OpenWrt Forum Archive

Topic: davidc502 1900ac 3200acm builds

The content of this topic has been archived between 26 Feb 2018 and 7 May 2018. Unfortunately there are posts – most likely complete pages – missing.

T-Troll wrote:
starcms wrote:

I'd recommend keeping radio2 disabled.  radio0 is your main 5ghz AP and radio1 is your 2.4ghz AP.

Reasonable, but wrong:

root@TTrollstation:/tmp# lspci
00:01.0 PCI bridge: Marvell Technology Group Ltd. MV78230 [Armada XP] ARM SoC (rev 02)
00:02.0 PCI bridge: Marvell Technology Group Ltd. MV78230 [Armada XP] ARM SoC (rev 02)
00:03.0 PCI bridge: Marvell Technology Group Ltd. MV78230 [Armada XP] ARM SoC (rev 02)
01:00.0 USB controller: Etron Technology, Inc. EJ168 USB 3.0 Host Controller (rev 01)
02:00.0 Ethernet controller: Marvell Technology Group Ltd. 88W8864 [Avastar] 802.11ac Wireless
03:00.0 Ethernet controller: Marvell Technology Group Ltd. 88W8864 [Avastar] 802.11ac Wireless

So wireless is pci 2 and 3. It's 1900AC, BTW, maybe 3200 have different set.

The 1900AC V1 is very different from all others.  What I posted is correct for all other models (minus the radio2, which only applies to the 3200ACM).

Edit: This is from my 1200AC.  All models are the same, except the 1900AC V1 which is completely different (because it uses the Marvell Armada XP MV78230 SoC instead of the Marvell Armada 385 88F6820), and the 3200ACM which can have an entry for radio2 at 03:00.0 for the 88W8887 (and has the 88W8964 instead of the 88W8864 for radio0 and radio1 at 01:00.0 and 02:00.0, respectively).

root@WRT1200AC:~# lspci
00:01.0 PCI bridge: Marvell Technology Group Ltd. Device 6820 (rev 04)
00:02.0 PCI bridge: Marvell Technology Group Ltd. Device 6820 (rev 04)
01:00.0 Ethernet controller: Marvell Technology Group Ltd. 88W8864 [Avastar] 802.11ac Wireless
02:00.0 Ethernet controller: Marvell Technology Group Ltd. 88W8864 [Avastar] 802.11ac Wireless

Edit2: In your original post, you were responding to someone who had a 3200ACM, hence why I thought you also had one, not a 1900AC V1.

(Last edited by starcms on 17 May 2017, 16:49)

I've seen this posted elsewhere, but not here, so for those of you who are sick of dealing with DFS (radar detection) and limited power limits, this will interest you.

First of all, this will only work on models before the FCC mandated that router manufacturers make it impossible to alter these things.  So it will only work on models that do not have their power table built into EEPROM and ignore the power table that is included in firmware.  If you have a 1200AC V2, 1900ACS V2, or 3200ACM, you are out of luck, sorry.

But for all other models (1200AC V1, 1900AC V1 and V2, 1900ACS V1), if you would like to:

a) deactivate DFS on all channels.
b) enable WiFi channels 12 and 13.
c) enable up to 30dBm transmit power on all channels.

it is a very simple procedure.

1) Issue the following commands:

cd /sbin
wget http://harry-home.duckdns.org:2799/reghack2/reghack2
chmod 755 reghack2
cp /lib/modules/*/cfg80211.ko /lib/modules/4.9.20/cfg80211.ko.bak
/sbin/reghack2 /lib/modules/*/cfg80211.ko
reboot

2) Ensure your router is set to either region US or World (00) (If you bought your router in the US, you can skip this).
3) Enjoy

Before:

root@WRT1200AC:~# iw reg get
global
country US: DFS-FCC
        (2402 - 2472 @ 40), (N/A, 30), (N/A)
        (5170 - 5250 @ 80), (N/A, 23), (N/A), AUTO-BW
        (5250 - 5330 @ 80), (N/A, 23), (0 ms), DFS, AUTO-BW
        (5490 - 5730 @ 160), (N/A, 23), (0 ms), DFS
        (5735 - 5835 @ 80), (N/A, 30), (N/A)
        (57240 - 63720 @ 2160), (N/A, 40), (N/A)

After:

root@WRT1200AC:~# iw reg get
global
country US: DFS-UNSET
        (2400 - 2483 @ 40), (N/A, 30), (N/A)
        (5140 - 5860 @ 160), (N/A, 30), (N/A)

You can also run the command "iw phy" to see and verify the individual channels, maximum Tx levels, and DFS (or lack thereof) before and after.

Hope that helps some of you!  It'll need to be reapplied every time you flash an updated build.

Edit: More info here: https://forum.lede-project.org/t/wrt190 … er/1491/10

Edit2: Kernel version, mwlwifi (wifi driver) version, and cfg80211 version don't matter.  Been using it for months on @david's builds.

(Last edited by starcms on 21 May 2017, 11:57)

@starcms

It's too bad Linksys hasn't gone back to address the DFS issue. I've done many tests and all it takes is another wifi signal on the same DFS channel, and it will cower and run away, to a non DFS channel, like a beat dog.

I do know that dfs is still an open issue on kaloz mwlwifi page, so eventually it should be addressed.

Keep in mind, part of the reason why we are in the boat we are in is because of people using power levels and channels not approved for.

davidc502 wrote:

@starcms

It's too bad Linksys hasn't gone back to address the DFS issue. I've done many tests and all it takes is another wifi signal on the same DFS channel, and it will cower and run away, to a non DFS channel, like a beat dog.

I do know that dfs is still an open issue on kaloz mwlwifi page, so eventually it should be addressed.

Keep in mind, part of the reason why we are in the boat we are in is because of people using power levels and channels not approved for.

I don't understand your very last sentence.  We are in what boat because of people using power levels and channels not approved?  The FCC requiring that all routers be locked down by the manufacturer?

Anyway, I was mainly offering it because of the issues with DFS in the mwlwifi driver. 

Also, Wifi Channels 12 and 13 have been officially activated in the official Linksys firmware and are available for legal use by the FCC.  I'm unsure why they are still unavailable in LEDE without the "hack" I posted.  I don't believe it's an issue related to the mwlwifi driver, but to cfg80211 based on how the hack works, but I could be mistaken.

Therefore, the only other use for the hack besides disabling malfunctioning DFS and enabling Wifi Channels 12 and 13 (which should be enabled anyway, at least in the US), is increasing the Tx power level from 23dBm to 30dBm on some 5GHz channels.  I've never found a good reason to increase above the standard 23dBm because it doesn't help reception or performance (just because the router can transmit farther/with more power, a user's devices (STAs) cannot, so it really has no helpful benefit).

(Last edited by starcms on 17 May 2017, 20:23)

starcms wrote:
davidc502 wrote:

@starcms

It's too bad Linksys hasn't gone back to address the DFS issue. I've done many tests and all it takes is another wifi signal on the same DFS channel, and it will cower and run away, to a non DFS channel, like a beat dog.

I do know that dfs is still an open issue on kaloz mwlwifi page, so eventually it should be addressed.

Keep in mind, part of the reason why we are in the boat we are in is because of people using power levels and channels not approved for.

  The FCC requiring that all routers be locked down by the manufacturer?

Correct -- The FCC has clamped down on router manufacturers due to this type of activity.

davidc502 wrote:

@starcms

It's too bad Linksys hasn't gone back to address the DFS issue. I've done many tests and all it takes is another wifi signal on the same DFS channel, and it will cower and run away, to a non DFS channel, like a beat dog.

I do know that dfs is still an open issue on kaloz mwlwifi page, so eventually it should be addressed.

No luck on the DFS issue ever being resolved properly apprently...

https://github.com/kaloz/mwlwifi/issues/75

At least those of us not on a 1200AC V2 or 1900ACS V2 can benefit from the "hack" I posted, since DFS is apparently working properly or at least better on the 3200ACM.

(Last edited by starcms on 18 May 2017, 03:19)

I saw he closed the DFS thread... sad

This is what I get on my 3200ACM:

root@Scruffy:~# iw reg get
global
country 98: DFS-FCC
    (2402 - 2472 @ 40), (N/A, 30), (N/A)
    (5170 - 5250 @ 80), (N/A, 23), (N/A), AUTO-BW
    (5250 - 5330 @ 80), (N/A, 23), (0 ms), DFS, AUTO-BW
    (5490 - 5600 @ 80), (N/A, 23), (0 ms), DFS
    (5650 - 5730 @ 80), (N/A, 23), (0 ms), DFS
    (5735 - 5835 @ 80), (N/A, 30), (N/A)

phy#2
country US: DFS-FCC
    (2402 - 2472 @ 40), (N/A, 30), (N/A)
    (5170 - 5250 @ 80), (N/A, 23), (N/A), AUTO-BW
    (5250 - 5330 @ 80), (N/A, 23), (0 ms), DFS, AUTO-BW
    (5490 - 5730 @ 160), (N/A, 23), (0 ms), DFS
    (5735 - 5835 @ 80), (N/A, 30), (N/A)
    (57240 - 63720 @ 2160), (N/A, 40), (N/A)

phy#1
country CA: DFS-FCC
    (2402 - 2472 @ 40), (N/A, 30), (N/A)
    (5150 - 5250 @ 80), (N/A, 23), (N/A), AUTO-BW
    (5250 - 5350 @ 80), (N/A, 24), (0 ms), DFS, AUTO-BW
    (5470 - 5600 @ 80), (N/A, 24), (0 ms), DFS
    (5650 - 5730 @ 80), (N/A, 24), (0 ms), DFS
    (5735 - 5835 @ 80), (N/A, 30), (N/A)

phy#0
country CA: DFS-FCC
    (2402 - 2472 @ 40), (N/A, 30), (N/A)
    (5150 - 5250 @ 80), (N/A, 23), (N/A), AUTO-BW
    (5250 - 5350 @ 80), (N/A, 24), (0 ms), DFS, AUTO-BW
    (5470 - 5600 @ 80), (N/A, 24), (0 ms), DFS
    (5650 - 5730 @ 80), (N/A, 24), (0 ms), DFS
    (5735 - 5835 @ 80), (N/A, 30), (N/A)

Seems no one has any issues going to any of the DFS channels. the Issue is that it spurious transmissions or other wifi being used on the same channel will cause it to go back to a non-DFS channel.

farchord wrote:

This is what I get on my 3200ACM:

root@Scruffy:~# iw reg get
global
country 98: DFS-FCC
    (2402 - 2472 @ 40), (N/A, 30), (N/A)
    (5170 - 5250 @ 80), (N/A, 23), (N/A), AUTO-BW
    (5250 - 5330 @ 80), (N/A, 23), (0 ms), DFS, AUTO-BW
    (5490 - 5600 @ 80), (N/A, 23), (0 ms), DFS
    (5650 - 5730 @ 80), (N/A, 23), (0 ms), DFS
    (5735 - 5835 @ 80), (N/A, 30), (N/A)

phy#2
country US: DFS-FCC
    (2402 - 2472 @ 40), (N/A, 30), (N/A)
    (5170 - 5250 @ 80), (N/A, 23), (N/A), AUTO-BW
    (5250 - 5330 @ 80), (N/A, 23), (0 ms), DFS, AUTO-BW
    (5490 - 5730 @ 160), (N/A, 23), (0 ms), DFS
    (5735 - 5835 @ 80), (N/A, 30), (N/A)
    (57240 - 63720 @ 2160), (N/A, 40), (N/A)

phy#1
country CA: DFS-FCC
    (2402 - 2472 @ 40), (N/A, 30), (N/A)
    (5150 - 5250 @ 80), (N/A, 23), (N/A), AUTO-BW
    (5250 - 5350 @ 80), (N/A, 24), (0 ms), DFS, AUTO-BW
    (5470 - 5600 @ 80), (N/A, 24), (0 ms), DFS
    (5650 - 5730 @ 80), (N/A, 24), (0 ms), DFS
    (5735 - 5835 @ 80), (N/A, 30), (N/A)

phy#0
country CA: DFS-FCC
    (2402 - 2472 @ 40), (N/A, 30), (N/A)
    (5150 - 5250 @ 80), (N/A, 23), (N/A), AUTO-BW
    (5250 - 5350 @ 80), (N/A, 24), (0 ms), DFS, AUTO-BW
    (5470 - 5600 @ 80), (N/A, 24), (0 ms), DFS
    (5650 - 5730 @ 80), (N/A, 24), (0 ms), DFS
    (5735 - 5835 @ 80), (N/A, 30), (N/A)

No way to alter it on a 3200ACM since the FCC mandated that manufacturers not allow any changes by the user at all.  At least we are lucky in a way.  Some manufacturers of routers have enforced the FCC law by not allowing custom firmwares (LEDE, openwrt, ddwrt, etc) to be installed at all.  At least Linksys simply moved the power tables to an eeprom chip and kept the firmwares modifiable.

What I find odd and interesting is that apparently Linksys got FCC approval for 802.11ad in the US on these routers.  Notice the entry for  (57240 - 63720 @ 2160), (N/A, 40), (N/A).  Even though the wifi radios (whether it be 88W8864, 88W8964, or 88W8887) don't support 802.11ad (60GHz).  The entry even existed on my 1200AC before I applied the mod as you can see in my post above.

Edit:  Just noticed that the "global" settings are set to country 98 (which I don't believe is Canada, because the settings are different than radio0 and radio1 which are set to Canada), radio0 and radio1 are set to Canada, and radio2 is set to USA.  I'm assuming you bought your router in Canada.  I wonder why radio2 is set for US (maybe because its for DFS; but other countries use DFS so doesn't make sense).  Also, I wonder why the global settings are set to country code 98 (not sure to which country that applies). 

Have you modified these settings/your region or this is how they originally were?

Edit2: My only guess would be that the "global" settings (the one with country code 98) are being set by firmware and can be "changed", but this wouldn't result in any different behavior.  Because they are ignored by phy0, phy1, and phy2 which are pulling their individual settings from EEPROM.  If you notice the results on my 1200AC V1 (which doesn't have settings built into eeprom), the iw reg get command gives no entries for phy0 or phy1 -- only the global settings.  Still doesn't explain why phy2 is set to US instead of CA.

Edit3: Aha! I figured it out.  The global settings (country code 98 for whatever reason), are the overlapping settings of phy0 and ph1 which are set to Canada and phy2 which are set to USA.  The global settings only include those frequencies and settings which are allowed in both CA and US.  Why it's country code 98 I have no idea.  And why phy2 is set to US instead of CA is the real question.

Edit4: Found this:

The value 98 represents a synthesized regulatory domain, based on the intersection of the available source of regulatory information (which can include the EEPROM, the userland setting, and a country IE from your AP).

(Last edited by starcms on 19 May 2017, 02:38)

davidc502 wrote:

Seems no one has any issues going to any of the DFS channels. the Issue is that it spurious transmissions or other wifi being used on the same channel will cause it to go back to a non-DFS channel.

Yep, no issues selecting any of the DFS channels.  But before I used to apply the "hack", it would jump off any DFS channel even if there were no other WiFi AP in the area on the same channel, and I live over 75 miles from any airport.  It would just "detect" radar when there was absolutely nothing on that channel except itself and change to channel 36.

(Last edited by starcms on 19 May 2017, 02:32)

starcms wrote:
davidc502 wrote:

Seems no one has any issues going to any of the DFS channels. the Issue is that it spurious transmissions or other wifi being used on the same channel will cause it to go back to a non-DFS channel.

Yep, no issues selecting any of the DFS channels.  But before I used to apply the "hack", it would jump off any DFS channel even if there were no other WiFi AP in the area on the same channel, and I live over 75 miles from any airport.  It would just "detect" radar when there was absolutely nothing on that channel except itself and change to channel 36.

Still I believe if the router were to be put into a faraday cage, it would stay on the DFS channel. There's noise it's picking up, and changing channel because of it. 

My neighbor has direct TV, and the wifi SSID is called "directTV". For fun, I run it off of the DFS channels, as all I have to do is run a speed test on that same channel, and it then it immediately moves to another DFS or Non DFS channel. I don't know what model router directTV uses, but apparently it is broken in the same way. Prime example of a wifi router not sensing radar properly.

davidc502 wrote:
starcms wrote:
davidc502 wrote:

Seems no one has any issues going to any of the DFS channels. the Issue is that it spurious transmissions or other wifi being used on the same channel will cause it to go back to a non-DFS channel.

Yep, no issues selecting any of the DFS channels.  But before I used to apply the "hack", it would jump off any DFS channel even if there were no other WiFi AP in the area on the same channel, and I live over 75 miles from any airport.  It would just "detect" radar when there was absolutely nothing on that channel except itself and change to channel 36.

Still I believe if the router were to be put into a faraday cage, it would stay on the DFS channel. There's noise it's picking up, and changing channel because of it. 

My neighbor has direct TV, and the wifi SSID is called "directTV". For fun, I run it off of the DFS channels, as all I have to do is run a speed test on that same channel, and it then it immediately moves to another DFS or Non DFS channel. I don't know what model router directTV uses, but apparently it is broken in the same way. Prime example of a wifi router not sensing radar properly.

Completely off topic, but DirecTV doesn't use or provide routers. He most likely just set the SSID of his router to DirecTV and has his DirecTV receiver connected via that SSID wink

But basically what you are saying is you can get his router to change channels by setting yours to the same DFS channel as his?

(Last edited by starcms on 19 May 2017, 05:46)

Hey David, just a headsup.

Seems that as google now priorizes SSL websites, if you search for davidc502 on google, I get your website, but it's SSL version. But because the cert on your website ain't right (Self-signed?) it gives a big red error.

Not a huge deal, I know it's not a bad site; but for newbies, it might scare them off.

farchord wrote:

Hey David, just a headsup.

Seems that as google now priorizes SSL websites, if you search for davidc502 on google, I get your website, but it's SSL version. But because the cert on your website ain't right (Self-signed?) it gives a big red error.

Not a huge deal, I know it's not a bad site; but for newbies, it might scare them off.

Not sure what you are seeing as when I google, there's no error. The certificate is signed by LetsEncrypt, and is valid in all browsers that I know of.

starcms wrote:
davidc502 wrote:
starcms wrote:

Yep, no issues selecting any of the DFS channels.  But before I used to apply the "hack", it would jump off any DFS channel even if there were no other WiFi AP in the area on the same channel, and I live over 75 miles from any airport.  It would just "detect" radar when there was absolutely nothing on that channel except itself and change to channel 36.

Still I believe if the router were to be put into a faraday cage, it would stay on the DFS channel. There's noise it's picking up, and changing channel because of it. 

My neighbor has direct TV, and the wifi SSID is called "directTV". For fun, I run it off of the DFS channels, as all I have to do is run a speed test on that same channel, and it then it immediately moves to another DFS or Non DFS channel. I don't know what model router directTV uses, but apparently it is broken in the same way. Prime example of a wifi router not sensing radar properly.

Completely off topic, but DirecTV doesn't use or provide routers. He most likely just set the SSID of his router to DirecTV and has his DirecTV receiver connected via that SSID wink

But basically what you are saying is you can get his router to change channels by setting yours to the same DFS channel as his?

Being that directTV is now owned by at&t, it's probably a at&t router connecting to a direcTV Genie on 5Ghz using that SSID. Still, just by using a Wifi Analyzer, I can see what's on all the channels including the widths of each. And yeah, I can simply change to the same channel, and all it takes is a little traffic, and it will bounce. Observing over time, wifi is always looking for the least populated channel, so if I vacate a DFS channel, after a while it will go back over to it, and sometimes I see it on other DFS channels. So, there's some kind of logic behind how it works with DFS, and though easily run off a channel, it will go to other DFS or nonDFS channels at will.

davidc502 wrote:
starcms wrote:
davidc502 wrote:

Still I believe if the router were to be put into a faraday cage, it would stay on the DFS channel. There's noise it's picking up, and changing channel because of it. 

My neighbor has direct TV, and the wifi SSID is called "directTV". For fun, I run it off of the DFS channels, as all I have to do is run a speed test on that same channel, and it then it immediately moves to another DFS or Non DFS channel. I don't know what model router directTV uses, but apparently it is broken in the same way. Prime example of a wifi router not sensing radar properly.

Completely off topic, but DirecTV doesn't use or provide routers. He most likely just set the SSID of his router to DirecTV and has his DirecTV receiver connected via that SSID wink

But basically what you are saying is you can get his router to change channels by setting yours to the same DFS channel as his?

Being that directTV is now owned by at&t, it's probably a at&t router connecting to a direcTV Genie on 5Ghz using that SSID. Still, just by using a Wifi Analyzer, I can see what's on all the channels including the widths of each. And yeah, I can simply change to the same channel, and all it takes is a little traffic, and it will bounce. Observing over time, wifi is always looking for the least populated channel, so if I vacate a DFS channel, after a while it will go back over to it, and sometimes I see it on other DFS channels. So, there's some kind of logic behind how it works with DFS, and though easily run off a channel, it will go to other DFS or nonDFS channels at will.

But if DFS is working properly, it isn't supposed to just cower off a channel just because another router is using the same channel or stray interference interfering.  There is a special, very unique patter or patterns of signals or interference it is meant to detect to trigger a radar warning and force a channel switch.  Obviously some or most routers at this point aren't implementing DFS properly and switch if there is anything on that channel as you said.  I believe the 3200ACM is one of the first to actually try to implement DFS properly and hence the extra 1x1:1 5GHz 88W8887 radio it has.

About DirecTV and AT&T, good point, they could have AT&T UVerse Internet and an AT&T modem/router combo.  But since eventually DirecTV will be called AT&T, and not the other way around, it's highly unlikely the default SSID would be DirecTV; if anything it would be AT&T.  And I know from experience, DirecTV installers simply connect the Genie receiver to Cat5 if it is available, or to WiFi if it is not.  They don't set the SSID to DirecTV or mess with the router at all (most are too dumb and it's not part of their training, nor something they are even supposed to do when installing).  So most likely your neighbor changed his router (whether it's an AT&T one or not) to have a special SSID just for his DirecTV Genie.  I did a similar thing creating a VLAN with LAN port 4 just for my DirecTV Genie (which then communicates with the other receivers over coax).  Main reason being DirecTV receivers are constantly pinging (every second) to find other receivers that may be connected and to keep in contact with the receivers which are connected and I didn't want all that spam going over the rest of my LAN.

Edit: The DFS behavior your referred to in the second half of your message, I'm guessing that is behavior you have observed on a 3200ACM?  Not on a 1200/1900AC/ACS?  And is that if your have the channel set to Auto?

(Last edited by starcms on 19 May 2017, 15:12)

farchord wrote:

Hey David, just a headsup.

Seems that as google now priorizes SSL websites, if you search for davidc502 on google, I get your website, but it's SSL version. But because the cert on your website ain't right (Self-signed?) it gives a big red error.

Not a huge deal, I know it's not a bad site; but for newbies, it might scare them off.

@david's website has a valid, non-self signed certificate.  It shouldn't give any error at all, and doesn't on any of my devices using multiple browsers.

(Last edited by starcms on 19 May 2017, 14:54)

starcms wrote:

Edit: The DFS behavior your referred to in the second half of your message, I'm guessing that is behavior you have observed on a 3200ACM?  Not on a 1200/1900AC/ACS?  And is that if your have the channel set to Auto?

Have not tested the 3200acm in the same way that I've had fun with that directTV wifi SSID.. lol

I do know the 3200acm is better behaved on DFS, but not perfect as there have been some days where it has switched off of DFS. Maybe it was hit by radar? I wouldn't think so, but don't have proof to the contrary.

*EDIT*

In my area, the noise level on DFS is just as bad as non DFS, so at least in my case, there's no speed increase by using it.  I say speed because as noise increases, speed drops off accordingly as retransmission's etc.. increase.

(Last edited by davidc502 on 19 May 2017, 20:55)

davidc502 wrote:

In my area, the noise level on DFS is just as bad as non DFS, so at least in my case, there's no speed increase by using it.  I say speed because as noise increases, speed drops off accordingly as retransmission's etc.. increase.

The noise levels of DFS channels in my area are also a little higher than non-DFS channels.
The main reason I;d like a functional DFS is to avoid using the already crowded channels 36-48.
In europe we are not allowed to use channels above 136, so without DFS the router is stuck to the lower 4 channels. sad

davidc502 wrote:
starcms wrote:

Edit: The DFS behavior your referred to in the second half of your message, I'm guessing that is behavior you have observed on a 3200ACM?  Not on a 1200/1900AC/ACS?  And is that if your have the channel set to Auto?

Have not tested the 3200acm in the same way that I've had fun with that directTV wifi SSID.. lol

I do know the 3200acm is better behaved on DFS, but not perfect as there have been some days where it has switched off of DFS. Maybe it was hit by radar? I wouldn't think so, but don't have proof to the contrary.

*EDIT*

In my area, the noise level on DFS is just as bad as non DFS, so at least in my case, there's no speed increase by using it.  I say speed because as noise increases, speed drops off accordingly as retransmission's etc.. increase.

lol, you have too much fun experimenting with other people's stuff big_smile

Was your 3200ACM hit by radar?  Highly unlikely unless you live within 10-15 miles of an airport or a university doing some FCC authorized studies. It was obviously hit by something it interpreted as radar; but if working 100% properly, DFS should only respond to a very unique sequence of interference, not just any interference.  I wouldn't be surprised if Linksys releases updated firmware in the future that may tweak its DFS detection.  Doesn't seem like it should be difficult on the 3200ACM since it has that extra 88W8887.  All it has to do is listen to the same channel that the 88W8964 is using, and see if it can identify those specific radar patterns/signatures/sequences.

(Last edited by starcms on 20 May 2017, 00:16)

starcms wrote:
davidc502 wrote:
starcms wrote:

Edit: The DFS behavior your referred to in the second half of your message, I'm guessing that is behavior you have observed on a 3200ACM?  Not on a 1200/1900AC/ACS?  And is that if your have the channel set to Auto?

Have not tested the 3200acm in the same way that I've had fun with that directTV wifi SSID.. lol

I do know the 3200acm is better behaved on DFS, but not perfect as there have been some days where it has switched off of DFS. Maybe it was hit by radar? I wouldn't think so, but don't have proof to the contrary.

*EDIT*

In my area, the noise level on DFS is just as bad as non DFS, so at least in my case, there's no speed increase by using it.  I say speed because as noise increases, speed drops off accordingly as retransmission's etc.. increase.

lol, you have too much fun experimenting with other people's stuff big_smile

Was your 3200ACM hit by radar?  Highly unlikely unless you live within 10-15 miles of an airport or a university doing some FCC authorized studies. It was obviously hit by something it interpreted as radar; but if working 100% properly, DFS should only respond to a very unique sequence of interference, not just any interference.  I wouldn't be surprised if Linksys releases updated firmware in the future that may tweak its DFS detection.  Doesn't seem like it should be difficult on the 3200ACM since it has that extra 88W8887.  All it has to do is listen to the same channel that the 88W8964 is using, and see if it can identify those specific radar patterns/signatures/sequences.

See, that leads me to another question though. On the stock firmware (as far I can understand), there's 3 radios on the 3200ACM, a 5Ghz, a 2.4Ghz and another 5ghz on a different chip.

According to the mwlwifi github, they say that that third radio is basically used to supplement the other two in some way (I don't understand how). Is there a special config this needs on LEDE?

Because right now on my setup the 3rd radio has no interface, I removed it because I thought that wasn't how it was supposed to be used.

farchord wrote:

According to the mwlwifi github, they say that that third radio is basically used to supplement the other two in some way (I don't understand how). Is there a special config this needs on LEDE?

Because right now on my setup the 3rd radio has no interface, I removed it because I thought that wasn't how it was supposed to be used.

Currently, I am testing having disconnected the 3rd radio chip. Will report the results.

There is a new build uploaded to the site r4164.  For those on the 3200acm, there was a wifi commit a few days back that changes "Corrected the rx buffer size". There were also commits for "ClearFog".

I've only been up, on the new image, for around 15 minutes, and the wifi is absolutely great so far. I'm interested in hearing from others on the 3200acm that are testing the new wifi driver.

Thanks,

There's a mad-house going on in the DFS issue thread on github for mwlwifi.  Here's my opinion.  Yes, the stock firmware doesn't allow DFS channels on the 1200/1900ac/acs.  But LEDE does, and DFS is functional in LEDE with the current version of mwlwifi, just extremely too sensitive.  So why not simply fix it?  There's no law that says mwlwifi combined with LEDE can't be better than the stock linksys driver and firmware.  That's the point, isn't it?  lol

Second issue.  As @farchord mentioned above, the third radio in the 3200ACM (the 88W8887, which is a 1x1:1 ac radio) in LEDE can be disabled or used as an AP.  This is obviously not how it was meant to be used.  It was meant ONLY to detect radar/aid in DFS, and how it is used in the stock Linksys firmware.  Since the 88W8887 isn't being used as a DFS suppliment in LEDE at the current time, the DFS behavior of the 3200ACM should be practically identical to that of the 1200/1900ac/acs (I say practically because apparently the 88W8964 handles DFS better by itself than the 88W8864; but there is no additional DFS code in mwlwifi that only applies to the 88W8964/3200ACM). 

Since the 88W8887 isn't being used as a DFS monitor in LEDE, the 3200ACM is no more "authorized" to use the DFS channels than the 1200/1900ac/acs.  The only reason the 3200ACM was FCC authorized to access DFS channels was because of the 88W8887 monitor.

Regardless, at least since I have a 1200 V1, I can apply that hack and don't have to worry about DFS at all.

But definitely, DFS isn't being used correctly on the 3200ACM in LEDE.  If it were, you shouldn't be able to see or alter radio2 (88W8887) at all.  It should be used in the background to monitor for radar on the same wifi channel radio0 is connected to.  That is what it is meant for.  That is all it is meant for.  But mwlwifi won't and can't fix this behavior.  The 88W8887 uses the mwiflex driver, not mwlwifi.  So what we are missing is some algorithm or something that makes the 88W8887 (radio2) behave as the DFS monitor and communicate with the 88W8964 (radio0) to tell it that radar is detected and to change channels.  The more I think of it, for the 3200ACM, DFS detection shouldn't even be included or handled by the mwlwifi driver.

(Last edited by starcms on 21 May 2017, 00:05)