• Content Count

  • Joined

  • Last visited

Reputation Activity

  1. Like
    krasoffski reacted to gtr33m in OrangePi Zero ALSA hardware not found on mainline   
    I'm trying to use an OrangePi Zero as an Alexa device, using  This is headless, but I need analogue out and in.
    Given that it's headless, I thought I'd use the mainline kernel instead of headless.  Everything installs fine, but ALSA does not see anysound card hardware.
    aplay: device_list:268: no soundcards found...
    I can try legacy, but I'd prefer to use mainline if possible.
  2. Like
    krasoffski reacted to tkaiser in H3 board buyer's guide   
    H2+/H3/H5 boards overview (2017/03 update)
    Since it has been a while since this topic has been updated and a lot of new boards have been released in the meantime it's time for a new overview.
    I'll add also H2+ and H5 based boards since in the meantime we learned that those SoCs are pin-to-pin compatible and recently vendors started to simply exchange H3 with H5 on some PCB (and vice versa in at least one occurence). From a software point of view H5 is quite different (using 64-bit Cortex-A53 CPU cores and ARMv8 instruction set, some early boot stages are also totally different compared to Cortex-A7/ARMv7 used in H3 and H2+) and it should also be noted that Armbian currently only provides OS images based on mainline kernel for H5 boards (so please forget about HW accelerated video decoding or 3D for now or maybe ever since none of the developers is in the mood to deal with Allwinner's BSP/legacy kernel for H5 (regarding 'BSP' just look above in post #2).
    While software support for H5 is currently somewhat different hardware features are pretty much the same as with H3 (still 3 to 4 real USB2 host ports and one USB2 OTG port: a simple register setting can switch the Micro USB port's PHY between the so called 'musb' controller used for OTG and a real EHCI/OHCI controller pair: with mainline kernel it will soon be possible to switch OTG to a real 4th USB2 host port with full feature set that still has not to share bandwidth with any of the other USB ports).
    CPU performance with H5 compared to H3 is slightly higher at the same clockspeed but some workloads that benefit from either 64-bit or ARMv8 instruction set are significantly faster (eg. software making use of NEON instructions might perform almost twice as fast and the best example is the stupid 'sysbench' CPU pseudo benchmark which shows over 10 times better scores on the same hardware when compiled with ARMv8 settings).
    In the following list I will also introduce some subjective 'categories' to deal better with the huge amount of boards we can use in the meantime:
    NAS category: these are the H3/H5 boards with Gigabit Ethernet IoT category: these are the small and cheap boards best suited for low consumption 'General purpose' category: all the other H3 devices, these are also those you should look for if you want a cheap device to run with X11, OpenELEC, RetrOrangePi or Lakka since they all feature HDMI and full legacy kernel support As already said the differentiation is subjective and partially misleading since new boards like NanoPi NEO 2 featuring Gigabit Ethernet are also that inexpensive, small and energy efficient that they could serve both as NAS and IoT nodes (actually you can somewhat control behaviour since GbE vs. Fast Ethernet makes a pretty huge difference in consumption so it's up to you). Boards that might fit in multiple categories are listed more than once to make comparisons more simple if you're only interested in a specific device category:
    NAS category (only due to Gigabit Ethernet available):
    Banana Pi M2+: H3, 1GB DRAM, 8GB slow eMMC, 1+2 USB ports useable, Wi-Fi/BT Banana Pi M2+ EDU: H3, 512MB DRAM, no eMMC, 1+2 USB ports useable NanoPi M1 Plus: H3, 1GB DRAM, 8GB slow eMMC, 1+3 USB ports useable, Wi-Fi/BT NanoPi M1 Plus 2: H5, 1GB DRAM, 8GB slow eMMC, 1+3 USB ports useable, Wi-Fi/BT NanoPi NEO 2: H5, 512MB DRAM, no eMMC, 1+1+2 USB ports useable NanoPi NEO Plus 2: H5, 512MB DRAM, no eMMC, 1+2+2 USB ports useable, Wi-Fi OrangePi PC 2: H5, 1GB DRAM, no eMMC, 1+3 USB ports useable OrangePi PC Prime: H5, 2GB DRAM, 1+3 USB ports useable, Wi-Fi/BT OrangePi Plus: H3, 1GB DRAM, 8GB eMMC, 1+4 USB ports useable (hub), Wi-Fi OrangePi Plus 2: H3, 2GB DRAM, 16GB fast eMMC, 1+4 USB ports useable (hub), Wi-Fi OrangePi Plus 2E: H3, 2GB DRAM, 16GB fast eMMC, 1+3 USB ports useable, Wi-Fi IoT category (cheap, small, energy efficient, most of them headless):
    NanoPi Air: H3, 512MB DRAM, 8GB slow eMMC, 1+1+2 USB ports useable, Wi-Fi/BT, no Ethernet NanoPi NEO: H3, 256/512MB DRAM, no eMMC, 1+1+2 USB ports useable, Fast Ethernet NanoPi NEO 2: H5, 512MB DRAM, no eMMC, 1+1+2 USB ports useable, Gigabit Ethernet NanoPi NEO Plus 2: H5, 512MB DRAM, no eMMC, 1+1+2 USB ports useable, Wi-Fi, Gigabit Ethernet OrangePi Zero: H2+, 256/512MB DRAM, no eMMC, 1+1+2 USB ports useable, Wi-Fi, Fast Ethernet OrangePi Zero Plus 2: H3, 512MB DRAM, 8GB fast eMMC, 1+0+2 USB ports useable, Wi-Fi/BT, no Ethernet but HDMI OrangePi Zero Plus 2: H5, 512MB DRAM, 8GB fast eMMC, 1+0+2 USB ports useable, Wi-Fi/BT, no Ethernet but HDMI General purpose (HDMI and full legacy kernel support: video/3D HW accelerated):
    Beelink X2: H3, 1GB DRAM, 8GB slow eMMC, 1+1 USB ports useable, Wi-Fi, Fast Ethernet NanoPi M1: H3, 1GB DRAM, no eMMC, 1+3 USB ports useable, Fast Ethernet OrangePi Lite: H3, 512MB DRAM, no eMMC, 1+2 USB ports useable, Wi-Fi, no Ethernet OrangePi One: H3, 512MB DRAM, no eMMC, 1+1 USB ports useable, Fast Ethernet OrangePi PC: H3, 1GB DRAM, no eMMC, 1+3 USB ports useable, Fast Ethernet OrangePi PC Plus: H3, 1GB DRAM, 8GB fast eMMC, 1+3 USB ports useable, Wi-Fi, Fast Ethernet OrangePi Zero Plus 2: H3, 512MB DRAM, 8GB fast eMMC, 1+1+2 USB ports useable, Wi-Fi/BT, no Ethernet pcDuino Nano 4: See above, it's just an OEM version of NanoPi M1 done for Linksprite Some important notes:
    The following boards are listed in more than 1 category due to advanced feature mix: NanoPi NEO 2, NanoPi NEO Plus 2 and OrangePi Zero Plus 2 H3/H5 CE/FCC certifications: Please check individually and don't trust in logos silkscreened on the PCB, even if it looks like 'CE' it might mean 'China Export' instead IO bandwidth: H2+/H3/H5 SoC features 3+1 USB2 ports but on a few boards an internal USB hub is used so while these expose more USB receptacles some ports have to share bandwidth. Also on these boards a buggy/slow GL830 USB-to-SATA bridge is used. Search for 'hub' above to identify them. eMMC: shows most of the times higher random IO performance compared to 'the average SD card', but some vendors use pretty slow eMMC on their boards (Xunlong being the exception with OPi PC Plus, Plus, Plus 2, Plus 2E and Zero Plus 2). Please do not overestimate eMMC -- there's no need to choose crappy/slow SD cards and if you follow the usual recommendations difference in performance varies not that much (for example eMMC on most boards shows pretty low sequential write speeds that will be easily outperformed by any good SD card and differences in random IO don't have to be that huge, simply watch out for SD cards showing A1 or even A2 logo) USB ports: Some of the IoT devices have two of the SoC's USB host ports available on a pin header to be used with soldering or combined with various Docks, HATs or 'Expansion boards' (search for '1+1+2' above). On OPi One/Lite the unexposed USB host ports are available at pretty tiny solder pads so only usable with a lot of soldering experience Wi-Fi/BT: all boards providing both Wi-Fi and BT rely on Ampak's AP6212 so performance is identical, the Wi-Fi only boards either rely on RTL8189ETV/8189FTV (slightly better Wi-Fi performance than AP6212) or Allwinner's XR819 (so expect low Wi-Fi performance with OPi Zero or NEO Plus 2 since implementation is low-end and currently driver sucks) Yeah, each vendor's naming scheme totally sucks. Partially there are rules involved (the 'Plus' then means eMMC with Xunlong or GBit Ethernet with FriendlyELEC... mostly) but please don't trust in and check always individually!  
    And now another few words on a different technical detail affecting both performance and thermal behaviour of the various boards: Voltage regulation / DVFS. TL;DR: the SoC can be fed with a variable voltage (VDD_CPUX), the lower the voltage the lower the temperature (less problems with heat/overheating), the higher the voltage the higher the maximum CPU clockspeed. So the best idea is to adjust this dynamically (low voltage/clockspeed when idle and only increasing both when needed). There are 3 variants to implement this: not at all, primitive or advanced (using a voltage regulator that's able to adjust VDD_CPUX in 20mV steps)
    Only 3 devices implement no voltage regulation at all: Banana Pi M2+/EDU (frying the SoC constantly at 1.3V therefore prone to overheating), Beelink X2 (no idea) and NEO 2 (only 1.1V therefore limited to 1008MHz cpufreq max since above instabilities might occur).  Some boards use SY8106A I2C accessible voltage regulator where we can use fine grained voltage settings (Armbian fine-tuned these for every board so far to achieve max performance). This applies only to the following Xunlong boards: OPi PC, PC Plus, PC 2, PC 3, Plus, Plus 2 and Plus 2E. All other boards implement a simple two voltage scheme and are able to switch between 1.1V (up to 912MHz possible with H2+/H3 or 1008MHz with H5) or 1.3V (1.2GHz max with H2+/H3 and 1.25GHz with H5) And finally to add some stupid rankings: the cheapest board is from Xunlong (Orange Pi Zero: $7), the fastest is from Xunlong (Orange Pi PC 2 for $20) and the one with best feature set and onboard peripherals is also from Xunlong (Orange Pi Plus 2E: $35). And that's only due to OrangePi PC 3 Prime still not being released at the time of this writing (since otherwise regarding both performance and features this specific Xunlong board... )
    Hope that helps
    Edit: OPi 3 is now known as OPi Prime and (almost) nothing has changed compared to the leaked pictures back from last August.
  3. Like
    krasoffski reacted to tkaiser in Orange Pi Zero mainline broken?   
    Here it's about board bring up. Which voltage is the SoC being fed with if no software started to control anything? Zero can switch between 1.1V and 1.3V but there's a default state when powering up and control is later turned over to software which then decides based on clockspeed which voltage to chose (remaining at 1.1V with lower clockspeeds and switching to 1.3V when exceeding 816MHz -- mainline -- or 912MHz -- legacy)
    Edit: to elaborate on that a bit (and to stand corrected if necessary and also to add stuff to @zador.blood.stained's TODO list  ):
    So the board's VDD_CPUX voltage is 1.3V when the board boots regardless of clockspeed set anywhere (what's happening when a 'reboot' has been issued before?). This should prevent any undervoltage issues at boot. Next step is booting u-boot, here a static cpufreq will be set (either u-boot default, IIRC 1008 MHz with sun8i, or in our case it's 480 MHz -- we chose this value since we prefer minimum consumption over boot duration) Next step is booting the kernel, with H2+/H3 legacy kernel here the userspace governor is set so the board remains at 480 MHz clockspeed while with mainline kernel schedutil governor takes over and results are not really predictable, at least now CPU cores might clock up to the highest limit allowed with sun8i dev kernel -- what's the value? 1296 MHz? -- and since DVFS is now active voltage starts to jump between 1.1V and 1.3V based on cpufreq) As soon as cpufreq-utils daemon is started it looks really weird since now we switch with legacy kernel from userspace to interactive governor while with mainline from schedutil to ondemand (everything ok, interactive is the better choice with legacy and we switched just recently to ondemand due to some strange reports regarding schedutil). From now on maximum cpufreq can be set by the user (h3consumption tool or /etc/defaults/cpufrequtils -- defaults for large H3 boards 1296MHz, for smaller H3 boards 1200 MHz and for NanoPi NEO/Air 912 MHz) DVFS settings switch voltage above 816MHz with mainline (since that's defined in official DT for Orange Pi One which we use for the Zero) and above 912MHz with legacy since after some extensive internal testing we don't give a sh*t about Allwinner's recommended defaults and let a few thousand Armbian H2+/H3 board users do some reliability testing in the field  
    For me personally questions 1 and 3 are interesting. What happens at reboot and what's maximum cpufreq with sun8i dev kernel before cpufrequtils daemon is loaded?
    And then I would propose to switch to userspace governor in sun8i dev kernel config too for obvious reasons (we discussed this back then when developing IoT settings for the smaller H3 boards and my tests revealed that boot times for the large boards aren't affected that much since their boot clockspeed in N° 2 above is 1008MHz and not 480MHz). Also I would prefer to switch voltages with mainline kernel on the primitive H3/H2+ board also above 912MHz and not 816MHz any more to get more test feedback for mainline DVFS implementation. But no need to hurry, opinions first!