations of those inputs, not just multiples of one. This is a harder problem because real-world signals are complex, containing many frequencies at once. A musical chord, a crowded radio band, or overlapping cell phone channels all contain multiple simultaneous frequencies. Each pair can interact to produce intermodulation products, and those products can interact with other signals to produce even more. The result is a web of unwanted tones that can degrade signal quality in ways that harmonic distortion alone wouldn’t predict.
Intermodulation in Audio Systems
In audio equipment, intermodulation distortion adds tones that aren’t musically related to the original sound. Harmonic distortion at least produces notes at octave intervals, which can sound relatively natural at low levels. Intermodulation products, by contrast, land at arbitrary frequencies that clash with the music. A chord played through a system with significant intermodulation distortion picks up harsh, dissonant overtones.
The threshold for hearing intermodulation distortion depends on the listener and the type of signal. Research on transient intermodulation distortion (the brief spikes that occur when a signal changes rapidly) found that the most sensitive listeners could reliably detect levels as low as 0.5%. At low levels, listeners didn’t perceive it as obvious distortion but rather as a subtle change in tonal character, a slight harshness or roughness in the sound. This is one reason high-fidelity audio equipment is designed with extremely low nonlinearity: even small amounts of intermodulation create coloration that trained ears can pick up.
Passive Intermodulation in Cell Towers
One of the most practically significant forms of intermodulation doesn’t come from amplifiers or active electronics at all. Passive intermodulation, or PIM, is generated by components that aren’t supposed to have any nonlinearity: connectors, cables, antennas, and waveguides. Any junction between two different metals can behave as a crude nonlinear element, especially when it degrades over time.
PIM is a major concern for cell tower operators. Loose cable connections, dirty connectors, corroded antenna parts, and aging hardware all create nonlinear junctions that mix transmitted signals together. The intermodulation products can fall directly into the receive band of the base station, effectively jamming the tower’s ability to hear incoming calls and data. A system might work fine when first installed, then develop PIM problems months or years later as weather and vibration loosen connections or corrode metal surfaces.
Engineers classify PIM into three categories: design PIM (caused by inherent material choices in components), assembly PIM (caused by poor workmanship during installation), and “rusty bolt” PIM (caused by environmental degradation over time). The rusty bolt variety is the most unpredictable because it develops gradually and can be difficult to locate. A single corroded connector on a tower can generate enough intermodulation to degrade service for thousands of users.
Your Ear Creates Intermodulation Too
The inner ear is itself a nonlinear system, and it produces measurable intermodulation products. The cochlea, the spiral-shaped organ that converts sound waves into nerve signals, actively amplifies incoming sound using specialized cells called outer hair cells. This amplification process is inherently nonlinear, and as a byproduct, the cochlea generates its own tones at intermodulation frequencies.
These byproducts are called distortion product otoacoustic emissions, or DPOAEs. When two pure tones are played into the ear simultaneously, the cochlea produces a faint third tone at a predictable intermodulation frequency. This tone travels back out through the middle ear and can be picked up by a sensitive microphone placed in the ear canal. Audiologists use this phenomenon as a hearing test: if the DPOAE is present and at normal strength, the cochlear amplifier is working properly. If it’s absent, something is wrong with the outer hair cells, indicating hearing loss. It’s a particularly useful test for newborns and others who can’t respond to traditional hearing assessments.
Reducing Intermodulation Distortion
The most straightforward way to reduce intermodulation is to keep signal levels well below a system’s limits. Because third-order products grow three times faster than the input signal, even a small reduction in drive level produces a disproportionate drop in distortion. This approach, called power back-off, trades maximum output for cleaner performance.
Filtering works for some intermodulation products but not all. Systems operating over less than an octave of bandwidth can filter out harmonic distortion and sum-frequency products, since those land outside the operating range. The close-in third-order difference products, however, can’t be filtered without also removing the desired signals. For those, engineers turn to linearization techniques that cancel the nonlinearity itself. One approach uses carefully balanced signal paths with opposing nonlinearities, so the distortion from one path cancels the distortion from the other. In optical systems, researchers have demonstrated a dual-polarization technique that adjusts two components of light to cancel the cubic nonlinearity responsible for third-order products.
For passive intermodulation, the solutions are more hands-on: using high-quality connectors with consistent metal plating, torquing connections to manufacturer specifications, keeping RF paths clean and dry, and regularly inspecting outdoor installations for corrosion. PIM testing with specialized instruments has become a standard part of cell tower commissioning and maintenance.