Let's talk about noise ...

Foundations of Amateur Radio

Today I'd like to talk about noise, but before I do, I need to cover some ground. Recently I explored the idea that, on their own, neither antenna, nor coax, made a big difference in the potential for a contact when compared to the impact of path loss between two stations.

I went on to point out that you'd be unlikely to even notice the difference in normal communications. Only when you're working at the margins, when the signal is barely detectable, would adding a single dB here or there make any potential difference.

In saying that, I skipped over one detail, noise.

Noise is by definition an unwanted signal that arrives together with a wanted signal at the receiver. In HF communications, noise comes from many sources, the galaxy, our atmosphere, and man-made noise from things like electrical switches, motors, alternator circuits, inverters and computers.

The example I used was my 10 dBm beacon being reported by an Antarctic station. My signal report was about 5 dB above the minimum decode level and based on signal path calculations, -129 dBm, or around an S0 signal level. What that statement hides is that this is in the context of a noise level that's lower than -129 dBm. Remember, a negative dBm value means a fraction of a milliwatt.

While you're considering that, think of the reality of an Antarctic station. This particular station, "Neumayer III" has three 75 kW diesel generators, a 30 kW wind turbine generator, 20 caterpillar trucks, 10 snowmobiles and 2 snow blowers and computers and technology to support 60 people, in other words, plenty of local noise.

This makes it all the more remarkable that my 10 dBm beacon was heard and that there was an amateur there to set-up the receiver in the first place.

Before I continue, picture mountain tops peaking through the top of a cloud layer as viewed from the window of an aeroplane. If the cloud layer increases in height, less and less mountain tops are visible, until at some point, only clouds are visible. Alternatively, if the cloud layer descends, more and more of the peaks are visible, until at some point no cloud remains and you see the mountains in all their magnificent glory.

In that analogy, mountains represent signals and the cloud layer is the equivalent of the noise floor, and in a similar way, signals can be heard or not, depending on the relationship between the level of noise in comparison to the level of the signal. There's a name for this, it's called the signal to noise ratio or SNR, where a value of 0 dB means that noise and signal are at the same level, negative SNR values mean that the signal is weaker than the noise, positive SNR values means that the signal is stronger than the noise. If you know the power level in dBm for both the noise and the signal, you can subtract the two and end up with the signal to noise ratio.

In reality, all receiving stations have to contend with noise.

If I arbitrarily set the local noise floor at -100 dBm, somewhere halfway between S4 and S5, I'll mostly get laughed at by many stations, either because it's too high or too low. In case you're wondering, I've worked my station in both S0 noise and S9 noise environments and it's fun trying either and comparing. It's one of the reasons I often use a mobile station, to get away from urban noise around me, and you don't have to go far, a local park might be far enough from local noise to whet your appetite.

Besides, -100 dBm is a nice round number to play with.

You might recall that a typical path loss number for a 2,500 km contact on HF on the 10m band is about 129 dB. With a noise floor of -100 dBm, we immediately know how much output power is required to be heard above the noise. If the received signal has to be at least more than -100 dBm and we know that the path loss is 129 dB, then our transmitted signal needs to at least be enough to make up the difference.

Said differently, if our output power is too low, the signal at the receive station will fall below the noise and they won't be able to hear us.

So, if we start at say 30 dBm, have a path loss of 129 dB, we'll end up at -99 dBm, which is 1 dB above -100 dBm. Said in another way, the SNR for this is 1 dB.

I'd like you to notice something.

I've said nothing about the noise floor at the transmitter. We could have low noise, or horrendous noise, either way, it makes no difference to the receiver. What it hears is entirely dependent on the noise floor at the receiving station.

I wonder if that observation changes anything about what you think the impact might be of adding an 18 dBi Yagi to your station?

I'm Onno VK6FLAB