Between 3 walls about 30 feet away I’m getting nearly 200mbit/s. Keep in mind I’m also on MT7986 running snapshots. I’m not sure how much has changed or differs between the 2 chips.
What if you change AX to AC and keep all other settings the same?
I had the intention to do so, but never did…
Please open one
Done:
5 Likes
Maybe slightly off topic but do you all see issues with Apple devices losing WiFi connection now and then? My wife has complained about this ever since getting RT3200's with OpenWrt. She has to manually reconnect often. Perhaps something related to power saving on Apple devices?
What software is that, I know it's iperf but I am talking about the gui part of it.
The iOS app iPerf
Seen with iPhone 11:
I do not see it with iPhone 13.
1 Like
Could be this bug in iOS 16.
I have a ‘TP-Link EAP615-Wall v1’ where the ‘ax’ option just doesn’t work when it’s enabled. Only ‘ac’ will work.
@Sokolum I have the same EAP615-Wall v1. For me ax seems to work fine across Android, iOS, Windows & MacOS.
I am running on:
OpenWrt 22.03.2 r19803-9a599fee93
Trying with my Iphone 12.
Anyway AC still enough for me.
How is your experience so far with your EAP? I just have it for a couple of hours.
Yeah, same here.
1 Like
For reference, I have two mini PCs in the house with Intel AX200 2x2 wireless cards and both antenna's installed. The first floor thin PC (manjaro linux) is a few walls and turns and rooms from a first floor Linksys EA8500 AP. The second floor PC (Windows 10 - WiFi drivers updated to most recent) is in a bonus room over the garage right next to the second floor Askey RT4230 Rev 6 AP.
On the first floor PC I got ~400 Mbps 80 MHz wide 802.11ac WiFi to the first floor EA8500 AP, and 250-300 Mbps 80 MHz wide 802.11ac WiFi to the second floor Askey AP. On the second floor PC I got a very steady ~400 Mbps 80 MHz wide 802.11ac on the second floor Askey RT4230W AP from the second floor PC in the same room.
With a recent RT3200 purchase set up on the second floor in the same location, I got a consistent ~500 Mbps on the second floor PC in the same room as the RT3200 on 80 MHz wide 802.11ax and 802.11ac. On the first floor PC, going up to the RT3200 on the second floor, I got 175-200 Mbps 802.11ac and 225-250 Mbps 802.11ax.
But, I also noticed the RT3200 802.11ax, while slightly faster than its 802.11ac at long range (from second to first floor of home), was definitely not as steady. Over the iperf3 trials on 802.11ax, most were in the 300 Mbps range, but enough trials dropped down to ~150+/-50 randomly to hurt the average, whereas average 802.11ac performance was slower, but steady. I think if I went a little further away, the RT3200 802.11ax would have gone to crap in a hurry, but I didn't do any more testing.
So that is what I see on 2x2 clients with a mt76 driver based RT3200. My general conclusion is the RT3200 isn't bad, but I don't think it performs quite as well at range as my ipq806x AP's do under 802.11ac, and the RT3200 802.11ax performance looked choppy enough at long range that I think I wouldn't bother with ax if I were putting it in service on a permanent basis and pushing its range (it's being set up for someone else, but I'll have it for awhile first).
Oh, and yes, I put the lines in the wireless file for he beam forming and bss on the RT3200. All tests with 22.03.2 stable on all devices. The ipq806x devices have on demand up/down thresholds down to 20/10 so the CPU speed increases at low threshod, and the RT3200 used the schedutil governor. I also tested iperf3 -R and didn't note any obvious differences in reverse direction. Except for that, either a PC or NanoPi R4S were used for iperf3 server duty, so not stressing the AP CPU's with double duty in other words.
In hindsight, I should have set the RT3200 up in place of my first floor AP too, and compared to the first floor PC. Another day - I'll update this post if I do with those results, including retest with the current Links EA8500 before swapping the RT3200 on the first floor.
1 Like
In general, it would be great to have precise measurements. E.g. a Markdown formated table that features following information:
- info about hardware and software (e.g. version of OpenWRT) of all involved devices (both accesspoint and client)
- Signal strength received (Example). E.g. with Wifianalyzer
- Info from speedtests using Iperf3 or Iperf2 (How to).
- Comparison of same location and same devices with different protocols (e.g. AX vs AC vs N) and different drivers/firmware/versions of OpenWRT or original software. (Example)
- Comparison of different distances (Example)
- Comparison of different layers of obstacles (e.g. walls) (Example)
- Device configuration, especially in case you have good results with your wifi and think it works very well.
Just saying it works fine or it does not work well does not help at all. I am glad we now have some people with different devices in this thread and not just Belkin RT3200/Linksys E8450 users...so please 
Did anyone try to build patched wifi stack (mac80211 + mt76) from https://git01.mediatek.com/plugins/gitiles/openwrt/feeds/mtk-openwrt-feeds/
Probably it is much more stable.
Steps to build are not clear to me... is this a replacement for OW's package/kernel/mac80211 and target/linux/mediatek/mt7622?
Wireless stack and mt76 driver it looks like, with some other ramips/mediatek target patches sprinkled on top. For 21.02, mind you.
You download OpenWRT, than download this mediatek repo and run scripts to replace mac80211,mt76 with custom version; after that you build OpenWRT, no need to change config.