It really irks me that Linksys advertises "Open source ready with OpenWrt" on their website but the 1900acs performs so bad on OpenWRT I cannot use it for wifi at all. I don't have a device that can even connect. Enabling the the wifi just doesn't work.
Is this normal?
Is seems there should be some sort of legal recourse for Linksys advertising this but not delivering on the promise.
Wifi has been great for me, and has been great for a very long time. I started of with the 1900ac version 1, and at the very beginning there were issues, but those issues were hammered out over time. I've also owned the 1900acs for a long time and it's been very reliable as well. I ended up giving it to my daughter to use while in college when I got the 3200acm which also works great (My daughters had that router for a little over a year, and has never complained of any issues). Now, when I first got the acm it was relatively new, and the wifi drivers didn't work, but after 30 days or so, there were several changes and eventually became stable.
In my opinion, and this is just my opinion, the wifi issues are over-blown. Every week there's an average of 500 downloads for the 1900/3200/1200/32x, and very rarely do anyone say anything negative about the wifi, and that's from a small 3rd party build. The main enchilada LEDE/OpenWrt, and large 3rd party builds like DD-WRT get far more downloads, and I rarely hear about them having issues. Granted, I don't monitor the issues they get.
I attribute what we are seeing more to the liking of the sheer amount of these units being sold every day. Of that large number we are probably getting around < 3% negative feedback of the wifi drivers. That is a unscientific number not based on facts, but just an idea of what I think is happening based on all the downloads these builds get every single day.
Over-all these wifi drivers have been really-really good for most people, but wifi being what it is not everyone is going to have the same experience. This is usually due to the plethora of different devices, and home/neighborhood environment which is the single biggest factor to wifi complaints. Well, that and some unrealistic expectations of how wifi should perform.
But don't take my word for it... Just look at all the work that's been done with mwlwifi over the years. You can read more about it on this link -> https://github.com/kaloz/mwlwifi/commits/master
Both the drivers and firmware are actively being worked on and improved on a regular basis. In fact MU-MIMO is currently being worked on, and we will have an open-source driver with MU-MIMO enabled which I think will be only the 2nd opensource driver with this capability (Might be the very 1st).
Well guys I'm seeing a lot of people complain about terrible wifi in that link given. How do I make sense of your saying everything is great?
I install Openwrt onto any other router and wifi works. I install it onto 1900acs and it just doesn't work in any way. What explains this? Why does Openwrt wifi work onto 3 different Netgear models without issues. Why does it work on 2 different TP-Link models without any issues?
I assume you just installed Openwrt and wifi worked? What explains my experience. I have quantity 2 of the router also so it's not that I got a bad hardware version or somehting.
The WRT1900ACS utilizes the same Marvell 88W8864 WiFi hardware and driver as the WRT1200AC and WRT1900AC (V1 & V2), and the only issue I've ever had with my 1900ACS is Windows rarely registers the SSID without manually typing it in and connecting, resulting with SSID 1, SSID 2, SSID 3, etc. as the SSID name.
Many quality issues on wifi result from using Auto, instead of a specific channel, utilizing channels that don't work well in the user's country/region, etc.
Please post output of:
cat /etc/config/wireless (code box please), removing the SSID name and password.
Channel 52 is a DFS channel. Generally you'll get more consistent results with a non-DFS channel (such as 36 or 48), unless you have lots of neighbors clogging up the band. Channel selection should of course always start with a survey to identify which channels are already in use and avoid them as much as possible.
I have tried all channels over the last couple years. I just picked a channel for purposes of posting here. And YES, the higher non DFS channels are all saturated with companies like Comcast who set up there automatically as part of their Wifi for everyone plan whereever they are (if they subscribers).
Why is OpenWrt being configured sub-optimally for the device which I have specifically installed for? I do not have to manually "optimally configure" for any other router I install for. At least that is, the default install is configured well enough that I can actually get a wifi signal and connect to the internet.
I am not in legacy mode. Does my config show me as in legacy mode? The OpenWrt setting page shows me as being in "N mode"
Well this seems to indicate that the LuCI config is not actually changing the setting then (The LuCI settings clearly indicate N has been selected as the mode). Perhaps this is a bug in the firmware?
WiFi is not enabled by default in my Netgear and TP-Link routers either. Yet when I enable a radio, I get a signal and my device can connect to that radio. this doesn't work though with the 1900ACS models. Why is it not considered a serious problem that the 1900ACS behaves in such a broken manner?
Since you're in the US, you're going to get better performance on channels in the 150s. The 5GHz wavelengths degrade rapidly compared to the 2.4GHz wavelengths, so simply because there's others using 5GHz channels in the 150s, it's not as likely to affect you.
legacy_rates '1' is enabled on radio0, and while @mk24 makes a great point regarding that option, without knowing precisely what that option is doing, it should not be there (my hunch is it was added when you selected N, as I've never seen that option on any of the WRT AC Series routers I've had over the past 3 years).
On a 5GHz router circa >2013, N would be considered legacy, as in order to garnish an extra 150Mbit/s throughput for N clients, it's arresting AC clients' throughput by 416.7Mbit/s.
N maxes out at 450Mbit/s and a 2.4GHz radio will guarantee at least 300Mbit/s for N clients
It may support the full 450Mbit/s, however I don't remember since I haven't had N only clients in ~5yrs
AC maxes out at 866.7Mbit/s, so when a 5GHz radio is configured for N only, the end result is a gain of, at most, 150Mbit/s in throughput on N clients, whereas AC clients have their throughput almost halved.
This is neither efficient, nor optimal, so unless an environment only has 5GHz N clients, and no AC clients, the 5GHz radio should be configured for AC only, with N clients utilizing the 2.4Ghz radio.