meeotch
New Around Here
Just got my RT-AC68U, and I was a bit surprised to find that it lagged behind the DLink DIR-868L (AC1750) that it's replacing. (I should mention that I have the T-Mobile version, with the bootloader & firmware replaced according to the guide over at slickdeals.) I did some tests, and despite having similar signal strength, I couldn't get similar throughput using the Merlin firmware. I had better luck with Tomato - but see the notes below.
These results are my best attempts, after dicking around with channels and a few other parameters. Most settings ending up default in the end, though. As you can see, Merlin was noticeably slower, both LAN & WAN.
Tomato did better, but it also seemed to take forever to make the wireless connection, and then an additional 30-60sec after connecting to the router for connections to my server and to websites to "kick in". (They would hang for a period after first making the wireless connection.)
I could continue to play around with alternate firmware versions (I've heard the Merlin fork is faster)... But I was hoping there might be a "best practices" setup for performance on one or the other firmwares. There a metric assload of options in there, and I imagine that most of them should just be left at defaults, so random noodling doesn't seem like the best strategy.
Code:
rssi/qual are signal strength & avg. quality
speed is throughput to the internet, measured at speedtest.net
thru is throughput via LAN to my file server (which has wired connection to the router)
DLink DIR-868L
rssi -28 -57
qual 97% 97%
speed 11/5Mbps 14/5Mbps
thru 3.5 8.3MB/s
RT-AC68U, Merlin 3.0.0.4_376.47_0
rssi -28 -53
qual 99% 99%
speed 9/3Mbps 11/5Mbps
thru 4.7MB/s 8.3MB/s
RT-AC68U, Tomato ARM--124-AIO-64K
rssi -39 -56
qual 99% 99%
speed 11/5Mbps 15/5Mbps
thru 5.3MB/s 9.4MB/s
These results are my best attempts, after dicking around with channels and a few other parameters. Most settings ending up default in the end, though. As you can see, Merlin was noticeably slower, both LAN & WAN.
Tomato did better, but it also seemed to take forever to make the wireless connection, and then an additional 30-60sec after connecting to the router for connections to my server and to websites to "kick in". (They would hang for a period after first making the wireless connection.)
I could continue to play around with alternate firmware versions (I've heard the Merlin fork is faster)... But I was hoping there might be a "best practices" setup for performance on one or the other firmwares. There a metric assload of options in there, and I imagine that most of them should just be left at defaults, so random noodling doesn't seem like the best strategy.