This article explores the design / analysis of a passive receive system for the 160m band, determining the maximum receiver noise figure (NF) to achieve a specified maximum S/N degradation (SND) by receiver internal noise.
For explanation of the metric Signal to Noise Degradation (SND), see Signal to noise degradation (SND) concept.
The antenna data table given uses a tabulated average gain for a set of interesting 160m antennas published by Tom Rauch (W8JI) at https://www.w8ji.com/receiving.htm . Read the whole article, it is interesting and relevant.
External noise is estimated using ITU-R P.372-16, and results are tabulated for the five noise environment categories used in ITU-R P.372-16.
The example and calculations assume linear systems, if there is significant nonlinearity that gives rise to significant IMD, IMD noise depends on the specific scenario (including receive spectrum) and is not captured by this analysis.
Maximium receiver NF for given SND
Let us assume that noise arrives equally from all directions. In that case, the average gain of the antenna is used to determine the noise power captured, \({Gain}_{avg}={Gain}_{max}-\text{Directivity}\).
The table above uses Rauch’s tabulated average gain for a range of 160m antennas, and calculates the maximum receiver NF to achieve the specified 1dB maximum SND.
The Rural environment (green) is probably a sensible target for most users, perhaps a little ambitious for very noisy suburban / city locations, make your own choice.
If your target is Quiet Rural, ambient noise is the greater of Quiet Rural and Galactic, except when foF2 is lower than 1.8MHz (which is unusual), a consequence of ionospheric shielding above foF2. At 1.8MHz (a consequence of ionospheric shielding above foF2) Quiet Rural and Galactic are almost the same anyway.
IMD
A word about IMD, IMD causes undesired mix products from input signals, some of which may appear in-band and are effectively noise, whether or not they appear to be white Gaussian noise or have discrete spectral peaks. The point is that IMD is an imperfection of all system, and adds noise that is not accounted for above. If the undesired signals are strong enough, the IMD bad enough, then IMD noise may be significant, ie degrade S/N more than predicted assuming a linear system.
Broadband antennas assist the capture of undesired signals, and so may exacerbate IMD noise.
Low average gain antennas help to reduce the amplitude of undesired signals reaching the first active device (the main cause of IMD), but still there may be significant IMD observed.
If you insert a broadband attenuator in front of the first active device (receiver or preamplifier), it attenuates both undesired signals that create an undesired mix product, and for example, if it is a third order mix, then a 10dB attenuator will reduce third order products by 30dB.
Front end attenuation is often used to reduce a receiver’s potential for IMD noise.
Another option is some form of front end selectivity to reduce the amplitude of out-of-band signals that may mix to undesired in-band products.
Analysis
It can be seen that antennas with high average gain (high efficiency) can tolerate a receiver with poorer (higher) NF.
Whilst it is relatively easy to make receivers with quite low noise figures, they may not have sufficient immunity to IMD to be practical at 160m when using a broadband antenna system.
So, whilst the intended receiver might have a best NF of 6dB, it might be prudent to consider that it may require an additional front end attenuation of say 20dB to achieve the needed immunity to IMD so the IMD noise is relatively insignificant.
Strongly directional noise
For noise captured equally from all directions, the noise captured is related to average gain, \({Gain}_{avg}={Gain}_{max}-\text{Directivity}\). Essentially this is the scenario dealt with in ITU-R P.372.
If ALL noise came from the back of the antenna, the relevant gain would be \({Gain}_{noise}={Gain}_{max}-\text{F/B}\). In a practical situation, it is unlikely that ALL noise comes from this direction, but it may be that a weighted average gain for noise calculation is a little less than average gain.
Realise that if ALL noise came from one direction, then the ITU-R P.372 conditions do not apply and its predictions are not relevant.
You could measure your own external noise more directly to inform an analysis.
Can I find SND from my station configuration?
Sure, let’s take an example where Fam=65.4dB (for Residential precinct from the table in the spreadsheet), Pennant antenna assume Gavg=-43.6dB, 2dB of feed line and Mismatch Loss, and receiver NF (with att ON) is 20dB.
Above is a calculation of expected SND of 3.1dB.