hAP AC^2 Wifi performance (almost solved)

Hey guys, so yesterday I got my first Mikrotik, after lots of reading and some tweaking I managed to get my 5GHz working at my max speed (around 300 mbps down and 300mbps up, but ONLY for my smartphone!), but no matter what I do my 2.4GHz is stuck at 50mbps down and 50mbps up.
Here is what I tried to do:
I updated the firmware to the latest one - 6.46.3
I disabled b band (set to 2GHz-G/N)
I tried setting channel both to 20 and 20/40MHz XX no difference in either
I tried different frequencies, some were way worse, a few capped at the aforementioned 50mbps up/down
I tried setting Antena gain to 2
I tried setting TX Power to 17
None of these settings managed to help me get over that limit.

Here is how I am testing my speed:

  1. speedtest.net, the interesting thing that I am noticing here is that, on the download test it slowly builds up and gets to 50mbps and maintains it, on the upload test it jumps to 100+ mbps and then goes down to 50mbps, kind of like an ESP in a car, you push the throttle but it doesnt let you, it seems like its trying to get higher speed, but something is pushing the brakes. This test was done with a smartphone (Xiaomi Mi 8 Lite), the same phone reaches 270-280 mbps down and up on 5GHz. A test with a laptop (equipped with Intel AC 3165, and attached to the power adapter, no power saving activated) yields better results, 60-70mbps download and 110 mbps upload. The same laptop on 5GHz reaches 200mbps download (and then slowly drops down to 150mbps) and maintains ~270mbps upload
  2. I tried iPerf test between a laptop and a wired computer (within the local network), the test was consistently capped at 100mbps for 2.4GHz, and 220mbps for 5GHz
  3. I tried transferring a file between these two same computers, locally, speeds varied anywhere between 3 and 10 MB/s (varying all the time) for 2.4GHz, and 10-25 MB/s for 5GHz

All devices are in the same room and within 1 to 3 meters of the router, I tried placing them right next to the router to eliminate obstacles and distance as a factor, no change.
I live in a house, there are no other WiFi networks around me, the only WiFi devices in my home are 2 laptops and 2 smartphones and I do not own a microwave ( :stuck_out_tongue: ), at the time of testing, only the device that was being tested was consuming network resources. I have also ordered a new Intel 9260 to eliminate my old WiFi card on the laptop as a factor, will have it on Tuesday, but I’m not so sure that it is 100% causing this problem, as seen by tests from other devices, and because my current wifi is supposedly capable of handling speeds of about 400-ish mbps.

Update: Today I tried with a completely different laptop that is equipped with Intel 9560, the tests are better, but still far from what I am expecting (with my current knowledge).
2.4 GHz 150-200 mbps speedtest.net, and 170mbps local transfer with iPerf
5 GHz 220-250 mbps speedtest.net, and 270 mbps local transfer with iPerf
Now, even if we disregard the speedtest.net results, the local transfers still seem pretty low?

Okay, its been a few days and no responses yet which makes me think that either I did not ask my question correctly/did not provide enough information, or no one has suggestions for the issue? I could not find a thread with directions on how to ask questions in this forum. Could someone let me know if there’s an issue with my post, or simply no one has other suggestions than what I’ve already tried?

What I have read is that your performance is good. MikroTik writes their own drivers and it seems to be close to the max. Other forums mentions this speed also. Your question is complete so why nobody answers…

I think your speed is normal. My ISP speed is 840/220, with hAP ac2 I got 470mbps speedtest and 270mbps iperf3 in 5GHz, actual transfer movie file from my NAS to phone is about 48MB/s, all tested with my iPhone 11 Pro Max

Just some ideas./maybe well known or not relevant for you/ that might help looking for the reason of the 50 Mbps transfer rate. 2.4 GHz is crowded and dangerous for interference. It is not because you have no other wifi in house that there is no other wifi signal in your house. Don’t look to the devices, use the “SCAN”, “freq usage”, “Snooper” tools on the wireless interface to find out. (Or the “interface wireless spectral scan” in the Terminal window) And yes only channel 1 6 or 11 should be used. Usage of other channels will give destructive side-channel interference. Having your devices close to the router does not help for co-channel interference. Wifi is very polite. If a device is transmitting everything else WAITS. Even if that transmission is seen below -86dBm in strength, and your connection is at -27 dBm, your wifi waits. Every device has a different cut-off for recognising another is transmitting, the noise floor for my MKT now is -111 dBm.

And then what can you expect. Look at the “registration” in the wireless tab page. TX and RX rate is there for your connected device: things like “72.2 Mbps-20MHz/1S/SGI”. Its your data rate, the number of spatial streams, and the guard interval. To get higher than 72.2 Mbps you need at least 2 spatial streams. Here it is just 1. (Spatial streams depend on the number of 2.4GHz radio’s in the AP and in the device, and on the signal quality) Using 40 MHz bandwidth on 2.4 MHz band is not practical. You take away 2/3 of the available channels, and chances for interference increase.

Above -30 dBm signal strenght distortion at the receiver starts to kick in. Reduce the transmit power. as needed.

Data transfer rate is around 50% of the interface rate. It halves again in case of bidirectional communication. (Wifi is half duplex, and the 802.11 protocol has a lot of overhead also)

iPhone 11 Pro MAX focuses on Wifi 6 (802.11ax) , and also tries to use 802.11ac on 2.4 GHz according to on-line information. 802.11ac on 2.4GHz is non-standard and is not on the Mikrotik AP’s. So don’t put the iPhone 2.4 GHz on ‘Auto’ or ‘802.11n/ac’ maybe. Didn’t find the number of 2.4 GHz radio’s or antennas.for the iPhone.

Thanks for the replies guys.

Now, some things I still cant figure out, just because most of the information is very vague about it are:
When I am connected to my 2.4GHz network, on the status of my WiFi it says that the speed is 300 mbps, does that mean that if I start downloading I can reach max of 300 mbps or if I start uploading a max of 300 mbps (assuming that I am only downloading or only uploading), or does it mean that half of it is my max download and the other half is max upload. Like I can reach a maximum of 150 mbps download and the other 150 mbps are not accessible for download and reserved for upload (in case its needed while downloading)? In essence, if I reach 150mbps while transferring a file over the 2.4GHz I am at the maximum and everything is actually normal?
Same goes for the 5 GHz network, if it says that my speed is 867 mbps, … same question?

P.S. I am upgrading my radio on the laptop today to Intel AC 9260 2x2 will see how that one goes. Also I did use the Scanner and Frequency Usage tools, there is absolutely nothing in my home that the Mikrotik detects that is using any of the frequencies of the WiFi, just 2 veeery distant wifi networks.

Screenshots of Scanner and Freq. Usage tools:
Scanner 2.4ghz - https://prnt.sc/r0rn0l just 2 very distant 2.4ghz wifi networks
Freq usage 2.4ghz - https://prnt.sc/r0rnm8
Scanner 5ghz - https://prnt.sc/r0roi4
Freq usage 5ghz - https://prnt.sc/r0ro5m

WiFi air interface is half-duplex (time-division duplex) meaning air-time has to be divided between both directions. There’s no hard division hence almost 100% of air time can be allocated for one direction in certain period of time and to the other direction in next period of time.

The wifi speed indicated shows maximum theoretical bitrate. However due to several technical reasons (all are part of standard) only approximately half is realistically achievable. When signal strength (in both directions!) is high and external interference low.

Well well, nice and quiet in your location. Not seen that often.

Your distant neighborg probably has set his AP on “auto”. And vendors just take some channel then, and don’t stick to the gentlemans agreement of only using non-overlapping channels 1,6 and 11. (Europe could use 1,5,9,13 if 802.11b is disabled, but that’s not commonly used.)

Signal is 16 dB above the noise floor, that’s enough to make devices wait for it to end any transmission. The freq usage is low, that is an advantage.
It’s not co-channel interference, if you don’t use channel 9, but it will interfere (disturb) wifi signal on channels 7-8-9-10-11. So CCQ will drop (see WLAN1 status tab), and it can cause more retransmits .

For full and detailed analysis of wifi performance I use two (quite long) documents:
https://www.duckware.com/tech/wifi-in-the-us.html
https://www.mist.com/documentation/wi-fi-fundamentals/
They are not the “wifi bible” but explain in detail and length what to expect from wifi performance.
In short : you never get the interface rate as data rate (the AP beacon and probe overhead, and the specific wifi timing for collision avoidance are always there)
In bidirectional communication you get only half of this data rate.
If you use TCP, then TCP congestion avoidance kicks in as well (the car ESP equivalent) , slowing down the transmission when there is packet loss. (https://en.wikipedia.org/wiki/TCP_congestion_control)

Okay, I just upgraded to 9260, there is night and day difference.
Local transfer (of a shared file) on the 2.4GHz network is at 20-22 MB/s which is 160-180 mbps
Local transfer (of the same file) on the 5GHz network is at 50-55 MB/s which is about 400-440 mbps

After all thats been said, if I understand correctly, this result is actually normal/expected/and even possibly on the high end of expectations?

One additional question, I am going to use two old TP-link routers to extend my wifi to more distant locations in my house (I am not expecting any performance from them), the only question that I have is, since they will be using 2.4GHz only, I should set them all on different channels, for example Mikrotik on ch1, TPlink1 on ch6 and TPlink2 on ch11, correct?

Yes.

One additional question, I am going to use two old TP-link routers to extend my wifi to more distant locations in my house (I am not expecting any performance from them), the only question that I have is, since they will be using 2.4GHz only, I should set them all on different channels, for example Mikrotik on ch1, TPlink1 on ch6 and TPlink2 on ch11, correct?

Yes, if you connect them with ethernet and put all three AP with only 20 Mhz Channel Width on 2.4 Ghz (with 40 Mhz they’ll take 2 channel like 1+6 or 6+11).