I haven't noticed any negative effects in fact the wireless has been solid but I haven't used the WRT3200 in production much since I'm still working on McDebian firmware updates.

@nitroshift

I know the powertables are hardcoded.
I thought I tried to say the same in my message.

I hope my unit will have ETSI european powertables.
I can change the US power adapter if need be, but not the powertables.

(Last edited by adri on 30 Jan 2017, 22:24)

@nitroshift do you know if the country code is set in the firmware bin file? If so it's might be possible to change that.

Loaded LEDE and fired it up for the first time... Set up wifi setting and saved.  5Ghz wifi isn't starting.  Don't know if that's a good place to start

31 00:31:34 2017 daemon.notice netifd: radio0 (10582): command failed: Not supported (-95)
Tue Jan 31 00:31:34 2017 daemon.notice netifd: radio0 (10582): sh: out of range
Tue Jan 31 00:31:34 2017 daemon.notice netifd: radio0 (10582): Configuration file: /var/run/hostapd-phy0.conf
Tue Jan 31 00:31:34 2017 kern.err kernel: [  553.380113] ieee80211 phy0: adapter does not exist
Tue Jan 31 00:31:34 2017 kern.err kernel: [  553.384979] ieee80211 phy0: adapter does not exist
Tue Jan 31 00:31:34 2017 kern.err kernel: [  553.389795] ieee80211 phy0: adapter does not exist
Tue Jan 31 00:31:34 2017 kern.err kernel: [  553.394634] ieee80211 phy0: failed execution
Tue Jan 31 00:31:34 2017 kern.err kernel: [  553.398929] ieee80211 phy0: adapter does not exist
Tue Jan 31 00:31:34 2017 daemon.notice netifd: radio0 (10582): Could not set interface wlan0 flags (UP): I/O error
Tue Jan 31 00:31:34 2017 daemon.notice netifd: radio0 (10582): nl80211: Could not set interface 'wlan0' UP
Tue Jan 31 00:31:34 2017 daemon.notice netifd: radio0 (10582): nl80211: deinit ifname=wlan0 disabled_11b_rates=0
Tue Jan 31 00:31:34 2017 daemon.notice netifd: radio0 (10582): nl80211 driver initialization failed.
Tue Jan 31 00:31:34 2017 daemon.notice netifd: radio0 (10582): wlan0: interface state UNINITIALIZED->DISABLED
Tue Jan 31 00:31:34 2017 daemon.notice netifd: radio0 (10582): wlan0: AP-DISABLED 
Tue Jan 31 00:31:34 2017 daemon.notice netifd: radio0 (10582): wlan0: CTRL-EVENT-TERMINATING 
Tue Jan 31 00:31:34 2017 daemon.notice netifd: radio0 (10582): hostapd_free_hapd_data: Interface wlan0 wasn't started
Tue Jan 31 00:31:34 2017 daemon.notice netifd: radio0 (10582): Device setup failed: HOSTAPD_START_FAILED
Tue Jan 31 00:31:35 2017 daemon.err insmod: module is already loaded - xt_multiport
Tue Jan 31 00:31:35 2017 daemon.err insmod: module is already loaded - xt_comment
Tue Jan 31 00:31:35 2017 daemon.err insmod: module is already loaded - xt_length

.

(Last edited by davidc502 on 31 Jan 2017, 01:35)

Somehow int wlan0 was renamed to radio0.network1.

Strange

Added 'default_radio0 ' back to the wifi-iface and wifi came back up. Unknown reason why it was dropped from the config.

config wifi-iface 'default_radio0'
        option device 'radio0'
        option mode 'ap'
        option ssid 'TDS_5Ghz'
        option encryption 'none'

@Chadster766

Work is being done to use the official wifi firmware blob.

nitroshift

So I received my 3200ACM today.  For baseline testing I swapped out my ACS and configured the rango unit in its place.  Initial testing is pretty positive.

Other than the nuisance wifi leds not working (which I've seen reported elsewhere and haven't looked into at all) the only issue I've noticed is with the 5G wireless performance.

As of now the Shelby unit is rocking (this is the 1900ACS, for comparison) -- I can load multiple high bandwidth streams and they hold around 25 MB/sec each.  I started a file transfer on the rango unit and was getting better than 30 MB/sec, but when I started a second transfer (iperf over openvpn to test throughput) the file transfer bottomed out at around 3 MB/sec.  Also, I usually keep a VNC tunnel going to a remote machine and the screen quality reduced to poor and stayed that way.  I tried a few things but only rebooting improved both transfer speed and VNC video quality.

So a promising start but definitely room for improvement on the 5G.

I'm digging through the factory firmware image, and i can identify several of the power tables.
I may try messing with one of the images to do what i want it to.

This doesn't sound like complying with the firmware license, though not complying with local law due to FCC madness is a no go as well. Currently the driver checks whether to update power levels or not, so I assume the firmware wouldn't check again and you can possibly get away with just modifying mwlwifi. Anyway, hope it will be an ETSI unit for me ...

In response to the LED's not working, it appeared to be a good idea to manually configure them in the LED section. So, 3g and 5g LED configurations were added, and appeared to work initially, but within a few minutes, all the wifi clients dropped, and could not reconnect. Until I removed the LED configurations, it remained that way. I did this twice and both times had the same result.

As for the 5g driver - it seems stable and speeds are good. I have a couple of iphone 7's that can supposedly take advantage of tri-band, but in preliminary tests, they wouldn't take advantage though the router was set to 160mhz. Using a Amazon fireTV box connected via 5g wifi, youtube was selected with automatic video playing, and when I woke up in the morning, video's were still playing.

So, besides the LED's, and 160mhz (tri-band), it's working well.

@davidc502

LED patches are still laying around on my work machine, pending tests when i receive the giveaway unit.

nitroshift

Has anyone seen this yet?  Where 5Ghz is still being broadcast, but no one can join?

System and Kernel log show the same issue over and over.

Wed Feb  1 00:35:50 2017 kern.err kernel: [  365.770004] ieee80211 phy0: cmd 0x801d=MEMAddrAccess timed out
Wed Feb  1 00:35:50 2017 kern.err kernel: [  365.775864] ieee80211 phy0: return code: 0x001d
Wed Feb  1 00:35:50 2017 kern.err kernel: [  365.780421] ieee80211 phy0: timeout: 0x001d
Wed Feb  1 00:35:50 2017 kern.err kernel: [  365.784619] ieee80211 phy0: failed execution

Edit...

Adding the following as well.  The GUI during this time is very unresponsive, and the statistics are showing blank.

[  712.326933] wlan0: failed to set key (2, ff:ff:ff:ff:ff:ff) to hardware (-5)

(Last edited by davidc502 on 1 Feb 2017, 01:48)

Ops! I see I'm a little late to the party.   https://github.com/kaloz/mwlwifi/issues/118

Thinking about this issue... I wonder if it is something else other than the driver.  The reason why I say that is because during this time, logging into Luci, many things were not working.

@yuhhaurlin, has confimed that in the 118 issue, also with the updated DTS setting the wifi LEDs is know to crash the wifi as you discovered.

Processor % Utilized is much higher than on the acs which averaged 2-4% under normal load.  The 3200acm is averaging around 10% under normal usage.  Considering the acm has a faster processor, if anything one would expect the number to go down, and not up.

EDIT

Looks like after 10 to 15 minutes the processor has settled down...Will continue to monitor.

(Last edited by davidc502 on 1 Feb 2017, 04:37)

Hi people what builds are you lot testing? davidc502 or the new RC? I just got mine in the mail and don't know what to flash first (smirk)

Just a note that I'm rethinking how best to integrate the Rango device into my network for testing.  Initially I was going to replace my Shelby with it but that's turning out to be a bad idea (it's currently in that spot up and running, with some downsides).  I've seen data rates slow to a crawl (< 1 MB/s) and a wifi crash requiring a reboot when multiple simultaneous demands are placed on the 5G network.  Typical use I haven't seen any obvious issues.

I'm going to move it to a lower demand isolated network for more structured testing and as the driver improves I'd like to see it back in the "prime" spot.

I did verify that WDS is working on this unit as well so it can be used easily as a wifi bridge.  I haven't worked with it or tested that with other units so no idea how it impacts performance.

@tapper I built from LEDE trunk (and updated mwlwifi to the latest commit) but you might want to start with the RC.  Trunk has updated busybox to 1.26.2 so you'll need to take care in reconstructing your .config (unless you already have that done).

@InkblotAdmirer

May I ask what "latest mwlwifi" you are talking about? As my patch to update to the latest version was already upstreamed.

Thanks!

Hi i tryed out the RC but it will not let me install any packages. I was going to flash over to gargoyle but i just get The uploaded image file does not contain a supported format. Make sure that you choose the generic image format for your platform. What can i do about this? The Gargoyle file is a .tar file

You will probably have to force using sysupgrade -F -n image-name from ssh. But if this was your fist flash, don't forget you could go back to OEM on the other partition.

@lifehacksback
Sorry for any confusion, trunk currently has the latest commit it just calls out the wrong date on the version.  So my patch is pulling the same version as trunk, just updating the version to the commit date.

Thanks mate. OEM firmware is so bad! (puke) It can't even be used with a screen reader i had to get my wife to help me flash in the first place. If a blind person got this router they would be scrooed for setting it up.

Understood, I only meant that you could go back to OEM in order to move forward with a gargoyle flash, rather than trying to force a sysupgrade from ssh.

In case someone is interested... the acm will run snort. With the wifi issues on-going, I thought it would be a good time to see if it would... It uses 80% of the RAM, but processor utilization looks good under normal traffic load. So far I'm impressed with its power.

Let me know if someone would like the snort and snort.conf

Thanks,