Belkin RT3200/Linksys E8450 WiFi AX discussion

@darekxan try newest master. 2.4 GHz regression was maybe fixed by

See here:

1 Like

SNAPSHOT r22191-f6a7ce2501 fixes it completely :rocket:

~ iperf3 -c ethernet_connected_host.local -P 4
[SUM]   0.00-10.00  sec   103 MBytes  86.2 Mbits/sec                  sender
[SUM]   0.00-10.02  sec   100 MBytes  84.2 Mbits/sec                  receiver
~ iperf3 -c ethernet_connected_host.local -P 4 -R
[SUM]   0.00-10.01  sec   103 MBytes  85.9 Mbits/sec                  sender
[SUM]   0.00-10.00  sec   101 MBytes  85.1 Mbits/sec                  receiver

Price seems to be changing daily ($49 yesterday) - but the RT3200 is available in U.S at Walmart for $50 today.


Let's buy them all up to help solidify the fact that we greatly appreciate and prefer the mt76 chipsets over some of the other more closed varieties. :slight_smile:



I bought the linksys e8450, it will be here tomorrow, I wondered if the device is now pretty stable with 2203 and master? I will compile my custom build, I hope is pretty good, I had R7800 and wrt3200.

1 Like

I've been using the Linksys E8450 for a while now. It's pretty stable for me. I'm running my own master build; PPPoE, with WED and H/W Offload enabled. It has OpenVPN TAP and Wireguard server and/or clients tunnels running on it.

AX range is pretty OK for me as well.


Does anyone using WED have this king of issue ?

Quick test can be done with
echo 1 > /sys/kernel/debug/ieee80211/wl1/mt76/sys_recovery
This emulates internal driver crush.

SER (System Error Recovery) process is activated to restart driver.
Unfortunately it can't handle the situation with mt7915 chip.

Yes, and I also opened this for what appears to be the same issue:

1333.680189] blk_update_request: I/O error, dev mtdblock2, sector 256 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 0
[ 1333.691246] blk_update_request: I/O error, dev mtdblock2, sector 256 op 0x0:(READ) flags 0x0 phys_seg 1 prio class 0
[ 1333.701788] Buffer I/O error on dev mtdblock2, logical block 32, async page read


I've got four of the RT3200 "under my care", and they've all been solid.

First one is at sister's (Colorado) for over a year now, running old snapshot with Syncthing emulating an Apple Time Machine. Also runs SQM (cake) to fix stuttering issues she'd been having with the Apple router (bufferbloat showed C-D grade depending on time of day, now shows A+ all the time). Wifi authentication was an issue to an ancient wireless printer she has (no WPA? something like that), so we just ran a cable to it. All other devices are wireless, and I keep asking about issues with connectivity, there are none.

Second is at a son's place (Illinois) for 6-7 months, running 22.03.3, also SQM and adblock. I think he has two wired devices (ws and laptop), but the rest (laptops and phones) are wireless. Again, I ask about troubles, but no issues.

Third is at another son's in-laws (Tapei) for about 3 months. Plain 22.03.3 running there, no reported issues and in fact a huge improvement over their previous configuration. Only a couple devices there, but all wireless. He does lots of video calls to them, and they were terrible before (again, a bufferbloat issue), works fine now.

Fourth is my local "experimental" one on a subnet here in my lab (California), running 22.03.3 for 8 months. I keep my primary workstation connected through it via Wifi (Intel AX200), so as to monitor stability. Its installed packages change all the time, as I play with it (sometimes experimenting before deploying to my APU2 main router), but I've had no issues at all.


As I understand it the RT3200s were not designed to work with 160MHz hence performance is worse than with 80MHz. Does that accord with your testing?

Well, i'm unsure about Apple M1's 160Mhz support, but I do seem to similar performance on 80 and 160Mhz. According to my tests current SNAPSHOT performs similarly for both configurations. MacOS and Inssider shows the network as 80MHz still, occasionally switching for a brief moment to 160Mhz. I don't have any other decent 5Ghz (besides iPhone 12 that performs adequately well) hardware to compare. There are at least one report of Intel AX200 performing better on 160Mhz.


Bought linksys e8450 and I want to install open wrt but it seems impossible to access the web linksys portal without wifi or only via ethernet cable to pc, it asks to have access to internet. Impossible to do something.

Whoops, you landed in the Developers forum. Have a look here and here to get going with it. You might also find some useful guidance here:


I have also noticed this regression in more recent versions of the stock firmware. I worked around it by connecting the WAN port of the Belkin/Linksys device to another router to supply it with a DHCP lease (but not with actual Internet connectivity). I don't think there is an easier way around it :frowning:


Multiple people have reported good results with 160 MHz in 5 gigabit band with ax on mt7915. See here:

mt7915 is 4x4 with 80 MHz, 2x2 with 160 MHz and 2x2 with 80 MHz dbdc. 160 MHz dbdc is not supported.


that's correct, but the device (and hence analog parts in the RF path) are meant to be used with 80 MHz channel bandwidth in 4T4R mode. What you got here is btw. MT7915+MT7975PN supposedly. DBDC would be different frontend types, so this is not supported on the RT3200 which uses the built-in 2.4GHz radio of the MT7622 SoC (which is more like MT7615)


Does anyone know why RT3200 5GHz radio power is limited to 23 dBm on U-NII-1 channels (36-48), even though on U-NII-3 channels (149-165) radio power up to 27 dBm is supported with country code set to US? Up to 30 dBm is allowed in the US on U-NII-1 and U-NII-3 channels.

Radio power is limited to 24 dBm on U-NII-2 channels (52-144), as expected with country code set to US.

1 Like

I think it just uses the wireless-regdb which says 23 except for the comment which I guess is not processed.

# 5.15 ~ 5.25 GHz: 30 dBm for master mode, 23 dBm for clients
(5150 - 5250 @ 80), (23), AUTO-BW


Thank you - I would never have found that. The RT3200 is the first OpenWrt target I've used that with country code set to US could not set transmit power up to 30 dBm on 2.4 GHz channels 1-11 (RT3200 limited to 28 dBm), U-NII-1 channels 36-48 (RT3200 limited to 23 dBm) and U-NII-3 channels 149-165 (RT3200 limited to 27 dBm).

I've experimented with more OpenWrt targets than I care to admit over the years that could set tx power up to 30 dBm: TP-Link Archer C7, Linksys EA6350v3 and EA8500, Askey RT4230W, Dynalink DL-WRX36. I half suspected the RT3200 amplifiers were just not capable of a full watt. Looks like it is the defaults used for the OpenWrt builds instead.