Stanislav Sinyagin Posted July 18, 2017 Posted July 18, 2017 FriendlyARM has recently launched a new board in its NEO family: NanoPi NEO Plus2 As of Armbian_5.32.170704_Nanopineo2_Ubuntu_xenial_dev_4.11.8.img, few things aren't working: out of 1GB RAM, only 512MB is visible the USB socket closer to the edge (USBN1) is not visible, even with overlays Wifi and Bluetooth are not visible (AP6212A) (it's not a complaint, but just a statement
Igor Posted July 18, 2017 Posted July 18, 2017 Somebody needs to create proper u-boot configuration and kernel device tree BTW. Engineering samples, which we received, are different - they don't have AP6212A but "beloved" XR819 known from Opi Zero. I assume the rest is the same, but it has to be checked.
Stanislav Sinyagin Posted July 18, 2017 Author Posted July 18, 2017 thank goodness they decided for a better chip I just checked, the Ubuntu image that FriendlyARM is providing sees all those components correctly.
Stanislav Sinyagin Posted July 18, 2017 Author Posted July 18, 2017 almost forgot: there's onboard 8GB eMMC chip (actually preloaded with Ubuntu image), and it's also not visible if I boot the Nanopineo2 Armbian image. It will be great to have a way to push Armbian onto that chip too.
etatto Posted July 19, 2017 Posted July 19, 2017 And mine has got an AP6212 without the A The pictures on FriendlyARM site show an AP6212A on the PCB, but the features listing and the pdf schematics clearly indicate AP6212
Stanislav Sinyagin Posted July 22, 2017 Author Posted July 22, 2017 On 19.7.2017 at 4:10 PM, etatto said: AP6212 without the A haha, mine too. I just didn't bother to look at the chip
tkaiser Posted July 22, 2017 Posted July 22, 2017 On 18.7.2017 at 3:15 PM, Igor said: I assume the rest is the same, but it has to be checked. Voltage regulation is also different (none on the engineering samples, after I explained to FriendlyELEC's Weidong that voltage regulation just don't work with Allwinner's shitty legacy kernel but there's no problem with mainline he said they want to redesign NEO Plus2 to switch between 1.1V and 1.3V again and NanoPi M1 Plus 2 should even get I2C accessible voltage regulation like on larger Orange Pis. On 19.7.2017 at 4:10 PM, etatto said: And mine has got an AP6212 without the A Don't fool yourself. Ampak just recently started to mark AP6212A correctly, the AP6212A chips we find on many boards within the last half year all show 'AP6212' instead. It's only a matter of firmware (and maybe still a tweak is needed with mainline kernel, didn't had a look since 1T1R/2.4GHz Wi-Fi is low performing crap anyway)
@lex Posted July 22, 2017 Posted July 22, 2017 1 hour ago, tkaiser said: voltage regulation just don't work with Allwinner's shitty legacy kernel but there's no problem with mainline Would you mind elaborating how a fixed 1.1v for all frequencies on H5 works on mainline and not on legacy? I would like to understand how this should work, from a hardware point of view an then from a software point of view, may i suggest a specific thread? Yes i know you have discussed this for the H3.... Feel free to move, delete or whatever.
zador.blood.stained Posted July 22, 2017 Posted July 22, 2017 8 minutes ago, @lex said: Would you mind elaborating how a fixed 1.1v for all frequencies on H5 works on mainline and not on legacy? Fixed voltage should work fine both on mainline and legacy, but if you want switching it on legacy - good luck&have fun since legacy relies on the ARICS firmware for DVFS, and at least people who tried to make it work didn't have any success.
tkaiser Posted July 23, 2017 Posted July 23, 2017 10 hours ago, @lex said: I would like to understand how this should work, from a hardware point of view an then from a software point of view, may i suggest a specific thread? As far as I know with Allwinner's H5 BSP an OPi PC 2 is still not able to clock higher than 1008 MHz since here voltage regulation is simply broken (it might have changed but I stopped looking into Xunlong's/Buddy's Github repo since causing headache). FriendlyELEC chose NanoPi NEO 2 to be a fixed voltage SBC for this reason: since they couldn't get voltage regulation to work. Then in March Icenowy and me were talking to them and I mentioned that voltage regulation is working already with mainline kernel. They discussed internally, decided to drop BSP support for their H3 and H5 boards and since they now can make use of DVFS to also re-design both NanoPi Neo Plus2 and NanoPi M1 Plus 2. 1
tkaiser Posted July 23, 2017 Posted July 23, 2017 Stupidity 2.0: https://github.com/BPI-SINOVOIP/BPI_WiFi_Firmware/commit/6dee7aabad92112e548b551c5acb9611d15e5b33
bamartin125 Posted August 22, 2017 Posted August 22, 2017 I just received my NanoPi Neo Plus2 yesterday and am having some issues with the FE provided image regarding UART2's RX (interestingly the UART2 TX side works and the other UARTs work as well). I am willing to see what I can do about getting the armbian build working with this board since I notice that it still needs work. I guess the first step would be to download the armbian source to a virtual machine and try to build the target for my board. It looks like the priority to get armbian going with this board is to get the uboot patch done. Can someone point me to any hints/resources to get started with uboot build configuration? Maybe it is just a matter of looking at what FE is doing in their uboot build (I haven't looked yet) and replicating that for armbian while looking at armbian uboot patch examples for similar boards.
antony Posted August 23, 2017 Posted August 23, 2017 Quote RAM detected 512MB. free total used free shared buff/cache available Mem: 495024 28120 382452 3456 84452 441272 And from dmesg [ 0.000000] memory : 0xffff800000000000 - 0xffff800020000000 ( 512 MB) I didn't check reported RAM with FE shipped image. So far I tested: - Armbian build nanopineoplus2 Kernel 4.11 Xenial - Boot from micro SD - Install to eMMC
gdm85 Posted September 20, 2017 Posted September 20, 2017 FYI, as I reported in this thread, the second USB works with their official Ubuntu Xenial core image (didn't test any other of theirs). So now I am trying to figure out if it's a kernel configuration issue or (more likely) DTB/kernel driver level.
antony Posted September 21, 2017 Posted September 21, 2017 17 hours ago, gdm85 said: FYI, as I reported in this thread, the second USB works with their official Ubuntu Xenial core image (didn't test any other of theirs). So now I am trying to figure out if it's a kernel configuration issue or (more likely) DTB/kernel driver level. I see two usbs [ 314.305133] sd 0:0:0:0: [sda] 60566016 512-byte logical blocks: (31.0 GB/28.9 GiB) [ 314.334507] sd 0:0:0:0: [sda] Attached SCSI removable disk [ 335.621769] sd 1:0:0:0: [sdb] 15482880 512-byte logical blocks: (7.93 GB/7.38 GiB) [ 335.636197] sd 1:0:0:0: [sdb] Attached SCSI removable disk
teknoid Posted October 4, 2017 Posted October 4, 2017 @Igor, could you please provide a debian based installation for the neo plus2? I only found the ubuntu xenial image for the moment. In the mean time I manually installed stretch to my neo plus2 device: preparations: - cross compile uboot and kernel from friendlyarm's git - install ubuntu image (or dietpi or something else) on external sdcard drive installation: - boot system from sdcard - dd the nanopineo2 image to internal emmc drive - still boot kernel from sdcard but set root to emmc (setenv bootargs "console=ttyS0,115200 earlyprintk root=/dev/mmcblk3p1 ro") - update uboot on /dev/mmcblk3 - install kernel - create boot.scr from following boot.cmd: # Recompile with: # mkimage -C none -A arm64 -T script -d /boot/boot.cmd /boot/boot.scr setenv bootargs "console=ttyS0,115200 earlyprintk root=/dev/mmcblk3p1 ro" ext2load mmc 0 0x43000000 boot/dtb/allwinner/sun50i-h5-nanopi-neo-plus2.dtb ext2load mmc 0 0x46000000 boot/Image booti 0x46000000 - 0x43000000 - run apt-get dist-upgrade to move up to stretch Works perfectly: | \ | | __ _ _ __ ___ | _ \(_) | \ | | ___ ___ |___ \ | \| |/ _` | '_ \ / _ \| |_) | | | \| |/ _ \/ _ \ __) | | |\ | (_| | | | | (_) | __/| | | |\ | __/ (_) | / __/ |_| \_|\__,_|_| |_|\___/|_| |_| |_| \_|\___|\___/ |_____| Welcome to ARMBIAN 5.31 stable Debian GNU/Linux 9 (stretch) 4.11.2 Linux sabre 4.11.2 #1 SMP Sun Oct 1 16:01:05 CEST 2017 aarch64 System load: 0.13 0.05 0.01 Up time: 1 day Local users: 2 Memory usage: 11 % of 994MB Swap usage: 13 % of 127Mb IP: 192.168.1.245 CPU temp: 29°C Usage of /: 29% of 7.2G storage/: 1% of 367G [ General system configuration: armbian-config ]
teknoid Posted October 4, 2017 Posted October 4, 2017 btw. I use both internal USB ports: one is conneted to a harddisk /dev/sda and the other fires my ES9018 DAC via USB XMOS interface. hje@sabre:~$ dmesg |grep sda [ 3.656921] sd 0:0:0:0: [sda] 781422768 512-byte logical blocks: (400 GB/373 GiB) [ 3.657914] sd 0:0:0:0: [sda] Write Protect is off [ 3.657918] sd 0:0:0:0: [sda] Mode Sense: 43 00 00 00 [ 3.661772] sd 0:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA [ 5.881354] sda: sda1 [ 5.887091] sd 0:0:0:0: [sda] Attached SCSI disk hje@sabre:~$ aplay -l **** List of PLAYBACK Hardware Devices **** card 0: sndsun8ii2sdac [snd-sun8i-i2s-dac], device 0: sun8i-i2s-dac pcm5102a-hifi-0 [] Subdevices: 1/1 Subdevice #0: subdevice #0 card 1: Codec [H3 Audio Codec], device 0: CDC PCM Codec-0 [] Subdevices: 1/1 Subdevice #0: subdevice #0 card 2: x20 [xCORE USB Audio 2.0], device 0: USB Audio [USB Audio] Subdevices: 0/1 Subdevice #0: subdevice #0
Igor Posted October 5, 2017 Posted October 5, 2017 Check latest nightly: https://dl.armbian.com/nanopineo2/nightly/ Untested and also not ready for production!
Lope Posted October 5, 2017 Posted October 5, 2017 I'm testing the nightly image: Armbian_5.32.170704_Nanopineo2_Ubuntu_xenial_dev_4.11.8.img for NanoPi Neo2 H5. Every time if my Neo2 has been off for many minutes, if I power it up it doesn't boot on the first attempt. If I unplug and replug power a few seconds later (either at the 220v end or the microUSB, it boots up flawlessly. Anyone experienced this? Any ideas? Maybe it's a power stabilization issue. I guess there should be a 2 second or whatever delay before it starts booting so it can be as reliable as other Linux SBC's. Other than that, surely there's a watchdog built in to make it reboot if it doesn't boot successfully?
Igor Posted October 5, 2017 Posted October 5, 2017 This is somehow expected behaviour for the kernel in development. It doesn't work properly! Yet. This image is for testing - not for using since it's not stable enough. Try this image: https://dl.armbian.com/nanopineo2/Debian_stretch_dev_nightly.7z It should be in better shape, but it's also not ready. It lacks thermal protection.
teknoid Posted October 5, 2017 Posted October 5, 2017 Hi @Igor I just tested the nightly neo2 image on my second plus2 device - It works. But I _strongly_ recommend using the official friendlyarm's 4.11.2 kernel. The linux-sunxi kernel included in your armbian image does not really fit at the moment. If you replace the kernel I think you can release an official armbian for NanoPi NEO2 / Plus2 devices - coming hopefully in ~35 days ;-) root@192.168.1.166's password: _ _ ____ _ _ _ ____ | \ | | __ _ _ __ ___ | _ \(_) | \ | | ___ ___ |___ \ | \| |/ _` | '_ \ / _ \| |_) | | | \| |/ _ \/ _ \ __) | | |\ | (_| | | | | (_) | __/| | | |\ | __/ (_) | / __/ |_| \_|\__,_|_| |_|\___/|_| |_| |_| \_|\___|\___/ |_____| Welcome to ARMBIAN 5.33.171006 nightly Debian GNU/Linux 9 (stretch) 4.11.2 System load: 0.30 0.08 0.03 Up time: 10 min Memory usage: 5 % of 994MB IP: 192.168.1.166 CPU temp: 41°C Usage of /: 16% of 7.2G
Igor Posted October 5, 2017 Posted October 5, 2017 39 minutes ago, teknoid said: If you replace the kernel I think you can release an official armbian Read one post up (merged from another open topic). We are at 4.13 but also that kernel is simply not finished. Comparing unfinished work to some quickly patched EOL kernel is pointless. In any case, you will be more or less forced to move to our image or the same from copycats since board makers usually don't provide updates. 39 minutes ago, teknoid said: coming hopefully in ~35 days ;-) Put cash on the table and we will hire professional help to make it sooner. You can also join the project with your own resources. Much better than complaining.
teknoid Posted October 5, 2017 Posted October 5, 2017 7 minutes ago, Igor said: since board makers usually don't provide updates. I know... Odroid C2... it's a nightmare. Thats why I moved to Nanopi and I was very surprised that I got a recent 4.x kernel up and running with all board features working that I need. As long as I can simply replace the kernel by myself everything is fine...
Igor Posted October 5, 2017 Posted October 5, 2017 5 minutes ago, teknoid said: Odroid C2 Is getting mainline support which means it's interesting again. We convinced FriendlyARM to rather proceed with kernel 4.x since this board is simple & headless and most of the base functions are mainlined. Our perspective is different than board makers - quick fixes are usually waste of time.
chrisf_ Posted October 19, 2017 Posted October 19, 2017 On 22/08/2017 at 4:01 PM, bamartin125 said: I just received my NanoPi Neo Plus2 yesterday and am having some issues with the FE provided image regarding UART2's RX (interestingly the UART2 TX side works and the other UARTs work as well). Did you (or anyone else) get UART2 RX working reliably? I'm seeing very strange behaviour on the same board (extra RX characters received when TX is sent - no loop back!) and wondered if it might be a HW problem (though it's probably my code!).
farfeduc Posted October 20, 2017 Posted October 20, 2017 Hello, I am currently using a Raspberry Pi 3 for a small project and I would like to use the NanoPi Neo Plus2 instead. In my project, I use a program which reads and writes the RPi GPIO through /dev/ttyAMA0 (which if I am not mistaken is for GPIO 14 and 15). I have not been able to find an equivalent on the NanoPi Neo Plus2. Could you point me in the right direction ? Do I have to change the kernel or something ? Would it work if I install Armbian instead of the preinstalled Ubuntu Core ?
chrisf_ Posted October 20, 2017 Posted October 20, 2017 On 19/10/2017 at 9:19 PM, chrisf_ said: Did you (or anyone else) get UART2 RX working reliably? I'm seeing very strange behaviour on the same board (extra RX characters received when TX is sent - no loop back!) and wondered if it might be a HW problem (though it's probably my code!). Ignore above post. Problems were elsewhere. UART2 works fine.
Recommended Posts