Unstable WiFi on both RT-AC68U and RT-AC87U, forced to reconnect wireless clients

[RamGuy]

Greetings,

I'm having some stability issues with both Asus RT-AC68U and RT-AC87U on both the 2,4GHz and 5,0GHz band.

It has nothing to do with range or the signal itself, my clients are bascially spread out within 5-15 meters and all have great singal strength and quality. The bandwidth and connection itself is working very well on all the devices.


My problem is that both cellphones (iPhone 6 + 6 Plus), our iPad Air 2 and our two MacBook Pro's (one 13" Retina and another 15" Retina) seems to completely drop the WAN connection using the WiFi.

They dont drop the WiFi and still have great connection but WAN connections get slow or not working at all and a simple disconnect and reconnect fixes this emiddietly.


I have tried with the latest official Asus firmware, and with both 378.50 and 376.49_0-5. Its the same issue with all of them.


The one thing I've tried thus far is to run WiFi Analyser (Android) and pinpointed the best channels for both 2,4GHz and 5,0GHz in my area as the "auto" function doesnt seem to be doing anything. But this haven't done anything to improve upon this issue. I have also tried to disable and enable various advanced features to no avail.


In terms of our clients both iPhones have been through iOS 8.1.X, various iOS 8.2 and now on the latest iOS 8.3 BETA so they been through various software editions with various WiFi drivers, same goes for the iPad Air 2.

The MacBooks have been on Mac OS X Yosemite 10.10.2 and 10.10.3 with the same issues. I have also tested with a Samsung Galaxy Note 4 and Nokia Lumia 930 having the same issues.



None of my desktop clients running cable is having any issues at all. What might be the source of my problems? Seems odd that both routers would be defective? And they have worked for quite some time before, but not as of the past 1-2 months or so. I'm not 100% sure when this issue started to kick in to be honest, but now wireless clients needs to reconnect at least 4-5 times per day making the WiFi extremely unstable and forcing us to disconnect them alltogether as many push notifications on our phones etc rely on working WAN over WiFi for us to get notifications at all and we don't want to risk our phones being in our pockets with WAN bugs for several hours until we notices.

 

[johtaja]

Are those EU routers? I have read that Apple products (as well as Intel wifi cards) have troubles with 802.11d+h regulation, and in newer Asus fw's in EU routers it is forced and can't be turned off. Older Asus firmwares dont have that, I mean those that have only lower 5GHz channels available. Did this problem come in some fw update? If so, try older FW in routers, or you could modify CFE to regulation free. Or try hggomes's Merlin forks, they are modified to regulation free.

 

[InspectHerGadget]

I would try turning it off for 30 minutes and turning it back on.

This seemed to work for me and I'm getting similar issues to you.

I'm using latest merl in release.

If you have been fiddling with settings then I'd do a reset as well.

It is possible that a simple quick power cycle may not completely reset everything, the thinking being that with so many electronics in the box it may take time to drain current and reset everything.

I seem to be getting very stable wifi now on my Air 2 and Nokia 1520.

 

[RMerlin]

Are those EU routers? I have read that Apple products (as well as Intel wifi cards) have troubles with 802.11d+h regulation, and in newer Asus fw's in EU routers it is forced and can't be turned off. Older Asus firmwares dont have that, I mean those that have only lower 5GHz channels available. Did this problem come in some fw update? If so, try older FW in routers, or you could modify CFE to regulation free. Or try hggomes's Merlin forks, they are modified to regulation free.

In EU Asus enforces 802.11h. However, Macs tend to work better if you also enable 802.11d, so you need to set the reg mode to 802.11d+h on the Professional tab (it defaults to just 802.11h).

 

[RamGuy]

Are you sure? Mine defaults to 802.11h+d no matter what I do on the 5.0GHz band. Doesn't help if I manually set it to off or 802.11h it still pops right back to 802.11h+d. Seems like modifying the bootloader is the only way around this? Changing to a non-DFS channel wont change anything?

 

[john9527]

Are you sure? Mine defaults to 802.11h+d no matter what I do on the 5.0GHz band. Doesn't help if I manually set it to off or 802.11h it still pops right back to 802.11h+d. Seems like modifying the bootloader is the only way around this? Changing to a non-DFS channel wont change anything?

No, it's coded that way in the firmware.

Note to Merlin....you need to make a companion fix in init-broadcom.c to only force reg_mode=h if it's "off" or "d".

EDIT: Nevermind.....see you already did the init-broadcom update

 

[highwire]

I'm glad I have a CA version and not an EU version of the 87U because if I enable 802.11h support in 5GHz professional tab, all my Apple devices have atrociously bad wifi performance. When it's disabled though, it works well. There doesn't seem to be a way to set 802.11d on the CA version of the 87U.

 

[Bobbulf]

We also had problems with random drops on signal and loss of connection on various iPhone 6's, iPads, Mac and PC's on 87U (Norway). I had a suspicion that when there were clients far away from the router, and the beamforming function tried to send strong signal in that direction, it could be a problem when many clients were connected on different floors and directions, and the beam should go "everywhere". I turned the beamfunction off (wireless/professional,explicit and universal beamforming), and the signal is much more stable for us now.

 

[RamGuy]

I'm glad I have a CA version and not an EU version of the 87U because if I enable 802.11h support in 5GHz professional tab, all my Apple devices have atrociously bad wifi performance. When it's disabled though, it works well. There doesn't seem to be a way to set 802.11d on the CA version of the 87U.

That's correct, there is no way to enable 802.11h+d on the RT-AC87U on 5.0GHz, I guess this has something to do with the Quantenna WiFi NIC and the lack of customization in the drivers / firmware. Seriously this Quantenna nonsense is in of itself a reason to drop the RT-AC87 and go for the RT-AC68 or the RT-AC3200 instead.. Broadcom seems to be much more mature and customizable in terms of the drivers / firmware.


Doesn't really matter here in Norway though, as we seem to be forced into having 802.11h+d enabled on the 5.0GHz regardless.. Or 802.11h when it comes to the RT-AC87 as it doesn't seem to support 802.11d for some reason.

We also had problems with random drops on signal and loss of connection on various iPhone 6's, iPads, Mac and PC's on 87U (Norway). I had a suspicion that when there were clients far away from the router, and the beamforming function tried to send strong signal in that direction, it could be a problem when many clients were connected on different floors and directions, and the beam should go "everywhere". I turned the beamfunction off (wireless/professional,explicit and universal beamforming), and the signal is much more stable for us now.

This was the first thing I suspected so I disabled beamforming on the SSID that phones and tablets connect to with no improvement. Also tried to enabled and disable AirTime Fairness without any luck.


I have loaded Hoggmes custom fork of Asus Merlin as it features the capability of disabling regulations such as 802.11h and 802.11d before I consider starting to modify the bootloader.. Thus far it seems to be doing wonders, nothing has yet to hang and the 802.11ac connection towards our MacBook Pro's seems much healthier.

 

[highwire]

I should clarify that enabling 802.11h on the 5GHz professional page of my 87U only degrades the wifi connection of Apple products. My Windows, TV, and other devices continue to work well with it enabled. So I think the issue here is with Apple, not ASUS's or Merlin's F/W. Given all the recent reports of horrible wifi performance with recent Apple releases, I'm not really surprised by this.

 

[RamGuy]

It seems to be affecting mostly Apple devices and Windows devices running Intel NICs from what I'm able to gather from the Internet.

One could always argue that pushing these locked firmware's before companies like Apple have made sure it will actually work as intended is rather stupid. And why is 802.11h(+d) even being enforced in the first place, even on non-DFS channels like 36, 40, 44 and 48?


I guess the sole reason why Apple haven't really been keeping track on this issue is because very few routers and firmware actually enforces 802.11h(+d) in the first place. Most defaults to having it off completely, even when using Auto (with DFS channels being included into the mix).

And considering this is a European Union only thing thus far, and quite few manufactures have started to enforce this in the first place. I don't know of any other than Asus in the consumer market? I guess Apple haven't got that many reports of this specific problem so it's not the highest priority in neither Mac OS X or iOS at the moment.


From what I understand the European Union does not require enforced 802.11h(+d) to be delivered through new routers and new firmware release until 2016 so Asus seems to be ahead of the curve here for some reason.


EDIT:

Regardless, not having the option for 802.11h+d on the RT-AC87U is still odd. Routers with Broadcom SoCs have been supporting 802.11h+d for literally ages now and from what I have gathered online it seems like one would opt for 802.11h+d over 802.11h-only in about every occasion where you would want to enable regulation mode.

 

[Morbeas]

I have a RT-AC87R running the latest Merlin 378.50 firmware and I can not stay connected on the 5 GHz radio (2.4 GHz channel seems to be OK) to save my life. It's constantly dropping clients. Is this simply because Asus needs to update their drivers or faulty hardware or what? I can't believe I spent all this money on a cutting edge router only to have nothing but problems since day 1.

So frustrating.

 

[L&LD]

I have a RT-AC87R running the latest Merlin 378.50 firmware and I can not stay connected on the 5 GHz radio (2.4 GHz channel seems to be OK) to save my life. It's constantly dropping clients. Is this simply because Asus needs to update their drivers or faulty hardware or what? I can't believe I spent all this money on a cutting edge router only to have nothing but problems since day 1.

So frustrating.

That is exactly what you're signing up when spending money on a cutting edge router. For some, the frustration is more than offset when they have helped worked the bugs out after some weeks or months of beta testing for the manufacturer.

If you can't return, sell or otherwise use another router for now, your job is to find what is causing the dropouts, if you can.

Buying a cutting edge anything is a double edge sword. But the thing that is guaranteed is that it is up to the customer to ensure it works for them before being stuck with it.

 

[RamGuy]

It might seem like I was a bit too quick.. I'm still having the same issues but not as often as before. And it only seems to be the Asus RT-AC87U and not the RT-AC68U at all. Which sounds about right as I didn't have any of these issues before replacing my RT-AC66U with a RT-AC87U so it might be the RT-AC87U being defective, or some incompatibility with the quantenna wifi nic.

 

[oranheim]

I have loaded Hoggmes custom fork of Asus Merlin as it features the capability of disabling regulations such as 802.11h and 802.11d before I consider starting to modify the bootloader..

Where do I find the Hoggmes custom fork?

 

[L&LD]

PM directly.

http://www.snbforums.com/members/hggomes.8609/

 

[millercentral]

Also, do you have the SSIDs for 2.4Ghz and 5Ghz named the same? I've seen Apple products struggle to select and stay on a band when both bands have the same SSID name. The end result is drops/WAN loss.

 

[gotdade]

Could this be related to the problems with how OS X handle 802.11h Power Constraint Information Element described in
https://discussions.apple.com/thread/6825371

 

[sovereignty68]

Greetings,

I'm having some stability issues with both Asus RT-AC68U and RT-AC87U on both the 2,4GHz and 5,0GHz band.

It has nothing to do with range or the signal itself, my clients are bascially spread out within 5-15 meters and all have great singal strength and quality. The bandwidth and connection itself is working very well on all the devices.


My problem is that both cellphones (iPhone 6 + 6 Plus), our iPad Air 2 and our two MacBook Pro's (one 13" Retina and another 15" Retina) seems to completely drop the WAN connection using the WiFi.

They dont drop the WiFi and still have great connection but WAN connections get slow or not working at all and a simple disconnect and reconnect fixes this emiddietly.


I have tried with the latest official Asus firmware, and with both 378.50 and 376.49_0-5. Its the same issue with all of them.


The one thing I've tried thus far is to run WiFi Analyser (Android) and pinpointed the best channels for both 2,4GHz and 5,0GHz in my area as the "auto" function doesnt seem to be doing anything. But this haven't done anything to improve upon this issue. I have also tried to disable and enable various advanced features to no avail.


In terms of our clients both iPhones have been through iOS 8.1.X, various iOS 8.2 and now on the latest iOS 8.3 BETA so they been through various software editions with various WiFi drivers, same goes for the iPad Air 2.

The MacBooks have been on Mac OS X Yosemite 10.10.2 and 10.10.3 with the same issues. I have also tested with a Samsung Galaxy Note 4 and Nokia Lumia 930 having the same issues.



None of my desktop clients running cable is having any issues at all. What might be the source of my problems? Seems odd that both routers would be defective? And they have worked for quite some time before, but not as of the past 1-2 months or so. I'm not 100% sure when this issue started to kick in to be honest, but now wireless clients needs to reconnect at least 4-5 times per day making the WiFi extremely unstable and forcing us to disconnect them alltogether as many push notifications on our phones etc rely on working WAN over WiFi for us to get notifications at all and we don't want to risk our phones being in our pockets with WAN bugs for several hours until we notices.

I have exact same issue. WiFi on mobile devices gradually dies. WiFi is still connected but internet won't work until I turn off & on the WiFi function on the client devices.

 

[sfx2000]

Could this be related to the problems with how OS X handle 802.11h Power Constraint Information Element described in
https://discussions.apple.com/thread/6825371

Has nothing to do with power management - but yes, OSx and iOS do respect the 802.11h IE's in the beacon... 11h has to do with max Tx power on any STA, which in a BSS, both the AP and the client's are STA's...

Don't get 11h confused with 11e...