What is a dB?

lgllp

We use certain types of antennas because they have “gain” – they increase signal strength. They don’t magically create extra energy though, they direct radio signals into a smaller beam in the same way a spot light does. The higher the gain the narrower the beam and the more concentrated power – in certain directions only though!
We often talk about this gain in terms of ratios, because the gain is independent of the power level – an antenna that doubles the signal strength of a one watt transmitter will also double the signal strength of a ten watt transmitter.
You can also have losses – a poor cable might lose one half (or more) of the signal power. This is also independent of the absolute power level – one watt becomes half a watt, ten watts becomes five, etc.
Ratios are messy to use though when using several of them together – imagine “I have a seventeen times antenna with a one third power loss in my cable and one tenth of my power lost in the connectors”. It’s hard to figure out just how much overall gain or loss that represents.
To make it easier, we work with deci-Bells (dB), which are defined as ten times the logarithm of the ratio. The beauty of this is that to multiply gains (or losses) we just add the dB numbers. (Gain will be a positive dB number, while loss will be negative, so subtracted) So now “I have a 15dB antenna, with 3dB of cable loss and 2dB of connector loss – a total of 10dB gain”. Much easier!
Here are some examples of dB ratios:
+3dB is doubling the power, -3dB is only a half.
+6dB is Quadrupling power, -6dB is only a quarter.
+10dB is 10 times the power, -10dB is only a tenth.
+20dB is 100 times the power, -20dB is only one percent.
(You can now see why the “ten times” factor is used – it avoids fractions in the log values for commonly encountered ratios)
All of the above deals with RATIOS, which are independent of power. So it’s meaningless to say “How much power will a 23dB antenna give me” – the antenna gets as much power as the transmitter feeds it, but focuses it to give the effect of a 23db stronger (200 times more powerful) transmitter – in the focus direction only. In fact, the effective width of the strengthened signal is specified as “degrees of beamwidth” and is the number of degrees between the points at which the gain falls off by 3dB.
We often talk about the “Effective Radiated Power” and that is the equivalent power level produced by the transmitter/antenna combination (in the highest gain direction). So a 1 watt transmitter with a 20dB (100 times) antenna is producing a ERP of 100 watts.
High gain antennas have narrow beamwidth, and so can be harder to align. It should be noted that gain can be achieved in either the horizontal or vertical plane (or both). Most antennas have quite different patterns of radiation in the vertical and horizontal planes, with different antennas chosen for the different pattern according to their function. Antennae also have a “polarisation” – either horizontal or vertical, which defines the way in which the radio waves are transmitted, but the main connection between signal strength and polarisation is that having different polarisations between transmitter and receiver gives another 20dB or more of loss!
Finally, another term encountered is dBm, and this one IS related to power. Zero dBm is defined as one milliwatt (one thousandth of a watt) and so transmitter powers are often expressed in terms of this reference. This is NOT the same as antenna gain – this is real power. Receiver sensitivity is also expressed in terms of how many dBm or what power level is required to receive a signal. So if I have a 20dBm transmitter, a 15dB gain antenna at each end, 110dB of path loss (signal attenuation in the atmosphere) and a –90dBm receiver, can I get a signal through?
Add them up: 20 plus 15 minus 110 plus 15 gives -60dBm, so I will theoretically have a signal of 30dB more than the minimum I need, so it should be a good path. Of course, this is only theoretical, trees and other obstructions can reduce the signal dramatically, but this is a good starting point before a real-world trial.