R1D2 wrote

Don't buy the whole equipment w/o having tested beforehand! Get an EAP110-Outdoor from Amazon for testing, you can send it back if it does not fit your needs.

As for iperf rates in EAP110-Outdoor's standard setup (no tweaks): ~20 to 24 Mbit/s pure data speed (~ 54 Mbit/s WiFi speed) for 802.11b/g connections and between 70 to 85 Mbit/s pure data speed (~ 150 Mbit/s WiFi speed) for 802.11n connections, both at 20 MHz channel width. Distance to the EAP close to the clients used for measuring. Of course bandwidth will be reduced if more clients connect. In order to get best airtime you could use fixed 802.11n mode, but if there is a single client not supporting 802.11n, then you will have to use mixed mode.

Not sure a) what WiFi modes an iPhone supports and b) how sensitive iPhone "antennas" (the piece of sheet metall surrounding the case) are.

However, if you deploy four or five EAP110-Outdoor, there might be too much interferences in an area of 80x250m to get the best bandwidth possible. I would not deploy more than three 2.4 GHz outdoor devices in such a small area.

Ok- good to know.
In any case, I have ordered 1 EAP110 from amazon for a field test as you have suggested

As for the iphone, it certainly supports good transfer rates as I have seen 300+mbps on iperf on the 5ghz radio and 100+ on 2.4ghz when in proximity of an indoor Netgear R7000..
The transfer rate seems to be an issue with Ubiquiti equipment not playing well with non Ubi devices - 40mhz seems to be a no go with the M2 (unless it is directly talking to another Ubnt)

I also tested with a macbook (2016) which has a very capable antenna and card (i have even seen 70-80 MB/s file copy rates from a Gigabit wired NAS on a Netgear R7000 AC mode) as well as a One Plus Two phone -
None of them will connect to the M2 unless I drop the channel width to 20 mhz

Problem is that the data rate gets capped at 65mbps at 20mhz width and real world rates drop to as low as 4-5 mbps even when very close to the antenna.
I verified the LAN speed at the AP end and it does 90+ mbps on iperf so the cabling is not an issue

There are other folks as well on the Ubnt forums complaining about 40mhz channel width not working with client devices with Ubnt radios in AP mode..

In all honesty, not at all impressed by Ubiquiti esp considering their equipment costs almost 2x vs comparable TP-Link models in India..
I sure hope the EAP110 works out well

R1D2 wrote

I have not much experience with UBNT gear, only did test an AirGrid M2 HP without much success, but using 40 MHz channel width in 2.4 GHz is a problem for every device, since it needs 8 channels simultaneously. With 20 MHz it still needs 4 channels in 802.11n and 5 channels in 802.11b/g mode. Given the fact, that the 2.4 GHz band is most often overcrowded, 40 MHz is often not useable w/o much interferences.

Also, in 802.11b/g bandwidth gets capped at 54 Mbit/s WiFi rate yielding ~ 24 Mbit/s data speed. So make sure your device use 802.11n for best results. All MacBooks released in last ten years support N mode and prefer this mode over B/G when negotiating the protocol at connection setup.

To be honest, I had never even bothered with the channel width of any of my indoor network equipment but always got pretty decent throughputs even in 2.4Ghz
Now that I am looking at the internal router (Netgear 7000) and AP (another r7000), , they are both have a checkbox enabled that says 20/40 mhz coexistence

If I am close to the r7000, i get very high realworld throughput (500mbps+) on 5ghz while pretty decent throughput on 2.4 although I cannot say what channel width is being used

Still not sure what is wrong on the m2 - although one odd thing seems to be the mode which is set to b/g/n and cannot be modified
Even a decade old wrt54g that I have as a wired AP in the porch seems to be performing far better

Waiting for the EAP110 and hoping it works better!