AX86 Giving Slow Speeds (Brand New)

[L&LD]

Okay, since I'm in testing mode with half a dozen routers right now, I turned on QoS, rebooted, and now the highest speeds I get are about 495Mbps download (upload is not affected on my i7 desktop connected via 1GbE Ethernet).

I turned off QoS and I waited 15 minutes or so and the speeds are the same (I'm using the speedtest built into the 386.1 Beta 2 firmware).

I reboot once more, but on instinct, I test with a different server. Bam! 940Mbps down, 950Mbps up! I then test via the Edge browser on my i7 and get even better speeds (1Gbps down, 980Mbps up).

I then go back to my own ISP's server and... back down to 625Mbps down (upload not affected). Testing via Edge shows similar low speeds for download.

So, is it the router, or is it the ISP (throttling us)?

I'm still in test mode, will report anything new later.

 

[JChoi]

You won't get more than 80 to 90 Mbps in speed tests using 2.4GHz (for devices with two antennas). Less than that for devices with one antenna.

But it has 3 antennas. LOL

 

[JChoi]

Okay, since I'm in testing mode with half a dozen routers right now, I turned on QoS, rebooted, and now the highest speeds I get are about 495Mbps download (upload is not affected on my i7 desktop connected via 1GbE Ethernet).

I turned off QoS and I waited 15 minutes or so and the speeds are the same (I'm using the speedtest built into the 386.1 Beta 2 firmware).

I reboot once more, but on instinct, I test with a different server. Bam! 940Mbps down, 950Mbps up! I then test via the Edge browser on my i7 and get even better speeds (1Gbps down, 980Mbps up).

I then go back to my own ISP's server and... back down to 625Mbps down (upload not affected). Testing via Edge shows similar low speeds for download.

So, is it the router, or is it the ISP (throttling us)?

I'm still in test mode, will report anything new later.

So I turned on QoS. But not sure what options I should be checking... I don't even know what GeForce Now QoS is

 

[L&LD]

Why would you turn QoS on? Did you not see when I did so my speeds plummeted to almost half?

I am tempted to do a full reset right now, but I am still in experimenting mode.

 

[ColinTaylor]

But it has 3 antennas. LOL

<not sure whether you meant that as a joke, but> Your router does but your clients don't (as shown in your screenshot).

 

[SoCalReviews]

My router goes to channel 11.

 

[SoCalReviews]

Okay, since I'm in testing mode with half a dozen routers right now, I turned on QoS, rebooted, and now the highest speeds I get are about 495Mbps download (upload is not affected on my i7 desktop connected via 1GbE Ethernet).

I turned off QoS and I waited 15 minutes or so and the speeds are the same (I'm using the speedtest built into the 386.1 Beta 2 firmware).

I reboot once more, but on instinct, I test with a different server. Bam! 940Mbps down, 950Mbps up! I then test via the Edge browser on my i7 and get even better speeds (1Gbps down, 980Mbps up).

I then go back to my own ISP's server and... back down to 625Mbps down (upload not affected). Testing via Edge shows similar low speeds for download.

So, is it the router, or is it the ISP (throttling us)?

I'm still in test mode, will report anything new later.

If you wear a tin foil hat you might get 1 Gbps.

 

[L&LD]

Okay, as 'proof' to my suspicions that turning on QoS and turning it off isn't the same as simply not turning it on at all (after a full reset and minimal config without using a saved backup config file), I found out the following. (Continuing from my post above).

• I 'withdrew' consent from Administration, Privacy for QoS. Rebooted. No change in speeds via the RT-AX86U's speed test*.
  • *Note: From a fresh reset and JFFS format after flashing RMerlin 386.1 Beta 2, speeds were 940 down and 940 up.
  • After 'touching' QoS, speeds vary between 495Mbps down to 625Mbps down (upload speeds not affected).
  • If using a different Server than my ISP's, speeds back up to better than at fresh reset values, except for Ping and Jitter which is 20x to 5 times worse, respectively.
• I went back into the QoS menu and set it to 'On', but did not hit apply. The settings I used when I had actually turned it on and used it was still there!
• I changed them back to Auto Saved them and turned off QoS again. Rebooted the router.
• The settings I had changed originally (an hour ago now?) were still there (even with QoS actually 'Off' and seemingly put back to its defaults just before the reboot).

I have to go to a few customers and deliver their new routers now, but this is obviously a case of needing to fully reset the router once more (note to self: do not touch QoS in the future).

No tin foil hat required. The tests I've outlined are easily repeated and verified. Touching QoS with a 1Gbps ISP connection (symmetrical or otherwise) is just a bad idea.

 

[JChoi]

Okay, as 'proof' to my suspicions that turning on QoS and turning it off isn't the same as simply not turning it on at all (after a full reset and minimal config without using a saved backup config file), I found out the following. (Continuing from my post above).

• I 'withdrew' consent from Administration, Privacy for QoS. Rebooted. No change in speeds via the RT-AX86U's speed test*.
  • *Note: From a fresh reset and JFFS format after flashing RMerlin 386.1 Beta 2, speeds were 940 down and 940 up.
  • After 'touching' QoS, speeds vary between 495Mbps down to 625Mbps down (upload speeds not affected).
  • If using a different Server than my ISP's, speeds back up to better than at fresh reset values, except for Ping and Jitter which is 20x to 5 times worse, respectively.
• I went back into the QoS menu and set it to 'On', but did not hit apply. The settings I used when I had actually turned it on and used it was still there!
• I changed them back to Auto Saved them and turned off QoS again. Rebooted the router.
• The settings I had changed originally (an hour ago now?) were still there (even with QoS actually 'Off' and seemingly put back to its defaults just before the reboot).

I have to go to a few customers and deliver their new routers now, but this is obviously a case of needing to fully reset the router once more (note to self: do not touch QoS in the future).

No tin foil hat required. The tests I've outlined are easily repeated and verified. Touching QoS with a 1Gbps ISP connection (symmetrical or otherwise) is just a bad idea.

When you said reset it... You mean like... Pressing and holding that Reset button for 10 seconds and starting brand new? What do you mean by JFFS Format... I don't know how to do this.

Why would you turn QoS on? Did you not see when I did so my speeds plummeted to almost half?

I am tempted to do a full reset right now, but I am still in experimenting mode.

I turned it on because it was "Off" just to see if it fixed anything. But nope.

 

[SoCalReviews]

Okay, as 'proof' to my suspicions that turning on QoS and turning it off isn't the same as simply not turning it on at all (after a full reset and minimal config without using a saved backup config file)....

No tin foil hat required. The tests I've outlined are easily repeated and verified. Touching QoS with a 1Gbps ISP connection (symmetrical or otherwise) is just a bad idea.

Interesting info. Turning QoS on... causing performance issues has been brought up before in this forum. The performance issues continuing even after it has been turned off is sort of a new revelation. I wonder if Asus knows about this issue.

...and the tin foil hat suggestion was just for testing... it could help improve throughput due to time delay related signal reflections in the 5Ghz band... "possibly".

 

[SoCalReviews]

When you said reset it... You mean like... Pressing and holding that Reset button for 10 seconds and starting brand new? What do you mean by JFFS Format... I don't know how to do this.

I turned it on because it was "Off" just to see if it fixed anything. But nope.

He's saying keep QoS OFF and don't even turn it on at all if you have a 1Gbps internet service. If you turn it on you could have performance issues and to solve them you then might have to do a hard button reset on the Asus router as described by Asus here.

 

[JChoi]

He's saying keep QoS OFF and don't even turn it on at all if you have a 1Gbps internet service. If you turn it on you could have performance issues and to solve them you then might have to do a hard button reset on the Asus router as described by Asus here.

Ah yes, I had it on and did a reset... and thankfully i'm still having the 1gbps connected wired.

Just wireless is horrid at 5ghz with only getting 300s mbps download... upload is fine

 

[tallytr]

It looks higher than what speedtest.net gives me for the 5g devices

I would not post your clients MAC addresses, but that's just be, blank them out?

 

[tallytr]

Not sure, did someone ask about the wireless adapter in the PC/notebook? AC card? Maybe not set to Gigabit?

 

[JChoi]

I would not post your clients MAC addresses, but that's just be, blank them out?

I did not know that was a threat. Thank you.

Not sure, did someone ask about the wireless card in the notebook, maybe wireless N?
I think 300Mbps is the max for that card?

On the Laptop it's worse... But the Note10+ Phone is wifi6 compatible at 5gh it's 300 mbps.

 

[ColinTaylor]

I did not know that was a threat.

It's not.

 

[tallytr]

It's not.

Interesting, I always thought that would identify ones client like a house number or so, learnt something new I guess, thank you.

 

[ColinTaylor]

Interesting, I always thought that would identify ones client like a house number or so, learnt something new I guess, thank you.

It does uniquely identify a device on the local network. Unlike a public IP address it cannot be used to connect to from the internet and cannot be used to determine your ISP or geographic location. The only information it conveys is the manufacturer of the network adapter. So I don't see what "threat" posting a low resolution picture containing MAC addresses would pose. Maybe we have different interpretations of the word threat.

 

[JChoi]

It does uniquely identify a device on the local network. Unlike a public IP address it cannot be used to connect to from the internet and cannot be used to determine your ISP or geographic location. The only information it conveys is the manufacturer of the network adapter. So I don't see what "threat" posting a low resolution picture containing MAC addresses would pose. Maybe we have different interpretations of the word threat.

This is good info. Thanks.

But still not having good Wifi6 5ghz speeds -.-

 

[SoCalReviews]

This is good info. Thanks.

But still not having good Wifi6 5ghz speeds -.-

Are you saying your "Internet Speed Test" isn't getting good Wifi speeds?

What do the connection speeds say for your Devices under View List from the AX86U router's main Network Map page?

On the AX86U router I'm using right now it shows I have devices connected at 5Ghz ranging from an estimated 300 to 800+ Mbps. It all depends on the quality of the radios those devices have, the open 5Ghz channels that are available in your area, the bandwidth you are getting on those channels (20, 40, 80, 160Mhz), whether they are N, AC or AX, how many antennas they have, how far the devices are away from the router, the materials in the structure you reside in (concrete walls, wood, plaster, etc..) and the objects that the 5Ghz signal can reflect off of. All of this can effect that signal.

There are so many variables that can effect this and really 300Mbps isn't a bad speed in general. If you are downloading and uploading 4K video content at the same time via Wifi then of course you want a higher maximum throughput. If you want to get maximum Wifi speeds you are going to have to experiment more with the settings. Internet Speed Test isn't the best way to test it either since that can vary from one test to the next and especially over a 1Gbit connection. A better test would be transferring content via the router's Wifi from another device within your local network and if you require stable high Wifi 6 throughput over longer distances you should be using two AX86Us.