Is it possible the wrong standard is being used for 2.4GHz power?

Omnia hardware Omnia HW problems
1

Related to a different post of the safety of raising Tx power,
I decided to take a risk and raise it.
but the results suggested a different issue.

  1. I changed the channel on wifi1/5GHz, and the power went up as it should.
    (different 5GHz channels have different limits)
    So the system works as it should.

  2. But there is no difference for different channels in 2.4GHz.
    So I raised the power on wifi2/2.4GHz from the driver default of
    Tx-Power: 16 dBm / 39mW to 17 then to 18dBm / 63mW
    However, Current power remained unchanged at 16 dBm
    Reboots etc didn’t change current power.

I then check the standards for the Czech Republic to New Zealand
(where I am at the moment).

CZ Czech Republic
2.4GHz ch 1-13 100mW / 20 dBm

NZ New Zealand
2.4GHz ch 1-11 1000mW / 30dBm

It looks like the standard for the beautiful country of Czech Republic
is being used instead of the standard for the New Zealand.

Is it possible the wrong standards are being used to limit 2.4GHz power output?
(it may be the difference between 16 & 20 is to allow for antenna gain).

2

I did the same check after your topic about txpower and double checked if there is something hardcoded in reforis about region and in luci.

Check from console iw reg get my global is set to Poland as it should but fine grained settings for specific interface are set to US which is weird.

3

“iw reg get” shows the currently set regulatory domain
and gets the information about that domain from the kernel’s wireless regulatory database.

Mine is:

country NZ: DFS-ETSI
        (2402 - 2482 @ 40), (N/A, 30), (N/A)
        (5170 - 5250 @ 80), (N/A, 17), (N/A), AUTO-BW
        (5250 - 5330 @ 80), (N/A, 24), (0 ms), DFS, AUTO-BW
        (5490 - 5730 @ 160), (N/A, 24), (0 ms), DFS
        (5735 - 5835 @ 80), (N/A, 30), (N/A)

The first group (eg 2402 - 2482 @ 40 below ) is the frequency range and maximum bandwidth
for which this particular rule applies.
The second group (N/A, 30) is the maximum antenna gain and the maximum EIRP
(Effective Isotropic Radiated Power), in dBm

See Understanding the output of the ‘iw reg get’ command

You can compare it to the maximum allowed transmission power (EIRP) per country here:
Maximum wifi transmission power per country

So the regulatory information in the kernel is correct:
The country is NZ,
and max power is 30dBm (1000mW) in the 2.4GHz channel.

The problem is:
it seems that the correct county code (NZ) is not actually being used to set my power limits.
The information actually being used is that for Czech Republic (and China, HK,…)
20dBm / 100mW.

4

Some cards have the limit hardcoded in firmware. There might even be different releases of the card for different regions.

We have one mikrotik routerboard with a card that allows going to 1 W, but we had to root the routerboard, flash openwrt and use a different firmware (I think). The manufacturers are required to make it difficult or even impossible to breach the regulatory limits. Some take a shortcut and hardcode the 100 mW limit to be on the safe side. It is possible that the card refuses being switched to NZ region because thaat would require it to go over the hardcoded limit.

But all of these are just speculations.

5

Interesting experience.

However Europe has been on this ETSI standard since I think 2020, China is also on it,
the US, Canada and NZ are the same as the ETSI limits,
and Australia is slightly greater (for 2.4GHz).

So if someone is enforcing a limit of 100mW / 20dBm,
or choosing a card (for radio1) with this limit,
when most of the major markets are 1000mW / 30dBm
their shareholders, and their children, would probably poke them in the chest.
Even their mothers might have something to say.

Also worth noting, if I set radio0 to 2.4GHz,
the power goes up to 29dBm / 794mW, and the driver default is that maximum.
However as that needs to be set to 5GHz, that is not useful.

But whether it is a Openwrt mistake or a Turris mistake,
or a really old chipset meant for a remote restrictive market
I have no idea. But -50dBm at 8m in clear air,
without even an intervening mosquito, is a bit poor.
(my 7 year old Linksys WRT1900acs at 8m gives me a signal 0f -25dB
even with mosquitoes).

If no one here has a good explanation,
I should open a support case with this question

6

ok, I’ll open a support question, and post the results - hopefully this thread will stay open.

7

(I think these threads close if there is no message for 3 days

  • so adding this post to keep it open)
    I have submitted request to support as follows:

Issue:
On my radio1( wifi2) card, for the 2.4GHz band,
driver default for Tx-Power is: 16 dBm / 39mW
If i change manually to 17 or 18dBm / 63mW
the current power remained unchanged at 16dBm.
The max power available in the slider is 19dBm / 79mW.

Under the ETSI standard,
the max power is 30dBm (1000mW) in the 2.4GHz channel.

Europe has been on ETSI since (I think) 2020,
China is on the ETSI standard,
and the US, Canada and NZ are the same as the ETSI limits,
and Australia is slightly greater (for 2.4GHz).

So my question is:
Why is the Radio1 driver being limited to
16dBm rather than the ETSI standard of 30dBm?

Is the wrong standard being used, or is it just an old radio?

(Why do I ask:
It is important because the signal level of Radio1 is miserable.
Testing shows average signal of -50dBm at 8m in clear air,
without even an intervening mosquito.
For comparison, my 7 year old Linksys WRT1900acs
(running Gargoyle) at 8m gives me a signal of -25dB.
I have 4 of these Linksys WRT1900acs and I am looking to the
Turris Omnia 6 to replace them, but at that level of power it is impractical.
Other than that it is a beautiful router !)