Wi-Fi Signal Strength RT-AC86U v RT-AX88U

[Phil Outram]

You're providing 'an inaccurate statement'. Why don't you just simply answer the question instead? I've stated my testing procedure, maybe you can do the same?

I did not reproduce anything except seeing that removing an antennae breaks the router (no kidding). A single antennae/stream on the 2.4GHz band cannot give an AC client a 300Mbps connection speed at 40 feet and certainly not obtain over 100Mbps throughput either.

What you've stated is far more than what I can 'prove' with my router, sorry.

Again this is an inaccurate statement, you are confusing multiple antennas with connection speed of the client. Multiple antennas are required for MIMO not link rate. The only reason why the link rate would be reduced with a single antenna is because the signal is weaker not because multiple antennas are required to achieve that sync rate.

You can verify this with your own test. Remove all your antennas except for the one near switch port 8 and you'll have exactly the same results as your original test on the 2.4ghz band.

 

[thiggins]

First, @L&LD and @Phil Outram cool it with the personal attacks. Confine your comments to the topic, not each other.

I learned a lesson about not removing antennas long ago; it doesn't produce reliable results. Link rates get reported in management frames, which don't depend on MIMO. AP determines the MCS to set mainly by using STA RSSI.

Leave all antennas in place when you do your testing. I suggest focusing on throughput as the primary indicator. RSSI and link rate are secondary and can be used to theorize about why you might be getting the throughput you're getting.

 

[Phil Outram]

First, @L&LD and @Phil Outram cool it with the personal attacks. Confine your comments to the topic, not each other.

I learned a lesson about not removing antennas long ago; it doesn't produce reliable results. Link rates get reported in management frames, which don't depend on MIMO. AP determines the MCS to set mainly by using STA RSSI.

Leave all antennas in place when you do your testing. I suggest focusing on throughput as the primary indicator. RSSI and link rate are secondary and can be used to theorize about why you might be getting the throughput you're getting.

Not trying to attack anyone, to be honest I'm just trying to help out another member by sharing my findings. I'll admit I am frustrated but only because what seems to be obvious to me isn't getting across.

However in regards to the test, you can't do the test with all antennas in place because that is the only way to demonstrate the issue. You have to remove them to prove the point that a single antenna does not reduce the link rate or throughput. Although you have just confirmed this to be the case as well, so a moot point now. It only really effects MIMO, and may lower the link rate but only because the signal is weaker with just a single antenna transmitting not because they are required to achieve it.

Which takes us back to the original point, users are complaining about lower signal strength on the 2.4ghz band with the AX88U on 384.15, and the reason for this is because now only a single antenna transmits it. L&LD was stating that this could not be the case because the link rate was still 300mbps, but as you have just pointed out the number of antennas does not affect the link rate (other than by simply having a weaker signal) and therefore is not proof that multiple antennas still transmit the 2.4ghz radio on 384.15.

The entire point of this was that only a single antenna now transmits the 2.4ghz radio, and you can only really prove that is the case by removing them and demonstrating this to be the case. What I am trying to say is with L&LD test with all antennas attached he achieved a link rate of 300mbps at 40 feet, and I am saying he will get the same result removing all the antennas except for the one closest to switch port 8, and thus proving that only a single antenna now transmits this band.

To be clear, I have carried out this test myself and get the exact same results with 384.15 with either just the antenna near switch port 8 attached or all of them attached.

Test 1 - 384.15 with all antennas attached (without my wifi cable just attached normally) - Sync rate 300mbps, throughput was around 65Mbps copying to server share (server connected via gigabit ethernet)

Test 2 - 384.15 with all antennas removed except for the one closest to switch port 8 - results pretty much exactly the same as from test 1.

Test 3 - 384.15 with all antennas attached except for the one closed to switch port 8 - could not connect at all to the 2.4ghz band

So as I stated before, only 1 antenna is being used, does this now finally prove the point?

 

[thiggins]

I get that you are frustrated. But repeating the point is not going to win the argument.

Your argument would be more convincing if you replaced the antennas with 50 ohm terminators to ensure no signal leakage.

 

[Phil Outram]

I get that you are frustrated. But repeating the point is not going to win the argument.

Your argument would be more convincing if you replaced the antennas with 50 ohm terminators to ensure no signal leakage.

Ok I'll do that, should have them tomorrow morning via Amazon prime.

 

[PeterT]

Something I've wondered (but don't feel I'm qualified to really defend it or test it). Would leaving the antenna in place but covering it by a farrady cage show anything? For instance with the supposed 2.4Ghz antenna covered is there still access to 2.4Ghz?

 

[htismaqe]

I get that you are frustrated. But repeating the point is not going to win the argument.

Your argument would be more convincing if you replaced the antennas with 50 ohm terminators to ensure no signal leakage.

I was just getting ready to post this. I work on a test bench when I get new routers in and the first thing I do is remove the antennas to prevent them from transmitting all through the building. They don't stop transmitting to the TEST BENCH. Even without the antennas attached, the router still puts out a signal. Just not strong enough to leave my lab.

At a distance of 1 foot, there's a good chance there's enough signal leakage to taint the test.

 

[bluzfanmr1]

Is it just me or has there been a surge in newbies on here, most with posts numbering in the single digits, who are very disrespectful to those that have been here for years, have more than 10,000 posts, many of which are infinitely helpful to the others in the forum?

 

[Val D.]

Is it just me or has there been a surge in newbies on here

This is actually good. What is bad very few are searching/reading the information already available on SNB forums.

 

[Mrk26]

I also had issue with WiFi signal on latest 384.15, back to rock solid 384.13 solved all issues related to poor WiFi.

 

[RMerlin]

Folks, take a deep breath and relax, or I will put this thread under lock and key - and I'll keep the key.

 

[Val D.]

Test 1 - 384.15 with all antennas attached (without my wifi cable just attached normally) - Sync rate 300mbps, throughput was around 65Mbps copying to server share (server connected via gigabit ethernet)

Test 2 - 384.15 with all antennas removed except for the one closest to switch port 8 - results pretty much exactly the same as from test 1.

Test 3 - 384.15 with all antennas attached except for the one closed to switch port 8 - could not connect at all to the 2.4ghz band

So as I stated before, only 1 antenna is being used, does this now finally prove the point?

I believe your experiment proves my theory that one specific antenna is now used for something else in connection to new AX standards support. If you remove this antenna, some parameter inside the router goes below the set threshold and cuts off 2.4GHz band. If you have it in place, the 2.4GHz band works with more than one streams as expected, even though you don't have the other antennae attached. This is possible because the client you test with is close enough to the router. I remember with all 3 external antennae removed RT-AC86U (there is one internal for 5GHz) was able to provide acceptable coverage about 10ft around the router on both 2.4GHz and 5GHz bands. This "issue" is something coming from Broadcom drivers, part of the new ASUS firmware. Only someone from Broadcom may explain eventually how things work and why we observe this behavior. I don't see it as a "defect" though. Yes, it doesn't allow you to use the router with antennae in different places like before, but I'm in doubt the manufacturers take into consideration this specific use case. The router was never designed to be used this way in first place. Go back one firmware version and your problem is solved.

 

[CaptnDanLKW]

I learned a lesson about not removing antennas long ago; it doesn't produce reliable results. Link rates get reported in management frames, which don't depend on MIMO. AP determines the MCS to set mainly by using STA RSSI.

Leave all antennas in place when you do your testing. I suggest focusing on throughput as the primary indicator. RSSI and link rate are secondary and can be used to theorize about why you might be getting the throughput you're getting.

I don't own an AX88U but I've kept up on my thread readings and I have an observation.

The latest firmware adds the very new WPA3 and OFDMA protocols from 384_7756. IIRC, these require protected management frames. As our Jedi leader points out above, the link speed reported is in the management frames. Being new code it's possible the link speed isn't negotiated properly or reported accurately. (side thought - the client OS driver could equally be at fault for this possible scenario). If we assume the reported link speed is accurate, we can go down weird rabbit holes trying to test throughput vs. reported link speed trying to make sense of a odd results.

No conclusions, just another data point to consider.

 

[Phil Outram]

Is it just me or has there been a surge in newbies on here, most with posts numbering in the single digits, who are very disrespectful to those that have been here for years, have more than 10,000 posts, many of which are infinitely helpful to the others in the forum?

Post count is not a measure of somebody's knowledge. In my case this 'newbie' is a senior solutions architect with over 22 years of experience deploying servers, networking and other IT projects to large clients. So despite my low post count I am in no means a 'newbie' when it comes to such matters thank you very much!

I believe your experiment proves my theory that one specific antenna is now used for something else in connection to new AX standards support. If you remove this antenna, some parameter inside the router goes below the set threshold and cuts off 2.4GHz band. If you have it in place, the 2.4GHz band works with more than one streams as expected, even though you don't have the other antennae attached. This is possible because the client you test with is close enough to the router. I remember with all 3 external antennae removed RT-AC86U (there is one internal for 5GHz) was able to provide acceptable coverage about 10ft around the router on both 2.4GHz and 5GHz bands. This "issue" is something coming from Broadcom drivers, part of the new ASUS firmware. Only someone from Broadcom may explain eventually how things work and why we observe this behavior. I don't see it as a "defect" though. Yes, it doesn't allow you to use the router with antennae in different places like before, but I'm in doubt the manufacturers take into consideration this specific use case. The router was never designed to be used this way in first place. Go back one firmware version and your problem is solved.

That is exactly my thinking, and of course I have rolled back to 384.14 so that my environment still works.

I'm also in contact with Asus who are looking into it. Obviously I can stay on 384.14 but I don't like the idea of being permanently stuck on out of date firmware for security reasons (and of course I'd like the use of new features as and when then become available). Therefore I will continue efforts with Asus to see if I can get them to acknowledge the issue and perhaps they can resolve it for future releases. I am certain it is down to the new wi-fi drivers and perhaps it is now a design limitation and there is no fix, or maybe there is one and Asus/Broadcom need to resolve it. Only time will tell.

And in other news, the 50 ohm terminators turned up, and of course I get the same results as my previous test which proves that there is no signal leakage skewing the result. Only a single antenna now transmits the 2.4ghz radio.

 

[ColinTaylor]

And in other news, the 50 ohm terminators turned up, and of course I get the same results as my previous test which proves that there is no signal leakage skewing the result. Only a single antenna now transmits the 2.4ghz radio.

I think you'll have to summarise what tests you have now performed with your dummy loads vs. antennas as it is no longer obvious by trying to read through all the past posts.

So we'd be particularly interested in actual throughput numbers (at say 5 feet distance) rather than any reported link rate. And comparisons with all 4 antennas, 4 dummy loads, and 1 antenna (in positions a, b, c and d) + 3 dummy loads.

 

[Phil Outram]

I think you'll have to summarise what tests you have now performed with your dummy loads vs. antennas as it is no longer obvious by trying to read though all the past posts.

So we'd be particularity interested in actual throughput numbers (at say 5 feet distance) rather than any reported link rate. And comparisons with all 4 antennas, 4 dummy loads, and 1 antenna (in positions a, b, c and d) + 3 dummy loads.

I did the same 3 tests in post #43 above but I've got a day off tomorrow and I can certainly do them all again one last time so it is clear.

 

[ColinTaylor]

I did the same 3 tests in post #43 above but I've got a day off tomorrow and I can certainly do them all again one last time so it is clear.

Thanks. In theory, a baseline test (with all normal antennas in place) at relatively short range (~5 feet) using a multi-stream client ought to be producing a throughput in excess of 150Mbps. [Assuming 40MHz bandwidth and no nearby interference]

 

[RMerlin]

I'm also in contact with Asus who are looking into it.

You should make sure first that 384_7968 (the latest GPL for the RT-AX88U) still has the issue. 384.15 is still based on 384_7756, as I only received the 7968 GPL last night.

 

[thiggins]

I’ve cleaned up the thread and deleted personal attacks. If any new ones appear I’ll close the thread and temporarily ban the offenders.

 

[Val D.]

and of course I'd like the use of new features as and when then become available

Then you may have to stick to the original router configuration with all 4 antennae attached to it and look for other options for outside cameras. I would connect a separate 2.4GHz AP outside and keep the RT-AX88U functionality with the latest firmware available. I don't think future firmware updates will "reverse" the current situation.

Amazon is currently selling TP-Link Omada N300 WAP + PoE Injector, waterproof, dustproof, etc. for $38.
https://www.amazon.com/dp/B07CG3YRTR/?tag=snbforums-20
I'm sure you can get something similar from other manufacturers on a similar price or even lower off eBay. And powered by PoE you can place it anywhere you want to achieve maximum stable connection with whatever is outside. Isn't it a better solution?