How many dB of gain are needed to double the range?
6 dB in free space. Because path loss increases with distance at 20 dB per decade, doubling the distance brings exactly 6 dB of extra loss; conversely, if you add 6 dB to the link budget the theoretical range doubles. 6 dB corresponds to increasing transmitter power fourfold; a loss of 10 dB shortens the range by roughly a factor of 3.16.
This rule is the single most useful number in RF. Where you find the 6 dB does not matter; its effect on range is the same:
- Increasing transmitter power fourfold (for example from 10 mW to 40 mW) -> 6 dB. But the permitted ERP limit often does not allow this.
- Increasing antenna gain by 3 dBi at both the transmitter and the receiver -> 6 dB in total. It is also generally easier from a regulatory standpoint, because what is limited is ERP.
- Improving receiver sensitivity by 6 dB (for example using FSK instead of OOK) -> the same result, and without increasing battery consumption.
- Removing 5 m of RG58 cable -> roughly 2.2 dB recovered at 433 MHz; this gives back the roughly 23% of range lost because of the cable.
The reverse is just as merciless: if the noise floor rises by 10 dB, the effective range shortens by roughly a factor of 3.16 even though nothing changed on the transmitter side. This is why range suddenly halves on a crane site when an inverter starts up.
The rule is exact only for free space. In obstructed environments loss grows faster with distance, so 6 dB generally gains less than a factor of two; do not leave the calculation on the optimistic side.
Full topic: 433 MHz or 2.4 GHz?