@Bartvz , the 2018-1-25 commit was reverted; @dissent1 mentioned the commit after that. For him it was a big improvement, I can't report the same. Waiting for @nbd to official push it to trunk for further testing on my side.
Same here but it also seems my build environment is borked. Ah well, time to start using openwrt sources 
Are these new mt76 patches in the latest nightlies yet?
No, a faulty patch was reverted due to some bugs. There's a new patch in @nbd 's staging tree, but it hasn't landed in master just yet.
Where is? I can't find it, thanks.
EDIT: I found it, as unsaid.
compiled with latest patch and speed much better on xiaomi mi r3g (same chipset as dir-860l)
Latest mt76 drivers have landed in the master branch now
Yup, some time for a new build!
I'll probably upload a new one tomorrow evening.
2 Likes
Cant wait to test it
New build is up (r6009)!
Testing wireless performance with iperf3 running for 60 seconds yields the following results:
2.4 GHz band connecting at 144 Mbps:
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-60.00 sec 312 MBytes 43.7 Mbits/sec sender
[ 4] 0.00-60.00 sec 312 MBytes 43.7 Mbits/sec receiver
5 GHz band connecting at 866.7 Mbps:
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-60.00 sec 291 MBytes 40.7 Mbits/sec sender
[ 4] 0.00-60.00 sec 291 MBytes 40.7 Mbits/sec receiver
Compared to build r5442 WiFi performance seems a bit down. On that build I got the following numbers, 73.5 Mbits/sec for the 2.4GHz band and 57.4 Mbits/sec for the 5 GHz band. Keep in mind this is just one test, the router is situated in a noisy WiFi environment and at the moment of the test there were multiple clients associated with the router.
Looking forward for your results with this build!
Thanks for your build and test but sorry I don't understand...
2.4 GHz at 144Mbps is more fast than 5 GHz at 866.7 Mbps?
Yes, the 2.4GHz band has more throughput then the 5GHz band. In theory the 5GHz band should be faster due to more bandwidth available but if that is not being used 
1 Like
On my end:
2.4gig@20MHz SISO: 55Mbps / 50Mbps
5GHz@80MHz SISO: 200Mbps / 180Mbps
In my opinion the 2.4gig is a bit more stable, and the ramp up is quicker, more stable.
In general, not faster, not slower than the previous build, a bit more stable on 2.4gig.
There is a new version of the mt76 driver which should fix some issues with the 2.4 GHz band so I'll compile a new version tonight. However, I am unable to test this till Monday evening. Are there some brave souls who want to beta test?
1 Like
Hi Bartvz, I am on r5762 yet, I thought to test 6009 this weekend but if you have a beta version I would be glad to test it
Things went slightly differently, so I'll upload a new build in +/- 5 hours
And it's finally up, @Axl_Mas and other brave souls click here!
2 Likes
Some early benchmarks of the last beta build (r6022):
iperf3 running for 60 seconds, 1st laptop set as server and connected to the router with 100/10 ethernet, 2nd laptop connected through wifi at 2 meters from router, no other devices connected.
SQM deactivated.
r5762:
5G 47.0 Mbits/sec
2.4G 32.9 Mbits/sec
r6022:
5G 57.6 Mbits/sec
2.4G 51.4 Mbits/sec
Please note that the test is between r5762 and r6022, I've skipped the r6009 release.
With r6022 the auto channel on 5G failed, I tried to reboot the router several times but I solved only inserting the channel manually (the good thing is that with previous version the high channels on 5G were not working at all, now the connection seems stable with channel 157!)
I am going to stress the connection with some days of full bandwidth torrenting.
On my device on r6022, I'm not seeing a way to do channel 157. Mine tops out at 140.
@Bartvz FYI, the master branch contains new commits that fixes the VLAN functionality of the switch in the DIR-860L
3 Likes
New build tomorrow
1 Like