ASUS RT-AC87 Firmware - Official Releases

[Richard1864]

My wife works from home (laptop on 2.4GHz) , obviously she requires
a reliable environment, to elaborate, she VPNs to a remote server at least
8 hours a day. Usually she just leaves her laptop connected. She hasn't had
any disconnects.

My TV and Blu-Ray player are my only 5GHz devices. We watch movies often,
(large mkv, many in the 10 gig area) streamed from my Synology NAS,
and have never had a disconnect. (knock on wood)

The router usually has ~10 devices at a time connected to it.

My scenario: I have no Apple products. Because of the probs I've heard
here, I'm still running my unit at a minimum, meaning, no USB, no QoS, no
AI protection, etc. I have a few static addresses and port forwarding for SFTP
(no problem with port forwarding either). I made a change to the router when
I upgraded my NAS recently, but before that, it was up for over 30 days
without a hiccup.

With the wide range I'm reading here from lots of problems to no problems,
really makes me wonder if there are hardware issues, or at least maybe
hardware tolerance issues with these units. I don't know, maybe I just lucky
with the unit I have ??? ...and it's not for sale !

In our enterprise environment, these routers worked great up until a month or so ago, even with our Apple devices. I personally suspect that its a combination of what I said above and a hardware tolerance issue due to excessive heat with the routers. These guys run quite warm, averaging 74 degrees C and powered on for 48 hours, with no fans to help disperse the warm air out of the units, relying entirely on convection to disperse the heat.

When we put USB-power fans under them, the heat dropped to an average of 55 degrees C, with no connectivity issues or 5 or 2.4 GHz signal dropouts.

For comparison, the Netgear R7500's and R800's average temperature is 58 degrees C after being on and under heavy use for 48+ hours, and there is no fan in those units either.

The temperature in our buildings is set to 73 degrees. The R8000 has three radios and the 87R has two. That still doesn't explain the difference in internal temps.

 

[RMerlin]

In our enterprise environment, these routers worked great up until a month or so ago, even with our Apple devices. I personally suspect that its a combination of what I said above and a hardware tolerance issue due to excessive heat with the routers. These guys run quite warm, averaging 74 degrees C and powered on for 48 hours, with no fans to help disperse the warm air out of the units, relying entirely on convection to disperse the heat.

When we put USB-power fans under them, the heat dropped to an average of 55 degrees C, with no connectivity issues or 5 or 2.4 GHz signal dropouts.

For comparison, the Netgear R7500's and R800's average temperature is 58 degrees C after being on and under heavy use for 48+ hours, and there is no fan in those units either.

The temperature in our buildings is set to 73 degrees. The R8000 has three radios and the 87R has two. That still doesn't explain the difference in internal temps.

Are you measuring the QTN temperature (through the RPC API), or the BCM temperature? The AC87U's BCM SOC temperature is lower than with the AC68U so it should be at least as stable (and yes, it's still a bit higher than Netgear's R7000).

Hard to judge the QTN SOC temperature as I have no point of reference - one would have to measure it on an R7500.

Out of curiosity, can you compare the QTN SOC temperature between an AC87U and a R7500? You can retrieve it over telnet/ssh with the following command:

qcsapi_sockrpc get_temperature

My AC87U currently sits at 66C.

 

[RMerlin]

Quantenna is working on a full driver re-write, due out sometime Q1 2015. .

I hope this isn't something they just recently started, because considering the complexity of a wifi driver, that doesn't inspire me with confidence for the near future...

 

[yooper2001]

I guess that is as close to an admission that there is something wrong with this beast as we are gonna get.

 

[Richard1864]

I hope this isn't something they just recently started, because considering the complexity of a wifi driver, that doesn't inspire me with confidence for the near future...

They've been working on it since October. They used our information to make some modifications to it. It's the new driver Gary mentioned, I don't know what was changed tho.

 

[bodean]

They've been working on it since October. They used our information to make some modifications to it. It's the new driver Gary mentioned, I don't know what was changed tho.

Reason it takes them 4+ months to rewrite a driver?

 

[RMerlin]

Reason it takes them 4+ months to rewrite a driver?

Same reason it takes 4+ months for Asus to stabilize new routers: with so many different clients out there, just testing them out and fixing compatibility issues specific to some of them takes a lot of time.

Wifi technology is quite complex. It's not just about turning a radio on, and converting digital IP traffic into an analog signal to push through the chip. You have to manage the output per stream. You have to program and handle the error correction. Have to write the code that will handle the beamforming, varying output power per antenna based on the info you get from your clients. And on top of that, we're now also adding MU-MIMO into the equation.

And after that, you probably need to have the end result tested by RF specialists, possibly go through some certification steps.

Wifi technology code isn't something simple you can write in three weeks of spare time.

 

[DaveMishSr]

Are you measuring the QTN temperature (through the RPC API), or the BCM temperature? The AC87U's BCM SOC temperature is lower than with the AC68U so it should be at least as stable (and yes, it's still a bit higher than Netgear's R7000).

Hard to judge the QTN SOC temperature as I have no point of reference - one would have to measure it on an R7500.

Out of curiosity, can you compare the QTN SOC temperature between an AC87U and a R7500? You can retrieve it over telnet/ssh with the following command:

qcsapi_sockrpc get_temperature

My AC87U currently sits at 66C.

My AC87R is currently at 56° C. This may sound hilarious but I have hit upon a simple solution, empty toilet roll tubes. Yep you read that right, I said toilet paper tubes. I use six and stand them up on end under each corner with two in the middle and voila! I cannot remember what the unit's internal temps were like before but this trick easily knocked 10°C for me. Maybe that is why my 5GHz band is stable with no connectivity drops.

 

[Richard1864]

Are you measuring the QTN temperature (through the RPC API), or the BCM temperature? The AC87U's BCM SOC temperature is lower than with the AC68U so it should be at least as stable (and yes, it's still a bit higher than Netgear's R7000).

Hard to judge the QTN SOC temperature as I have no point of reference - one would have to measure it on an R7500.

Out of curiosity, can you compare the QTN SOC temperature between an AC87U and a R7500? You can retrieve it over telnet/ssh with the following command:

qcsapi_sockrpc get_temperature

My AC87U currently sits at 66C.

Our R7500's sit at 53 degree C QTN SOC temp. The R8000's appear to be happy at 5 degrees C QTN SOC temp. The 87R was 66 degrees C QTN SOC temp. We got them via the RPC API. Sorry for the delay getting back to you.

 

[RMerlin]

Our R7500's sit at 53 degree C QTN SOC temp. The R8000's appear to be happy at 5 degrees C QTN SOC temp. The 87R was 66 degrees C QTN SOC temp. We got them via the RPC API. Sorry for the delay getting back to you.

Thanks for the info.

(BTW the R8000 is based on Broadcom's Xtreme platform, it's not Quantenna-based)

Personally I'm a bit surprised. I remember someone else telling me how he found a pretty weak heatsink solution in quite a few of the Netgear routers he opened, and that was one of the reasons why he avoided Netgear's routers in general. I can only guess that the issue then could be in the thermal interface Asus is using between the SOC and the heatsink itself.

I vaguely remember at least one forum user having improved his Asus router's temperature just by removing the cheap thermal pad, and replacing it with quality thermal compound.

 

[RMerlin]

My AC87R is currently at 56° C. This may sound hilarious but I have hit upon a simple solution, empty toilet roll tubes. Yep you read that right, I said toilet paper tubes. I use six and stand them up on end under each corner with two in the middle and voila! I cannot remember what the unit's internal temps were like before but this trick easily knocked 10°C for me. Maybe that is why my 5GHz band is stable with no connectivity drops.

I remember reading that trick with PS3 owners as well. Personally I keep my PS3 on its side, which has a similar effect (and the model I have is flat enough to stand on its side without having to buy any kind of base). That greatly helped with the fan noise - that thing was getting real loud, real quick while playing a BD-ROM.

Your trick might be worth trying for those users having stability issues. Maybe Asus could have improved the thermals just by having taller rubber feet.

 

[scavenger]

Your trick might be worth trying for those users having stability issues. Maybe Asus could have improved the thermals just by having taller rubber feet.

I've been scratching my head about all the 5GHz woes in this thread, as mine has been relatively solid over several firmware versions (I had Nexus 4 and Touchpad dropouts several weeks ago which looked suspicious).

HOWEVER the very *first* thing I did with this router was to put a USB laptop cooler underneath it. I can't measure temps, as I'm running stock, but maybe that's worth trying for those who are having trouble.

 

[RMerlin]

I've been scratching my head about all the 5GHz woes in this thread, as mine has been relatively solid over several firmware versions (I had Nexus 4 and Touchpad dropouts several weeks ago which looked suspicious).

HOWEVER the very *first* thing I did with this router was to put a USB laptop cooler underneath it. I can't measure temps, as I'm running stock, but maybe that's worth trying for those who are having trouble.

The command I posted earlier would also work with stock (just enable telnet).

 

[pinkfloyd1173]

I've been scratching my head about all the 5GHz woes in this thread, as mine has been relatively solid over several firmware versions (I had Nexus 4 and Touchpad dropouts several weeks ago which looked suspicious).

HOWEVER the very *first* thing I did with this router was to put a USB laptop cooler underneath it. I can't measure temps, as I'm running stock, but maybe that's worth trying for those who are having trouble.

I have been doing that since I got this router and have had no issues like, these guys are having. But at the same time I use this mostly as a router, no adding external HD, or other stuff like that. 2.4&5hz have always been good no drop outs at all???

 

[scavenger]

The command I posted earlier would also work with stock (just enable telnet).

A-ha. Then I'm at 46C.

 

[mromero]

I've used a different trick to get the Asus RT-AC87U router 10 degrees lower than ambient. I mount it on one those wire shelves you see in supermarkets so it's butt naked underneath and the rack wires are a tad warm and I am thinking they act as a passive radiator.

When I have nothing better to do I may open it up and put in some quality thermal grease and drill some ventilation holes on this thing.

That is why I love our Cisco enterprise routers, all metal, easy to pop the hood, lots of space to work around and a quality fan you can easily clean or replace.

The command I posted earlier would also work with stock (just enable telnet).

 

[IAAI]

Update 12/01/2014 - Beta Release - Build 3123

Beta Release - Build 3123

1. This is not on the support site and servers for auto update.
2. Download here - https://drive.google.com/file/d/0Bw2xQ5benV_uZ25FTlFiOFlXVjg/view?usp=sharing
3. Short Fix List
Fixes:
1.TrendMicro Engine updated to fix log issues
2. Numerous bug fixes and performance improvements

Update
1.Expect public release shortly after qualification testing completes.
2.Another public release in early Q1 with new Quantenna driver updates.

any ETA for final build ?

 

[ankhazam]

So far this beta (or rather RTM) build seems far superior to the latest stable so instead of waiting go update to it

 

[neokoenig]

Agreed - it's noticeably better, especially the responsiveness of the Admin GUI. Doesn't fix my unusable 5Ghz channel though.

 

[arrow201]

...HOWEVER the very *first* thing I did with this router was to put a USB laptop cooler underneath it. ...

This is a great idea, I'm not having any probs with my router (so far ),
but why wait ? Went out and got a laptop cooler...worth the $15 I spent
on it. Thanks for the tip !