Decreasing 2.4G wifi level after changing from 374.34_2 to the latest Merlin fws

[Dimon]

Hi,

I have RT-AC68U. The Firmware 374.34_2 was the best for me because in my flat I have a stable wifi 2.4G signal level -65/-70dB behind 3 walls from the router (meausured by InSSIDer android program in my Samsung Galaxy Note). Speedtest.net resulted the speeds 20-30Mbit/s which is rather comfortable for web surfing. But there was a nat loopback problem on my AC68U on this FW. The next FW 374.35_2 has solved the NAT loopback problem but the Wifi 2.4G Wifi level decreased by 10-15dB! InSSIDer shows the levels between -80dB to -75dB! Sometimes the connection is losing at all. And the connection speed significantly decreased too (2-3 Mb/sec). The last FW 374.35_4 hasn't change this situation. After changing back to 374.34_2 the signal becomes strong again. I had many experiments during November. The problem is exists for me.
To my mind the reason of this signal decreasing is that 374.34_2 is based on 3.0.0.4.374_979 sources from Asus and the next 374.35_X are based on the older 3.0.0.4.374_339 sources from Asus. RMerlin, please can you comment this idea? Am I right? Are you going to make your new FW based on 374_979 or more fresh Asus sources which will use the maximum potential of RT-AC68U Wifi?

Thank you for the great FW!

 

[RMerlin]

Hi,

I have RT-AC68U. The Firmware 374.34_2 was the best for me because in my flat I have a stable wifi 2.4G signal level -65/-70dB behind 3 walls from the router (meausured by InSSIDer android program in my Samsung Galaxy Note). Speedtest.net resulted the speeds 20-30Mbit/s which is rather comfortable for web surfing. But there was a nat loopback problem on my AC68U on this FW. The next FW 374.35_2 has solved the NAT loopback problem but the Wifi 2.4G Wifi level decreased by 10-15dB! InSSIDer shows the levels between -80dB to -75dB! Sometimes the connection is losing at all. And the connection speed significantly decreased too (2-3 Mb/sec). The last FW 374.35_4 hasn't change this situation. After changing back to 374.34_2 the signal becomes strong again. I had many experiments during November. The problem is exists for me.
To my mind the reason of this signal decreasing is that 374.34_2 is based on 3.0.0.4.374_979 sources from Asus and the next 374.35_X are based on the older 3.0.0.4.374_339 sources from Asus. RMerlin, please can you comment this idea? Am I right? Are you going to make your new FW based on 374_979 or more fresh Asus sources which will use the maximum potential of RT-AC68U Wifi?

Thank you for the great FW!

374.979 does not contain any wireless driver for the RT-AC68U. 374.34 used the RT-AC68U driver from a previous RT-AC56/AC68U release but I don't remember which one (might be _134 or _205). 374.35_4 is based on the 374_339 wireless driver. Different wireless drivers can indeed impact performance in various ways.

One thing to try is to disable beamforming. In some scenarios it will actually decrease performance rather than improve it.

Asus is expecting a new RT-AC68U release this week, which was why I didn't pressure them into releasing the source code for 374_371. In addition to being too busy at work to do any GPL merge for the next couple of weeks (I just spent my Sunday moving our offices to a new location, wiring the network and reconnecting everything, and tomorrow I'll be installing a new server at a customer's office rather than unpacking stuff at our own office. I need a vacation...)

 

[Dimon]

RMerlin, thank you for your immediate assistance!

I played for some time with different combinations of beamforming switches on 374.35_4 for both 2.4G and 5G bands but I haven't received any positive result unfortunately. 2.4G wifi level was not higher than -80dB in my most distant room! I rolled back to 374.34_2 and returned to the level -65dB but with nonfunctioning NAT loopback
So, will be waiting for your next releases

Thank you

 

[RMerlin]

2.4 GHz performance is one thing that the new Broadcom SDK Asus will be using in the next release is supposed to address.

 

[DrTeeth]

I played for some time with different combinations of beamforming switches on 374.35_4...<snip>

PMFJI, but where are the beamforming settings located? I am using the SDK 5 of the latest firmware and cannot find it.

PS I have searched the forum before asking.

 

[RMerlin]

PMFJI, but where are the beamforming settings located? I am using the SDK 5 of the latest firmware and cannot find it.

PS I have searched the forum before asking.

There is no beamforming feature on the old SDK5.

 

[Dimon]

Success with wifi performance on the last FW!

Today I tried to measure wifi signals both on 2.4G and on 5G with the last Merlin FW 374.38_1 on my RT-AC68U. I found the level of 2.4G approximately equal with that on 374.34_2 (-70 -75dB in my far room) but with significally higher maxium speed measured by Speedtest.net. But more surprising for me was increased by 10-15dB level of 5G band (-75dB)! As a result I had a stable connection during all time on 30-40Mbit/s on Android device in my far room (2 thick walls from router) which was impossible on the earlier firmwares.

Great work! Thank you, RMerlin!

 

[RMerlin]

Today I tried to measure wifi signals both on 2.4G and on 5G with the last Merlin FW 374.38_1 on my RT-AC68U. I found the level of 2.4G approximately equal with that on 374.34_2 (-70 -75dB in my far room) but with significally higher maxium speed measured by Speedtest.net. But more surprising for me was increased by 10-15dB level of 5G band (-75dB)! As a result I had a stable connection during all time on 30-40Mbit/s on Android device in my far room (2 thick walls from router) which was impossible on the earlier firmwares.

Great work! Thank you, RMerlin!

All the credit goes to Asus/Broadcom, I simply used their newer drivers.