E4C

Receiver noise, selectivity and overload

Calculate noise floor from bandwidth, and explain image rejection, roofing filters, attenuators, overload and phase noise.

14 pool questions1 on the exam

Syllabus: Receiver performance: phase noise, noise floor, image rejection, minimum detectable signal (MDS), increasing signal-to-noise ratio and dynamic range, noise figure, reciprocal mixing; selectivity; SDR non- linearity; use of attenuators at low frequencies

The noise floor grows with bandwidth and with noise figure

Every 10× more bandwidth raises the floor 10 dB. Wide filters hear more noise.

Receiver bandwidth

Noise figure0 dB

Noise floor−144 dBm

−174 dBm is the thermal noise in a 1 Hz bandwidth at the input of a perfect receiver at room temperature.
Noise figure = how many dB the noise generated by the receiver is above that theoretical minimum.
MDS = minimum discernible signal: the weakest signal that stands out of the floor.

Bandwidth raises the floor10 log(1000 ÷ 50) = 10 log 20 = 13 dB

RememberFloor = −174 dBm + 10 log(bandwidth) + noise figure.

Match bandwidth to the signal, then move the passband off the neighbor

Match bandwidth to the signal. Use IF shift to slide the passband away from a nearby station.

Receiver bandwidth

IF shiftcentred

Resultinterference rejected

Control	What it does
Choice of bandwidths	match the modulation bandwidth: best signal-to-noise, least interference
IF shift	slides the passband to cut adjacent-frequency stations
Narrow roofing filter	attenuates strong signals near the receive frequency: better blocking dynamic range

RememberNarrow enough to drop the neighbor, wide enough to keep the signal.

A high IF pushes the image far from the signal, where a front-end filter can reject it

The image lies 2 × IF from the wanted signal. A higher IF moves it far enough away for the front-end filter to reject.

The mixer turns two frequencies into the IF: the wanted one and its image, 2 × IF away. Only a filter ahead of the mixer can remove it. A front-end filter or preselector also rejects strong out-of-band signals. A narrow IF filter acts after the mixer, and a notch filter removes only one frequency.

RememberHigher IF = image farther away = easier for the front end to reject.

Strong signals overload: attenuate them, filter them or lose to them

Example levels in dB. The attenuator cuts signal and atmospheric noise together; the receiver's own noise is the only thing left behind.

Input attenuation15 dB

S/N cost1.2 dB

On low HF bands atmospheric noise is generally greater than the receiver's own noise, even after attenuation. An attenuator cuts the signal and the atmospheric noise equally, and the receiver's own noise is still smaller, so S/N barely moves.

Strong-signal problem	Where
SDR overload	input exceeds the A/D converter's reference voltage
Capture effect	FM: a stronger signal on the same frequency suppresses the weaker one

RememberOn low bands, attenuate to stop overload: the noise was outside anyway.

Phase noise on the oscillator rides a strong neighbor onto your signal

Reciprocal mixing: oscillator phase noise mixes with a strong adjacent signal and makes noise right where you are listening.

Reciprocal mixing: local oscillator phase noise mixes with adjacent strong signals and creates interference to the desired signal. An SDR's master clock with excessive phase noise does the same: it combines with strong nearby signals to generate interference.

RememberStrong neighbor × noisy oscillator = noise on your frequency.

Loading lesson…

Go deeper

Compendium articles on the ideas in this lesson, beyond what the exam asks.

Check what stuck

Practise the 14 pool questions for this group. Answers are shuffled, and every one comes with a one-line "why".

Practise E4C →