The Ghost on 146.500
A 2 m FM operator at a busy suburban site hears a weak but perfectly readable signal on 146.500 MHz. It appears only when two nearby amateurs are transmitting simultaneously on 145.900 and 146.200 MHz; switch either of them off and it goes away. Neither of those stations can be heard directly on 146.500. The receiver's first IF is 10.7 MHz, so its image frequency is 125.1 MHz, and there is a broadcast FM transmitter about 1 km away. Inserting a 10 dB attenuator at the antenna socket makes the ghost signal vanish completely, while the two strong local signals remain usable.
1. Identify the mechanism, show the arithmetic, and rule out the alternatives.
Read the worked answer
Arithmetic
Third-order products from two signals f1 and f2 appear at 2f1 - f2 and 2f2 - f1.
2 x 146.200 - 145.900 = 146.500 MHz
(As a confirmation, the other product should fall at 2 x 145.900 - 146.200 = 145.600 MHz - worth checking for.)
Diagnosis
- The signal exists only when both transmitters are on. A real transmission, a harmonic or an image would not depend on two unrelated carriers being keyed together.
- Image check: 146.500 - 2 x 10.7 = 125.1 MHz. Nothing is being received there, so it is not an image.
- It is not overload of a real signal, because nothing is transmitting on 146.500.
Conclusion
This is third-order intermodulation generated by mixing of two strong signals in a nonlinear stage - most likely the receiver's own front end or first mixer.
2. Why does the 10 dB attenuator prove the product is generated inside the receiver?
Read the worked answer
The scaling law
A third-order product falls 3 dB for every 1 dB reduction applied to each of the two fundamentals. A 10 dB pad therefore drops the product by 30 dB, while a genuine on-frequency signal drops by only 10 dB.
What was observed
The ghost disappeared into the noise, but the two strong local signals - now 10 dB weaker - were still perfectly usable. That is exactly the 30 dB versus 10 dB signature, so the mixing is happening after the point where the attenuator was inserted, i.e. inside the receiver.
The counter-case worth remembering
If the interference had fallen only 10 dB and stayed readable, an attenuator would not have explained it, and you would be looking for a real transmitter on 146.500 or for an external mixing source - a corroded mast joint, a dissimilar-metal contact, a rusty gutter - near a strong transmitter.
3. Recommend fixes and note what this means for interference complaints from non-amateurs.
Read the worked answer
Fixes at the affected receiver
- Fit a bandpass or cavity/helical filter ahead of the receiver so the two strong signals are attenuated before the nonlinear stage.
- Leave a fixed 10 dB pad in line and make up the loss with audio gain if the receiver has the sensitivity to spare.
- Reduce RF gain rather than audio gain where the receiver allows it.
- Move the antenna away from the strong transmitters, or use a directional antenna pointed away from them.
- Choose a receiver with better dynamic range if this is a permanent site problem.
Before blaming anyone
Check your own transmitter into a dummy load with a spectrum display or a receiver on a separate antenna - confirm you are not radiating a real product on 146.500 that others could hear.
The EMC lesson
The same front-end nonlinearity exists in everyone else's equipment. A strong nearby amateur signal rectified in a neighbour's TV tuner or audio amplifier produces this same class of complaint, and the cure is the same shape: filtering and better screening at the affected device, reduced power, or greater separation - plus making sure your own feeder and antenna are not radiating harmonics in the first place.