ASUS still not making use of new power limits on U-NII-1 channel range

[maxbraketorque]

As I was looking through FCC data on the AC3100 and AC86U this morning, I noticed that ASUS is still appears to be limiting transmit power in the U-NII-1 channel range despite the new FCC rules that went into effect in 2014 that allow APs running on U-NII-1 to have the same transmit power as U-NII-3.

When I read through the AC86U reports submitted to the FCC, it looks like for Channels 36-48 in 802.11AC 80 MHz bandwidth in MIMO mode, the four antennas have a combined transmit power of 23.6 dBm while in beamforming mode, total power is 21.2 dBm. For Channels 149-161, transmit power values are 29.8 and 29.3 dBm respectively for the two modes. RSSI measurements of the two channel ranges confirms this. Setting the control channel to 36 (with 80 MHz bandwidth) gives an RSSI of -55 dBm, and for the control channel set to 149, the RSSI jumps to -43 dBm. Interestingly, within either the U-NII-1 or U-NII-3 channel range, altering the control channel has a noticeable effect on RSSI.

Here are my observations for the RT-AC86U in 802.11AC 80 MHz mode with beamforming enabled:

U-NII-1 channel range

• Control channel = 36 gives -55 dBm
• Control channel = 40 gives -54 dBm
• Control channel = 44 gives -50 dBm
• Control channel = 48 gives -50 dBm -- Seems to be the preferred control channel for U-NII-1 range.

U-NII-3 channel range

• Control channel = 149 gives -43 dBm -- Seems to be the preferred control channel for U-NII-3 range.
• Control channel = 153 gives -44 dBm
• Control channel = 157 gives - 45 dBm
• Control channel = 161 gives -48 dBm

 

[maxbraketorque]

I decided to do some additional testing to double-check these observations. The observations in my first post were made using my Galaxy Note 8 that only reported the RSSI value. I performed a second set of obsevations using my 2016 MacBook Pro where I can measure both RSSI, background, and "quality" which is RSSI minus background. The results are somewhat different.

Control Chan RSSI (dBm) Bkgrnd (dBm) Qual (dBm)

U-NII-1 channel range

• 36 -41 -93 52
• 48 -36 -93 57 -- Best control channel for U-NII-1 range.

U-NII-3 channel range

• 149 -39 -90 51
• 161 -39 -90 51

So for my Macbook Pro, Control Channel 48 remains the clear winner in U-NII-1, but U-NII-1 is now beating U-NII-3 not only for Quality index but also for RSSI, whereas for my Galaxy Note 8, the RSSI values were much higher in U-NII-3. This is in contradiction to the channel power values reported by ASUS in their FCC info where U-NII-3 had higher channel power. Perhaps the reason for the higher channel power is to overcome higher noise levels in that frequency range. Whatever the reason, the practical outcome is that Control Channel 48 is definitely preferred in the 5 GHz band range for the RT-AC86U for its current firmware (382.18848), and its probably worth investigating this kind of thing for your own router if trying to maximize range.

 

[ColinTaylor]

Try this command (I don't know if it works on your routers).

wl -i eth2 chanspec_txpwr_max

 

[sfx2000]

Turn BF off and try again... there's a glitch in the rules regarding TxBF and EIRP

In any event, vendors can do what they want as long as EIRP doesn't exceed regulatory requirements - some take a global view to reduce R&D and QA efforts.

Looking at EIRP and RSSI - needs more than just casual observations - one really need a spectrum analyzer and a controlled environment - e.g. conducted path/cabled up - and then one can start to make some accurate assessments on radiated performance - then one can back things up with a radiated chamber similar to what SmallNetBuilder uses for their testing.

 

[maxbraketorque]

Turn BF off and try again... there's a glitch in the rules regarding TxBF and EIRP

In any event, vendors can do what they want as long as EIRP doesn't exceed regulatory requirements - some take a global view to reduce R&D and QA efforts.

Looking at EIRP and RSSI - needs more than just casual observations - one really need a spectrum analyzer and a controlled environment - e.g. conducted path/cabled up - and then one can start to make some accurate assessments on radiated performance - then one can back things up with a radiated chamber similar to what SmallNetBuilder uses for their testing.

U-NII-1 channel values were the same with all BF disabled, while U-NII-3 channels were a few points worse in RSSI (and unchanged in background). In all the testing up to that point, I was very careful to maintain consistent conditions through the tests, including keeping myself in a consistent orientation relative to the computer. However this time, I got up to grab a glass of water in between observations, and when that happened, RSSI increased dramatically. So I redid all the observations with me away from the computer along with BF disabled, and this time, the Quality values were the same for U-NII-1 and U-NII-3. Background was still higher for U-NII-3 than for U-NII-1, but RSSI was higher in U-NII-3 by the same amount that background was higher.

I'm sure there is much to be learned from more detailed measurements, but it still doesn't seem like a bad idea to select the channel that gives best RSSI and/or Quality value.