Is 5GHz max conducted power controlled by firmware?

[drinkingbird]

Apologies for bumping a ~1 year old thread.

I just wanted to share that I've been having difficulty trying to understand why a 2x RT-AX86U in AiMesh has been struggling to provide similar (or better) coverage than an eero system it replaced and I have finally found my answer. It seems that in 2023 Broadcom's drivers (or at least the ones Asus incorporate into their firmware) still do not take into account the updated Canadian regulations and continue to limit output power on channels 36-48. I've been running a virtually "clean" setup (no firmware tweaks outside of the Quick Internet Setup) which auto-selects channel 36 for 5GHz and in certain places where the eero had no problem serving clients with acceptable 5GHz performance, the Asus has them dropping to 2.4GHz or sticking to a (painfully) slow 5GHz radio. Switched from auto channel select to 149 and it has made a world of difference - an instant 15dBm boost plus significantly improved measured throughput when not next to a router (+350Mbps in some locations!).

Hopefully this is helpful for anyone like me who's been searching Google like crazy to stumble upon the right keywords to lead them to this thread.

It may not strictly be power, the Asus has higher gain antennas which focus the signal in smaller areas. Re-orienting your antennas may help (have some at 45 degree angles). Keep in mind you're only adjusting the transmit power, your client still has to "get back" to the router and if it has lower power or worse antennas (it likely does) increasing tx power can have limited effect if any.

Also they have to deduct the gain of the antennas from the power, higher gain antennas require them to restrict the transmit power to a lower amount, since it is not wattage that is regulated, it is db (or really microvolts per meter or something like that). At least that is how it is in the US. However a higher gain antenna at lower power can actually perform better as it will be better at picking up your client signal than a lower gain antenna, but you still have to take into account the wave pattern. Think of the antenna having a donut on it (like a ring toss game). Lower gain = fatter donut, higher gain = skinny donut. Envision the donut as you adjust the antennas to point it at the areas you need coverage.

With a 3 antenna one it is easy, one straight up, one 45 degrees to the left, one 45 to the right, typically that is the best for most situations. You do not want to exceed 45 degrees as the antennas all work in conjunction and changing the polarity (exceeding 45 degrees) can actually cause problems.

With 4 or more antenna just pick approximately 1/3 of them to angle.

Since mine is on a high floor I have them also angled forward a bit to focus some of the energy a bit downward. Nothing drastic, probably like 15 degrees.

Don't go too crazy, like I said the antennas are a system, they are not independent, so you're trying to increase the area they coverage without decoupling them from each other.

 

[got_milk]

Perhaps not in all Asus products. RT-AC86U works well on both 36-48 and 149-161 channels. RT-AX86U works well on 149-161 only.

Where did you find the information about this?

I found out about this "limitation" through this thread, which then led to (admittedly unscientific) testing to confirm whether or not that was still the case.

Fair point that it might not apply to all products - hence why I mentioned the RT-AX86U specifically, since that's what I have. I'm currently using Merlin's 388.2 beta on both of my nodes but confirmed it was the same on stock 3.0.0.4.388.22525 firmware (which makes sense, that's all handled by closed-source stuff anyway).

Like I said, I thought it was worth pointing out since the last update on this was a bit over a year ago and anything could have changed via firmware updates since then. I thought it'd be helpful for others coming across this to know that in 2023 the current situation remains the case.

 

[Wadadli]

Point of order: Channels 36-48 (80 Mhz bandwidth) are allowed up to 1W. Channels 52-64 (U-NII-2A) are DFS channels and are limited to 200mW. I would assume that if you use 160Mhz bandwidth, then your o/p would be limited to 200Mhz. I could be wrong though.

 

[ColinTaylor]

Point of order: Channels 36-48 (80 Mhz bandwidth) are allowed up to 1W. Channels 52-64 (U-NII-2A) are DFS channels and are limited to 200mW. I would assume that if you use 160Mhz bandwidth, then your o/p would be limited to 200Mhz. I could be wrong though.

Not in Canada which is what this thread was discussing.

https://www.ic.gc.ca/eic/site/smt-gst.nsf/vwapj/rss-247-i2-e.pdf/$file/rss-247-i2-e.pdf

 

[DJones]

Merlin give you lots of control over your router being able to ssh gives you a great deal of bells and whistles to play with. But the onus to comply with regulations is at your discretion. Each router is tested to comply with regulatory standards per country so well I personally see no issue in tweaking the router if it’s possible to comply with new regulations if the manufacturer hasn’t; the legality maybe grey. That’s all I’ll say since discussion on how isn’t allowed on the forums since Merlin considers it “hacking” even though custom router firmware could and is considered just that in some countries.

Difference is he has permission from Asus and likely has conditions he must follow considering Broadcom drivers are proprietary and likely under NDA per partners along with the tools, instructions, commands. anyways maybe a future update will include increased power levels maybe it won’t no one here can change Broadcom drivers in Asus Merlin firmware. I would suggest leaving feedback to Asus as they would be the only ones who could legally change it and if that happens Merlin would get it in a future GPL.

 

[Wadadli]

Not in Canada which is what this thread was discussing.

https://www.ic.gc.ca/eic/site/smt-gst.nsf/vwapj/rss-247-i2-e.pdf/$file/rss-247-i2-e.pdf

The .pfd you are quoting is 6 years old. I was just going by the following:

 

[ColinTaylor]

The .pfd you are quoting is 6 years old. I was just going by the following:

The information in it is still correct and is the same as in the current Wi-Fi database and your picture. Channels 36-48 are restricted to 200mW unless you're using it outdoors with an appropriate licence. The DFS channels can use up to 1W with TPC (250mW without TPC). (N.B. To keep things simple I'm talking about EIRP rather than conducted power as the latter is dependent on antenna gain.)

wireless-regdb — Linux Wireless documentation

# Sources:
# https://www.ic.gc.ca/eic/site/smt-gst.nsf/vwapj/rss-247-i2-e.pdf/$file/rss-247-i2-e.pdf
# https://www.ic.gc.ca/eic/site/smt-gst.nsf/eng/sf11750.html (6 GHz operation)
country CA: DFS-FCC
    (2402 - 2472 @ 40), (30)
    (5150 - 5250 @ 80), (23), NO-OUTDOOR, AUTO-BW
    (5250 - 5350 @ 80), (24), DFS, AUTO-BW
    (5470 - 5600 @ 80), (24), DFS
    (5650 - 5730 @ 80), (24), DFS
    (5735 - 5835 @ 80), (30)
    (5925 - 7125 @ 320), (12), NO-OUTDOOR

 

[Wadadli]

The information in it is still correct and is the same as in the current Wi-Fi database and your picture. Channels 36-48 are restricted to 200mW unless you're using it outdoors with an appropriate licence. The DFS channels can use up to 1W with TPC (250mW without TPC). (N.B. To keep things simple I'm talking about EIRP rather than conducted power as the latter is dependent on antenna gain.)

wireless-regdb — Linux Wireless documentation

# Sources:
# https://www.ic.gc.ca/eic/site/smt-gst.nsf/vwapj/rss-247-i2-e.pdf/$file/rss-247-i2-e.pdf
# https://www.ic.gc.ca/eic/site/smt-gst.nsf/eng/sf11750.html (6 GHz operation)
country CA: DFS-FCC
    (2402 - 2472 @ 40), (30)
    (5150 - 5250 @ 80), (23), NO-OUTDOOR, AUTO-BW
    (5250 - 5350 @ 80), (24), DFS, AUTO-BW
    (5470 - 5600 @ 80), (24), DFS
    (5650 - 5730 @ 80), (24), DFS
    (5735 - 5835 @ 80), (30)
    (5925 - 7125 @ 320), (12), NO-OUTDOOR

Thanks Colin
In my chart, the author should not have used the phrase "FCC - Except Outdoor License Req. >200mW" It confused the crap out of me. I just took it as being the same as the FCC spec.

 

[sfx2000]

Tx levels on equipment are set and locked once the device has reached certification (FCC in the US, IC in Canada, Offcom in the UK, etc).

So even if the regulatory agency makes changes later on, older equipment would have to be recertified for the new levels, which typically doesn't happen.

 

[sfx2000]

"Can the 5GHz max conducted output be increased by the firmware from 200mW to 1W?

Getting back to the orignal question - Levels are usually either in the WiFI chip firmware and/or in the bootloader...

IIRC - broadcom is in the embedded firmware inside the WIFI chip itself.