Default Raspberry Pi 4 WiFi transmit power is set at 31 dBm. Is this a fixed figure or the maximum limit?

[Indigo12]

System: Raspberry Pi 4b 2GB in Flirc case running fully updated Raspberry Pi OS for light desktop use.

sudo iwconfig wlan0

Returns a default WiFi transmit power of:

Tx-Power=31 dBm

31 dBm seems excessively high (1000mW) versus typical mobile devices, which max out at 12dBm (15mW) according to the source below. As someone unfamiliar with RF engineering, I'm wondering why they would set the figure so high by default.

https://metis.fi/en/2017/10/txpower/

To limit RF exposure, I reduced transmit power to 12 dBM (the minimum needed to get a reliable connection) using the following command and added it to rc.local to ensure it is run on startup.

iwconfig wlan0 txpower 12

This is based on the presumption that:

1. The txpower figure is the maximum limit for WiFi transmit power.
2. The RPi dynamically adjusts the actual WiFi transmit power to what is required for a reliable connection, but not exceeding txpower.

Can someone confirm that I have understood these points correctly? If not, what should be done to set the WiFi power limits as above?

Under iw, txpower can be adjusted according to {auto|fixed|limit} parameters. What do these parameters do?

[pidd]

The typical maximum power out of the chip is less than 23dBm, that is before losses in filters, diplexor/switcher etc and I doubt the antenna has any significant gain.

I presume the 31dBm setting is just a nominal indicator of maximum output ...... as a binary max?

The Pi must be less than 23dBm erp to have achieved the compliance certifications.

It makes sense to have maximum output as default to reduce problems for initial setups. There are enough wifi problem reports occurring without intentionally having it set to a low power and having range problems.

RF exposure wise I doubt many have their working Pi's in their pockets, the RF exposure from mobile phones will be many orders greater than a non-body-worn Pi.

[Indigo12]

The typical maximum power out of the chip is less than 23dBm, that is before losses in filters, diplexor/switcher etc and I doubt the antenna has any significant gain.

I presume the 31dBm setting is just a nominal indicator of maximum output ...... as a binary max?

The Pi must be less than 23dBm erp to have achieved the compliance certifications.

It makes sense to have maximum output as default to reduce problems for initial setups. There are enough wifi problem reports occurring without intentionally having it set to a low power and having range problems.

RF exposure wise I doubt many have their working Pi's in their pockets, the RF exposure from mobile phones will be many orders greater than a non-body-worn Pi.

I came across this thread, which (though not specific to the RPi) seems to confirm what you say about WiFi transmitters in general.

https://www.snbforums.com/threads/power ... 0-4.13568/

The 31 dBm figure seems to be a 'user target' or initialization parameter that is then subject to dynamic adjustments based upon driver and country code settings. The final power output of the antenna would thus be in compliance with wireless emissions regulations.

The FCC wireless standards report for the RPi 4B confirms this, showing WiFi transmit power within the range 9.5 -28.2 mW (9.8 - 14.5 dBm).
For interest, the transmit power of the RPi 3B is much higher, at 138 mW (21.4 dBm).

RPi4B: https://fccid.io/2ABCB-RPI4B
RPi3B: https://fccid.io/2ABCB-RPI32

I'm wondering then, is there much practical benefit to adjusting txpower?

I've noted from experimentation that there is a minimum level (in my case 12 dBm) below which WiFi transmissions fail. Signal is not lost, given that the antenna can still pick up the router emissions, but web traffic stalls, consistent with loss of transmit power.

This would suggest that there is some benefit to reducing txpower, if the objective is to minimize WiFi footprint for whatever reason. However, are there likely to be any adverse/counterproductive effects on driver/antenna function if the default value is altered?

[pidd]

I've never found a need to reduce power because of interference despite having devices on different wifi networks very close together but its rare I have flat-out streams over wifi as I have a lot of ethernet and powerline network so its only portable devices that use wifi.

Some people (apparently) reduce power to limit the distance they can be snooped, if I considered wifi to be that insecure I simply wouldn't use it for anything private.