Hi Everyone,
I want to share my recent experience with successfully installing OpenWrt 23.05.2 on my TP-Link Archer A7 hardware version 5.8.
I was running OEM firmware 1.2.1 Build 20220714 rel.75573(5553). It's the latest on the TP-Link site as of this post, but still over a year old.
I was connected to LAN1 and used the GUI in the router to install the BIN file. In the GUI it was,
Advanced > System Tools > Firmware Upgrade > Manual Upgrade > New Firmware File > Browse (bin file for v23.05.2) > Upgrade
The file I used was openwrt-23.05.2-ath79-generic-tplink_archer-a7-v5-squashfs-factory.bin.
I did not need to rename the file. Installation completed without issue and I was able to log in using the default IP address.
Hopefully someone can update this successful installation (US region) in the wiki here:
https://openwrt.org/toh/tp-link/archer_a7_v5
My biggest motivation for OpenWrt was to get WPA3 on the A7, plus general security fixes that come with updated firmware. It's been well over a year and TP-Link has not released new firmware for the A7 so I'm guessing they'll never add WPA3...but I noticed they added WPA3 for the A9 back in 2020.
I'm generally happy with OpenWrt, and big thanks to all the devs and contributors to OpenWrt!
The router is one floor away from where I normally connect. My internet provider limits me at 300 Mbit/s. I did some performance comparisons. I ran many iterations of the speed test on google as an easy way to check.
My latency hasn't changed between OEM and OpenWrt, as it's still in the 4-6ms range.
On the 2.4GHz radio I was getting 60-80 Mbit/s on OEM, whereas OpenWrt gives 70-80 without flow offloading and 80-90 with flow offloading. Overall about the same but maybe a bit faster.
On the 5 GHz radio I was getting 220-290 Mbit/s with the OEM firmware, whereas with OpenWrt without flow offloading it's 100-150. With flow offloading it's 230-270. It's a bit slower on 5 GHz but I'm not complaining. It's still plenty for what I do. I'm getting better features overall, especially security-wise.