Orange pi zero wifi connection issue

[taner97]

With mainline kernel and its xradio module i cant connect to any open wifi networks. I tried wpa_supplicant network-manager and networking but this device just refuse to connect ap's that doesnt have encryption. This problem not related to performance nor coverage so please not say just crappy sources, dont have datasheet etc. and please try to fix this major problem.  Thanks.. 

[tao]

I found the wifi on the orange pi zero to be as bad as mentioned in various places in this forum. My needs and expectations were really low too.  I was disappointed, I love the zero, except for it's wifi performance.  I finally wound up sticking a usb dongle on the darn thing.  Not what I wanted, but it works "okay".

 

 I'm glad I found armbian.  It's a great distribution and lets me experiment with devices in all shapes, sizes, and cost.  And I get it for free!  Yeah, I wish the zero's wifi might work better, or work at all in your case, but hell, I love what these folks have done.  And that leads me to my point -- the wifi on the zero is crap.  Been said many times.  I find it hard to consider the software a "major problem" when it doesn't have decent hardware to work with.  Maybe you should leave a note on the Orange Pi forum asking them to improve their wifi hardware.  I'm pretty sure the armbian crewe would make a decent wifi component live up to it's potential. 

[Igor]

57 minutes ago, tao said:

Orange Pi forum asking them to improve their wifi hardware

I guess they realised and Zero2 is getting Ampak 6212, which is much better ... while I saw this crappy wireless chip on one Friendlyelec board (Neo plus 2)

[tao]

To clarify -- the Orange Pi Zero Plus 2 which has AP6212 for WiFi is much better than the WiFi on the Orange Pi Zero (which is bad even for us with low expectations). Would it be fair to say it will be good?  Or would it be more accurate to say it'll be acceptable?

 

If wifi is okay/good this is a nice looking board since it'll have wifi, Bluetooth,  8GB EMMC flash, and 512 MB RAM for ~ 19 USD.  The csi camera interface is nice too  but so far I haven't seen many good cameras that are compatible with the oranges (I really wish I could use my rpi cams on my oranges!).

[Igor]

1 hour ago, tao said:

Or would it be more accurate to say it'll be acceptable

Orane Pi Zero as an IOT node is good enough. To be more precise, transferring data in a client mode with low speed and low range demand, is acceptable, mainline (still development)  kernel is recommended in any case.

[taner97]

You're saying that mainline is recommended in any case but with legacy kernel i can connect to my open AP .Also in this thread

 you mentioned it as best driver but this problem only occurs on mainline.. If its lack of hw it cant be connected with legacy am i wrong? I appreciate your effort and i thank you for it but developers just given up with this xr819 driver and cant accept any discussion. I even tried to compile driver from sources and changed some flags but cant get it worked. I found a xradio to cw1200 port in git but cant compile it https://github.com/plaes/xradio. little help ? thanks again

[Bubba]

Ok I have the zero running a "version" of openwrt,  wifi works fine no issue for me, i have used it for hours of streaming video.

only doing a single stream at a time.

I did try to use Armbian builds from 5.20 ~ 5.26, but wifi was slow to connect, or would sometimes fail to connect, Tested with three different AP's (all have different chipsets). The zero is used as a client  not as AP.

 

I only did limited test as to range, and so on with the installs (home is only 1800 sq ft, 3 bedroom). my AP is mounted at 6 feet in my office and my daily use AP covers the whole house with very good to full signal coverage... BUT the max speed I have seen on the zero is around 17 meg. (not a real speedtest). In the small amount of using the zero outside i have done it still streamed netflix / hulu without buffering from 165 feet. 

 

Odd things i did notice, the armbian shows me many more AP's in my area than the WRT does, the WRT build runs MUCH cooler,

my router shows the armbian as connecting at 72 / where it shows the wrt build at 54, but as i said i did really did not do any point to point testing

 

 I had no luck getting the zero pluswifi to work on the OPI image I used,upon loading the AP6212 driver, once i open network manager and select my ssid and get the passphase window the system hard locks,OS becomes read only.

 

the wrt i am running is from http://pulpstone.pw/ , and yes i know it is a hacked up version using Armbian, but hey it works great for my limited use.

I will pickup some more SD cards and load up the newest Armbian and see what i get.

 

 am using a better antenna than the one provided on the zero

 

 

[Igor]

I think the only difference between this openwrt and our legacy is enabled power management (by mistake) in one of our early builds. Disable PM and wifi connecting shoul dramatically improve.

Wrote on mobile

[Bubba]

ok using a new updated install of Armbian_5.25_Orangepizero_Ubuntu_xenial_default_3.4.113 

using the stock ant. with the zero at 20 feet from the AP (Quality=70/70  Signal level=-28 dBm )

I disabled power management, and ran some iperf / ping test.

 

---The results below came from iperf test suggested here, they do not mean a thing in the real world use.---

ID] Interval       Transfer     Bandwidth
[ 10]  0.0-30.0 sec  0.00 (null)s  260195013222332 Bytes/sec

 

--- 192.168.0.1 ping statistics ---
55 packets transmitted, 55 received, 0% packet loss, time 54086ms
rtt min/avg/max/mdev = 2.729/52.794/151.154/27.817 ms
 

 

I did see that last image i had used was 5.27 mainline, which i see had known wifi issues.

 

Edited April 11, 2017 by Bubba
added that test a are meanless

[Bubba]

at 38 feet with two walls between the zero and the AP the results are just a little different; ping times increased about 15 ms

but Iperf shows about the same results.

armbianmonitor shows temp has been around 44 ~ 46 c at idle, upto to 63 c during upgrading and testing, I ran the upgrade at the same time to push to limits. 

on reboot it takes about 35 ~ 40 secs before I can get in.

 

not an issue as the fault was my the A/C / sunshine----I did noticed that if a I use a 1 amp P.S. vs  A 2 amp p.s. it runs a a little cooler---

 

 

signal was down to 50 / 70 and 58 dbi

Edited April 11, 2017 by Bubba
found reason for temp swings was NOT armbian

[taner97]

They dont seem to maintain legacy kernel but also dont give any offer for mainline issues. As i firstly said my problem is not related to speed performance ping coverage etc. its a general problem that blocks anyone who try connect public aps. if its not connecting wpa2 i understand but its just 802.11 stack without encryption and mainline driver lacks of this stupid connection. @Igor I dont understand why you leave my posts unanswered and commented off-topic problems.Im afraid to say that you and other devs behaves so arrogantly. If you want the community to grow just try to be more kind and give proper answers to reasonable problems for your OS. It's the manner of linux based distros and open source communities

[zador.blood.stained]

34 minutes ago, taner97 said:

I dont understand why you leave my posts unanswered and commented off-topic problems.Im afraid to say that you and other devs behaves so arrogantly. If you want the community to grow just try to be more kind and give proper answers to reasonable problems for your OS. It's the manner of linux based distros and open source communities

What kind of answers do you want? I believe everything was already answered in this sticky post: 

So I'll answer again:

• XRadio support package is a pile of crap
• This fact is listed as a known issue for the Orange Pi Zero
• In addition to the driver quality there are also firmware problems or possible incompatibility between the firmware and the driver
• Currently nobody is willing to waste time with this hardware unless vendors provide documentation, fixed firmware and driver (which most likely will never happen)
• Open source community is not obliged to fix every issue given to it by the vendor choice of the wireless chip

In addition:

• Mainline (dev branch) images currently are provided "as is" without any end user support. We are accepting build failure reports, Device Tree overlay requests, kernel configuration change requests and reports not related to the possible kernel issues (Edit: and obviously we are accepting pull requests with kernel fixes). Just the fact that we provide mainline based configuration and images does not mean that we are obliged to improve it and fix every existing issue - we are trying to do what we can, but at the current state it is mostly waiting for the improvements and integrations upstream.

 

Edit 2:

So currently there is not enough information to fix this right away. It is unknown what exactly is causing inability to connect to open networks and what exactly needs to be changed to make it work, and unless more information is provided it will be this way.

[taner97]

I have posted a git repo that merges cw1200 driver with xradio support and asked how can i compile but i gave no answer. I think its like pull request with little support request for a beginner. I dont know if you tried to build that but it sounds more stable for me if got it working. Also i cant find any declaration of this open ap connection problem as known issue.

[zador.blood.stained]

51 minutes ago, taner97 said:

I have posted a git repo that merges cw1200 driver with xradio support and asked how can i compile but i gave no answer.

Like any other kernel module - either compile it out of tree or add it as a patch to the build system. In the first case you may need to remove or blacklist existing driver and you may need to change DT according to what that driver needs for initialization, in the second case you may need to disable or replace existing XRadio patch. This may not be a simple task for a beginner since it requires some skills and knowledge.

 

58 minutes ago, taner97 said:

I dont know if you tried to build that but it sounds more stable for me if got it working.

Feel free to test and report. For now Orange Pi Zero wireless will remain in this state unless more information is provided.

 

1 hour ago, taner97 said:

Also i cant find any declaration of this open ap connection problem as known issue.

https://www.armbian.com/orange-pi-zero/   ->   "Known issues" tab

[tkaiser]

Should we add links like the following to OPi Zero download page? ("If you want to do fancy Wi-Fi stuff you need an external device like this. Onboard Wi-Fi NOT SUFFICIENT on this board")

• https://forum.armbian.com/index.php?/topic/1237-tutorial-opi-one-wireless-success/
• https://www.aliexpress.com/item/150mbps-RT5370-Mini-Wireless-USB-Adapter-Lan-Card-802-11n-g-b-USB-Wifi-Receiver-Wifi/32664731198.html 

 

 

[zador.blood.stained]

9 minutes ago, tkaiser said:

Should we add links like the following to OPi Zero download page? ("If you want to do fancy Wi-Fi stuff you need an external device like this. Onboard Wi-Fi NOT SUFFICIENT on this board")

If nothing changes regarding firmware/drivers, official (supported) mainline releases should be provided without the XRadio driver and with explicit notice like "Onboard wireless is not supported".

For now we should just link one of Xradio status threads/posts in the "Known issues" notice (or further expand on "connection issues", but not more than 2-3 lines), since the driver more or less works on legacy.

[Bubba]

I loaded the 5.27 mainline image and for now wifi works fine for me (no it will not connect to an open AP) running a few load test now

 

 I did solve a couple of the anomaly's, my temp swings where caused by the A/C vent blowing on the device (back room has own unit ) and the sun hits the table where the device was running a few mins during the day.

 

 

[tkaiser]

On 9.4.2017 at 7:40 PM, Bubba said:

260195013222332 Bytes/sec

 

So now that you figured out what you've been doing wrong regarding the temperatures you were reporting what about trying to clean up this thread and using this style here to mark stuff as irrelevant in your posts above? Would help a lot trying to keep Armbian forum a source of information and not confusion! Thanks!

 

BTW:  260195013222332 Bytes/sec are just ~242325 GB/sec or ~236 TB/sec. Impressive numbers if you can push Terabytes of data in seconds with this tiny hardware.

[Bubba]

OK tkaiser, I will try to do better.

I was only reporting what I see, some people here (hint hint) slam these DIRT CHEAP boards for the wi-fi.

 

I have found that NOT to be the case for someone who uses them as they should (IOT / micro pc for robot / drone control).

 

I was not worrying with trying to use 10.00 board to replace my 200.00 router, I was just making sure the boards where safe and usable for the kids to use in the projects, you know NOT start a fire or burn themselves, NOT be working on code to run on a board which would fail.

 

I load / stress testing for a better idea of the boards limits.  both the zero and zero plus are better than the batch of RPI's I got them, main reason is the Armbian crowd here.

my posting are to let others know that they should filter out those negative post, because in fact the wi-fi does work, and is usable as intended.

when someone gives NO info on their setup, nor do they troubleshoot, they just whine, those post are useless, NOTHING can be learned from that.

most people do not know how wi-fi works, they have never heard of Shannon's capacity limit,

 

[taner97]

I'm just laughin my ass off. Im using wifi since 2007 and back in time there is almost no one set passwords for it and its primitive function of wifi is being without encryption. Its not just only my case and should be exposed by anyone so i dont need to share debug info which are just useless. Also for all yours information i have f... dirt cheap 1.7$ esp8266 e-12 thing which have 80mhz cpu 96kB ram and i have no trouble while connecting to my open wifi and share same time(AP+STA) and i got near 8 Mbit/s speed from it. Again im not complaining for speed but connection. hint hint Bubba do you know taylor series and lagrange DE im studying it now lol. 

addition- My posts intended to do bug report and trying to do troubleshoot with supplying different driver sources but yours are completely flood and off-topic. Its discussed on above sticky thread (not mine) and i assume youre a troll for your last post

[Bubba]

 

taner97...Dude Chill life is short.
I am sorry if you felt that I was singling you out, sorry you took the post that way.

I was posting my observations in response the the whole thread.

For you I loaded the mainline and set up my 3 test AP’s as OPEN networks, verified your and  re-forced your post “ i can't connect to any open wifi networks”

and “but this device just refuse to connect ap's that does not have encryption.” then I took the time to try and force it to connect.

Glad you think it was a waste of YOUR time.

 

You have been doing wireless since 2007.... OK.. I got wi-fi gear older than you. 
 

I am VERY old, I bet I retired (the first time) from IT, I installed Muni / Corp wi-fi for almost 25 years;  before you ever touched a PC, can you say "Proxim"??? I knew could.

 I got stuck fixing a project which my grandkids need for robotics competition.

I picked up RPI’s from some website, then I got a ad for OPI and here we are. I am back at this trying to learn something to help them.

The last single boards I purchased before this was back in the 1990's, they ran Super DOS / and other OS most folks never heard of, so this whole arm cpu thing is all new to me
I got to get back under my bridge now. Good Luck

             \|||/
            (o o)
------ooO(_)Ooo------   
Kids what can do?
One should try to make one self to more than they are, never think you are the smartest one in the room, the moment you do. you are not
 

[prajil]

hie friends,

i have similar issue on WiFi configuration.

 

I'm using 'orange pi zero' with 'Armbian_5.25_Orangepizero_Ubuntu_xenial_default_3.4.113' image

i found solution by following steps:

sudo cp /etc/network/interfaces /etc/network/interfacesoriginal
sudp rm /etc/network/interfaces
sudo nano /etc/network/interfaces
----------------------------------
COPY
----------------------------------
auto lo
iface lo inet loopback

auto eth0
allow-hotplug eth0
iface eth0 inet dhcp

allow-hotplug wlan0
iface wlan0 inet manual
wpa-roam /etc/wpa_supplicant/wpa_supplicant.conf
iface default inet dhcp
---------------------------------
Ctrl + O
Enter
Ctrl + X
sudo nano /etc/wpa_supplicant/wpa_supplicant.conf

----------------------------------
COPY
----------------------------------
ctrl_interface=DIR=/var/run/wpa_supplicant GROUP=netdev
update_config=1

network={
ssid="SECURE-WIFI"
psk="yourpassword"
}

network={
ssid="OPEN-WIFI"
key_mgmt=NONE
}
----------------------------------
Change based on your WiFi Settings
Ctrl + O
Enter
Ctrl + X
sudo reboot

Your device should now connect to your WiFi network.

here is the link : "https://www.youtube.com/watch?v=8zGRAPguaSQ

[tkaiser]

5 hours ago, prajil said:

Your device should now connect to your WiFi network.

Same with the non anachronistic way (using nmtui as recommended). BTW: http://www.cnx-software.com/2016/11/10/allwinner-h2-linux-android-sdk-and-allwinner-xr819-wifi-driver-released/#comment-541282

[dgp]

On 09/04/2017 at 9:56 AM, taner97 said:

This problem not related to performance nor coverage so please not say just crappy sources, dont have datasheet etc. and please try to fix this major problem.  Thanks.. 

 

Sigh. I thought I'd spend the weekend working on the xradio driver again to see if I could work out why it drops so many frames and then I see this in the google search results while trying to find some more info on the cw1200 (if there is a register to query the state of the buffers). I find it really crazy that someone can come along and make a stink about people choosing not to do work for free and while providing almost no information to help debug their issue make out that it's an easy fix.

 

I guess I'll read a book or something instead.

[taner97]

@dgp please i beg you to fix it if you do i will give you 5000$ paycheck for spending your precious weekend just for me. If you do believe your writings please dont use unix/linux kernel (this is for me).Go for microsoft windows bullshit pay to get your privates sold bla bla.. The kernel that you using is being maintained for free and its free, there is a little part customized by allwinner and armbian(since mainline supports sunxi) and youre saying im not spending my weekend for it lol.If you prefer not to do anything for community i respect you but please keep it for yourself and dont waste your time for typing keep on reading your book or something instead. Me myself still trying to do make a progress from almost zero acknowledge , still waiting my celeron netbook to compile kernel there is almost 5 hours and this is not just for me there must be the people suffer from it.If i do manage to fix it i share my work and do proud of helping community not ....ask money for it.

@prajil

I thank you sincerely for trying to do something unlike others but as i said on earlier post i tried wpa_supplicant but its also not working for mainline kernel(4.10 -4.10.3 and probably earliers).If youre using open wifi i advice you to keep with legacy kernel.

[tkaiser]

14 minutes ago, taner97 said:

Me myself still trying to do make a progress

 

Seriously?

 

You do not even realize that @dgp is the person behind https://github.com/fifteenhex/xradio (one of the guys making this driver working in the first place). Please continue with your great contributions. I will from now on stop reading the H2/H3 subforum.

[taner97]

He seems to do commit 3 days ago but the person there talks arrogantly needs debug which must already known by him anyway i didnt know that and apologize for my words. It seems my hope from github is fizzle and i will from now on going with legacy kernel

[dgp]

19 hours ago, taner97 said:

but the person there talks arrogantly

 

You really couldn't make this shit up could you.

 

19 hours ago, taner97 said:

debug which must already known by him anyway i didnt know that and apologize for my words.

As far as I know connecting to open aps and being an open ap with hostapd works.

 

The main issues with the driver is that performance is awful because of all the frames it drops and that it's not structured very well and can lock up the kernel if it gets into a bad state. These things are also true if you compile the code dump from allwinner on their hotchpotch prehistorical kernel.

 

FYI - A weekend of work on stuff like this will cost you more than $5000. The source for the kernel may be "free" but getting people to do work on stuff you want costs money and lots of it. You might want to think of all the free work people here have put into putting together something useful before writing any more of your bizarre posts.

[tao]

I'd like to say that I appreciate the contributed work. It's easily overlooked and taken for granted, but it shouldn't.  By the work of many others I get the opportunity to play with new, inexpensive, and fun alternatives to the rpi.  And that is a hoot.  I shop and buy new orange pis and neo pis with the confidence that armbian will work.  And in those cases where there are issues I can ask my question here and not be mocked and shamed.  For all the work, and for the forum to lean ... Thank you. 

[taner97]

My words of

Quote

s problem not related to performance nor coverage so please not say just crappy sources, dont have datasheet etc. and please try to fix this major problem

is for that anyone who writes about xradio drivers are ignored irrespective of what is written so i want to clarify that im not be annoyed of low throughput and i just want to connect my dorms wifi.. So as long as this open wifi connection bug is not introduced and developers doesnt know about that im here to do bug reporting that maybe can easily solved which i'm believing to.. Sorry for my bad and long english sentences kudos for developers who are still into it  

[dgp]

27 minutes ago, taner97 said:

that im not be annoyed of low throughput and i just want to connect my dorms wifi..

 

The amount of retries needed to get a frame in/out of the wifi interface affects more than just throughput. It can break association, DHCP etc. Of course you know all of this because you know better than everyone else right?

 

28 minutes ago, taner97 said:

So as long as this open wifi connection bug is not introduced and developers doesnt know

 

It works for me from what I remember. I actually broke WPA/WPA2 support at some point trying to get the chip to pass encrypted frames directly to the kernel and connecting to open aps was the only thing that worked. 

As far as I can tell you haven't actually posted anything to show what happens when you try to connect to an open AP let alone anything that shows your issue is an issue with the xradio driver.

 

And... it could be an issue that is unsolvable without more details on the firmware running on the chip. The xr819 doesn't just take frames off of the air and pass them to the host. Maybe it thinks it should be decrypting frames it shouldn't? Who knows.

 

35 minutes ago, taner97 said:

im here to do bug reporting that maybe can easily solved which i'm believing to

 

Maybe it is easy to fix but where are the logs and packet dumps? I'm not sure what your idea of reporting a bug is but it's not "This doesn't work for me, fix it now!".  I don't think you have ever said where it fails to connect.. not associating outright is different to no passing any IP traffic etc.