The phenomenon is called Intermodulation, specifically third-order intermodulation when two signals produce a third frequency that falls into your band.

How it works

When two strong RF signals at frequencies f₁ and f₂ pass through a non-linear device (for example an overloaded amplifier, mixer, diode junction, corroded connector, or even rusty metal acting as a diode), they can mix and produce new frequencies.

Common third-order products are:

• 2f₁ − f₂
• 2f₂ − f₁

These are problematic because they often land close to the original signals, which makes them more likely to appear inside your observing band.

Example

Suppose two strong transmitters exist outside your band:

• f₁ = 900 MHz
• f₂ = 910 MHz

Third-order products:

• 2(900) − 910 = 890 MHz
• 2(910) − 900 = 920 MHz

If your receiver band includes 890 MHz, you’ll see interference even though neither transmitter is in-band.

Where the mixing usually occurs

In radio astronomy or SDR systems it commonly happens in:

• Front-end amplifiers driven into compression
• Mixer stages
• Poor connectors / corroded joints (passive intermodulation)
• Overloaded SDR ADCs
• Nearby metal structures acting as nonlinear junctions

The passive case is often called Passive Intermodulation (PIM).

In radio astronomy

At Hydrogen line (1420.405 MHz) observations, strong nearby transmitters can mix to create spurious lines that mimic astronomical signals, so front-end filtering and high-linearity LNAs are critical.

✅ Short answer: Intermodulation (specifically third-order intermodulation products).

A practical way to identify which transmitters produced an intermodulation spur is to work backwards from the observed interference frequency using the third-order formulas.

1. Start with the observed interference frequency

Let the interfering frequency be:

f_IM

Third-order intermodulation products come from:

f_IM = 2f1 – f2
f_IM = 2f2 – f1

This is part of Intermodulation.

2. Rearrange the formula

You can rearrange to find possible transmitters.

From

f_IM = 2f1 – f2

then

f2 = 2f1 – f_IM

So if you suspect a strong transmitter at f1, you can calculate the second frequency that would produce the spur.

3. Practical field method

Radio astronomers and RF engineers usually do this:

1. Measure the spur frequency precisely.
2. Look up strong transmitters within about 100 to 200 MHz of the band.
3. Plug combinations into the formulas:

2f1 – f2

2f2 – f1

4. See which pair produces the observed spur.

4. Example (similar to what happens near the hydrogen line)

Suppose you observe interference near the Hydrogen line at:

f_IM = 1420.405 MHz

You notice two strong transmitters nearby:

935 MHz (mobile base station)
450 MHz (UHF link)

Test:

2 x 935 – 450
= 1420 MHz

So those two transmitters could generate a spur almost exactly in the hydrogen line band.

5. Quick shortcut trick

Another quick diagnostic:

If you slightly attenuate one suspected transmitter (or it switches off), the spur will drop twice as fast in power because it depends on both signals.

Also, if one transmitter drifts by delta_f, the spur moves by:

2 x delta_f if it is the doubled term

• delta_f for the other term

That movement pattern helps identify which signal is f1 and which is f2.

Rule of thumb:

If you see unexplained narrowband RFI, check for pairs of strong transmitters where

2f1 – f2 approx f_observed

Here is a very practical spectrum-analyser / SDR trick used to recognise Intermodulation products almost instantly.

1. Look for symmetry around strong signals

Third-order products often appear symmetrically spaced around strong transmitters.

If you have two strong carriers:

f1 and f2

the intermodulation products appear at:

2f1 – f2
2f2 – f1

The spacing from the real signals is:

Delta_f = |f2 – f1|

So the spurs often appear exactly the same distance away as the spacing between the transmitters.

Example spectrum pattern:

890 MHz spur
900 MHz strong carrier
910 MHz strong carrier
920 MHz spur

Spacing between the real signals is 10 MHz, and the spurs appear 10 MHz outside them.

2. The “equal spacing” rule

If you see three or more narrow lines with identical spacing, suspect intermodulation.

Example pattern:

1420.0 MHz
1420.2 MHz
1420.4 MHz
1420.6 MHz

That evenly spaced comb pattern is often mixing products rather than real signals.

When observing the Hydrogen line, real hydrogen emission is usually broad and smooth, not a narrow comb.

3. Use attenuation as a test

Reduce receiver gain or insert attenuation.

Intermodulation products will usually:

• drop dramatically
• or disappear completely

Real signals generally reduce in a normal proportional way.

This happens because intermodulation requires nonlinear overload somewhere in the receiver chain.

4. Move the receiver tuning slightly

Shift the SDR tuning by a few MHz.

If the signal is real RF:

the signal moves exactly with the tuning.

If it is internal intermodulation, the spur may:

• stay fixed relative to a strong carrier
• or move in a nonlinear or unexpected way.

5. Waterfall appearance

On a waterfall display, intermodulation products usually:

• appear as perfectly straight vertical lines
• occur at mathematically related offsets
• mirror around strong signals.

Astronomical signals often:

• drift slowly due to Earth rotation
• broaden due to Doppler effects.

6. Gain test

Third-order intermodulation products increase approximately 3 dB for every 1 dB increase in signal level.

So if receiver gain increases slightly and the spur grows much faster than real signals, it is likely intermodulation.

Here is a very common real-world example in the UK that can create false signals near 1420 MHz due to Intermodulation.

A common UK intermodulation trap near 1420 MHz

The UK has strong mobile phone transmitters in these bands (among others):

• 800 MHz LTE
• 900 MHz GSM / LTE
• 1800 MHz LTE

These allocations are widely used by mobile networks according to Ofcom spectrum tables.

Two of these bands can combine to produce signals near the Hydrogen line.

Classic example

Suppose two strong transmitters exist:

f1 = 935 MHz (mobile base station)
f2 = 450 MHz (UHF link or telemetry)

Third-order intermodulation:

f_IM = 2f1 – f2

Calculate:

2 x 935 – 450
= 1420 MHz

Which lands almost exactly on the hydrogen line.

This type of mixing can occur in:

• overloaded LNA
• SDR front end
• corroded connectors
• metal structures acting as diodes
• nearby electronics

Another realistic UK pair

Mobile bands commonly include:

880–915 MHz uplink
925–960 MHz downlink

Example:

f1 = 950 MHz
f2 = 480 MHz

Then:

2 x 950 – 480
= 1420 MHz

Again very close to the hydrogen frequency.

Why radio astronomers often see this

Your receiver is tuned to 1420.405 MHz, but:

• two strong transmitters elsewhere
• mix in a nonlinear device
• generate a phantom spectral line

The SDR or LNA then detects it as if it were real.

Quick diagnostic

If the signal is intermodulation:

• it often moves when gain changes
• it disappears when attenuation is added
• it may appear only when strong transmitters are active

Real hydrogen emission does not behave like that.

Here is a quick professional sanity check commonly used to confirm whether a signal near the Hydrogen line is real or just Intermodulation.

It relies on the fact that astronomical hydrogen is affected by Doppler motion while interference usually is not.

The Doppler drift test

Because Earth rotates, the radial velocity of your telescope relative to the Milky Way changes continuously.

That causes the hydrogen frequency to shift slightly with time.

Typical drift:

about 1 to 5 kHz over an hour depending on where you point.

Interference does not drift.

So the test is simple.

Step-by-step method

1. Tune your receiver to around 1420.405 MHz.
2. Record a spectrum.
3. Wait 20 to 60 minutes.
4. Record another spectrum.
5. Compare the line position.

What you should see

Real hydrogen:

frequency shifts slightly
line shape remains broad and smooth

Example:

1420.401 MHz
later becomes
1420.398 MHz

Small but measurable drift.

Interference:

frequency stays exactly fixed
line often very narrow

Example:

1420.400000 MHz
still
1420.400000 MHz

No movement.

Another quick check: beam pointing

Point your antenna at different parts of the sky.

Real hydrogen emission:

• stronger near the galactic plane
• weaker away from it

Interference:

• stays about the same strength everywhere.

Another clue: spectral width

Hydrogen emission is broadened by gas velocity.

Typical width:

20 kHz to 200 kHz.

Interference:

often only a few Hz to kHz wide.

The “gain test”

Reduce receiver gain or insert attenuation.

Intermodulation signals often:

• collapse dramatically
• disappear entirely.

Real astronomical signals just scale down normally.

Professional radio observatories often use all four tests together, and they are very effective at spotting false signals.