jock

 View Profile  See their activity

  • Posts

    852

  • Joined

  • Last visited

 Content Type 

Forums

Member Map

Store

Crowdfunding

Everything posted by jock

  1. jock
    Hello. I have limited experience with windows, there is indeed the AndroidTool, then renamed to RkDevTool, which can be downloaded from rockchip github tools repository. It is not hard to use, but does not come with a real user guide, just a series of buttons that do things and you need some background to understand what they really do. What you want to do (backup eMMC to SD, transfer backup to other similar box) is already covered by multitool, in fact the backup/restore paragraph you cite is for EXPERTS because involves entering maskrom mode, USB male-to-male cable, and some low level linux tools. Multitool is by far easier to use.
  2. jock
    Frame this post and pin it in the forum registration page! Moreover, ssv6x5x chips are working just on rockchip platforms, no mainline kernel, definitely if anyone gets an ssv6x5x is not lucky at all. I think the best wifi you can find in common tv boxes is one of the broadcom derivatives (the AmPak AP* series): slim and good opensource driver, good throughput, often working bluetooth, generally available firmwares. I don't have enough experience with realtek chips, but looking at their drivers... well, they are not really well engineered and the latest burst of discovered vulnerabilities is just a confirmation of bad source code. Then there are the mixed bags: esp8089 (very good opensource driver, simple design, only 802.11n and no bt though), allwinner XR series (weak and unstable drivers in the past, don't know now). But, as @hexdump said, it is a lottery: you won't know what you will get. One consideration can be helpful: getting a box with wifi + bt will exclude all the wifi-only chips (ssv6x5x, allwinner XR, esp8089), which increases the chances to get broadcom or realtek chips, which at least work well with mainline kernel.
  3. jock
    @Wester_Minskno, you misread the datasheet. Each chip is 2gbit, you have 8, then it's 16gbit / 8 = 2 gbyte, no way out. The specs given by the manufacturer are fake.
  4. jock
    Are you sure it is 4gb? If it detects 2Gb probably it is 2gb, and 4gb is a faking spec from the vendor. It's hard to tell how much memory you have, you have look into the datasheet for the memory chip and multiply for the number of chips.
  5. jock
    The answer to this question is yes, but it still requires some manual intervention because the bits to get hardware accelerated videos is getting merged into mainline kernel in these months. From the tests I did on rk3288 (not rk3318), it is working pretty well, but requires compiling ffmpeg and mpv by hand, plus ffmpeg source code has to be compiled with some more patches. A note: GPU is not involved in hardware video acceleration, so tinkering with the driver won't provide any dramatic performance improvement.
  6. jock
    it could be interesting to know what did you change to make it work
  7. jock
    Yes, the rk3318/rk3328 has a spare sdmmc controller, sometime it is left unused, sometime it is used for wifi or sdcard. In your case, it seems that it is used for wifi, and it is strange that if you enable the right device tree overlay it freezes your board.
  8. jock
    In the previous pages I posted some instructions on how to let it work. The driver for various led drivers is available on github and needs to be manually compiled. I did not include it yet because it has some limitations I would like to fix before integrating into. Overclock: bad idea on rk3318, but your mileage may vary Why do you want to remove the metal can? It is a very bad idea, it shields the wifi from interferences and if you remove it, probably your wifi/bt will have troubles. There are various instructions over the internet, but what are you expecting from the dts?
  9. jock
    Well, installing the wrong firmware on the wrong box is usually the best recipe to brick the box, as you experienced by yourself. Just in case someone else reads and wants to do that, it is something that I absolutely DO NOT recommend to do unless you exactly know what are you doing. It's not your case, I see you're fine because you got the eMMC clk pin to put the board in maskrom mode, but not all rk3318/rk3328 have accessible eMMC clock pin. Anyway thanks for the large list of rk3318 firmwares, it will be interesting to scavenge for firmwares and blobs
  10. jock
    From the dts you posted, the wifi chip is attached to the "alternate" sdio bus. From the dts also there is no evidence that your wifi chip is sv6051p, that is anyway not supported in mainline kernel. To understand why your box is hanging on reboot if you select the alternate sdio bus there is the need for a detailed log from the serial. edit: photos of the board and chips are welcome and interesting
  11. jock
    @RetroFan90 Thanks for the photos, firmwares and all the details! The HK1 Max (aka YX_RK3328 board) is already very well supported since it is the board I got here and it is the main developing asset I got. The H96 Max looks quite ordinary box, I took a look to the dtb and it seems pretty standard to me, should work fine out of the box. Do you have issues with some peripherals with the H96 Max?
  12. jock
    You can try the instructions for another board (xt-q8l-v10) I worked on: https://forum.armbian.com/topic/7141-csc-armbian-for-rk3288-tv-box-boards-q8/ If it works for you, probably the basic things should work ok (usb, hdmi, sdcard, emmc, ...) but other peripherals like wifi, bluetooth, etc... may not work out of the box. Maybe it can be a starting point for further studies. AFAIR the rikomagic was somehow famous some years ago, and maybe there is a device tree already available in the linux kernel, but I'm not sure.
  13. jock
    I understand. Well, if you don't have any experience with Linux, it will be hard to suggest you to do anything valuable with such box. Surely Armbian, at the current state, is probably not going to be a good experience with netflix, multimedia things are still trying to get sorted out. Maybe you would give a look to LibreELEC, if your purpose is mainly multimedia.
  14. jock
    Hello, you didn't say what you want to do with your box. If you want to restore android, this is totally the wrong place. If you want to install/try armbian (and wipe out android), the other thread you already found has clear and simple step-by-step instructions on what to do for that. ps: link to the photos are not working
  15. jock
    The alsamixer is right, the devices have no hardware controls, so volume must be controlled from the source player. A sound server like pulseaudio can do all the mixing and volume control though.
  16. jock
    @Charles Bauer Apparently you did not read the first page carefully: Unbricking may be complicated, because neither me nor @fabiobassahad the chance to tinker with a board with eMCP. The problem is most probably related to memory initialization. A serial log is required for confirmation, but we already have seen a situation like that and I don't think this is different. Memory initialization is the first thing that is going to be done during bootstrap and thus, when it goes bad, the board is knocked down and requires manual intervention to get into maskrom mode. Doing this job require some skills in electronic and some non-common equipment because you need to find and ground the eMMC (eMCP in this case) clock pin. What you can do to help development is send the bricked board to @fabiobassa for him to analyze
  17. jock
    Stick with the autodetected (ap6334). Just follow the instructions in this last post: https://forum.armbian.com/topic/17597-csc-armbian-for-rk3318-tv-box-boards/?do=findComment&comment=127240
  18. jock
    @curse I think I spot the issue: from dmesg I see that the brcmfmac driver is using the standard nvram file, but the standard nvram does not work because it is for BCM4334 and not for AP6334. This time I propose you to paste this file over /lib/firmware/brcm/brcmfmac4330-sdio.txt, reboot and try again. If the problem is that one, it is very strange because the driver is supposed to find automatically the right nvram file, it has always worked that way but for some reason this time does not.
  19. jock
    @curse I'm looking into the dtb and everything seems to be in place. I see that the dtb tells the wifi chip is ap6330, but rk3318-box detects ap6334. They are two different chips, could you please try to change wlan-ap6334 to wlan-ap6330 in /boot/armbianEnv.txt file? I see there are many clones here and there, maybe some cloned the wrong id. edit: photos of the board and logs (dmesg) are still particularly appreciated edit2: BTW the behaviour you are describing is often related to a wrong nvram. First restore wlan-ap6334 in /boot/armbianEnv.txt, then download this file and write over the existing /lib/firmware/brcm/brcmfmac4334-sdio.rockchip,rk3318-box.txt file. You may even try this other version to see if it works and/or has better performance.
  20. jock
    @curse Sorry for the late answer, but I just spot the post edit right now. wlan and bluetooth, despite being on the same chip, are connected to the SoC in different ways: wlan is connected via high-speed sdio bus, bluetooth via common UART; also they are physically different chip parts that just share some things (the radio part), so it may happen that one work and the other does not. Since rk3318-config correctly reports the right chip, it is attached to the right sdio bus. Now there there may be some board peculiarities that does not make it work. In the first post of this thread there are some good things that help in debugging, in particular if you can provide photos of the board and the original firmware or the original dtb I can inspect it and try solve the issue. If you can spot some marking/signatures on the board, you can see if there is a match in rk3318-conf when it asks for "led/gpio configuration". At the moment there are just two supported boards: YK_RK3328 (found on my HK1 Max) and X88_PRO_B ( @lucky62's box); maybe yours is a different one that require some minor adjustment to make wifi work.
  21. jock
    Yeah, you're right: the thread title is a bit misleading, because rk3318 and rk3328 are fundamentally the same chip. It is so because I have no rk3328 to work on, so can't guarantee and test anything on that. People reported that the images works as well on rk3328 boards, so you're invited to try and report if it works for you. There are good chances that the images works fine, and ap6330 is well supported in mainline kernel, including bluetooth!
  22. jock
    @RaptorSDS Thanks for the links! I will check ASAP. The board is giving me some stability issues and, among other things, the wifi and emmc are having troubles with mainline kernel. It looks like the pin configuration of the mmc controllers is somehow wrong, but need to check against the original dtb to be sure...
  23. jock
    Thanks a lot!! It looks like the board are exactly the same, probably they are clones of some sort: mine is labeled IPB900, yours T95N_RK3229. The external chassis has printed T95V Pro, fantasy names I can see that the components, power regulators and soldering pads (leds, serial, diodes, ...) are placed in the same position and the soldering pads. Can't remember if you already uploaded the original firmware or dtb, did you?
  24. jock
    Nope, we don't just need something that runs, we need the original firmware because only the original device tree can tell us the missing pieces
  25. jock
    Hello guys, this time @fabiobassa and me needs a bit of help We encountered the board you can see in the photo. It is from the Indian manufacturer AEMS and has the IPB900 marking on the PCB. We could not find the original firmware because it arrived with a badly flashed firmware. It looks like the board is a bit different than usual, so some things are not perfectly working and it is also overheating a bit: we thing we could arrange things a bit better for this board, but we need the original firmware or at least the original device tree. If anyone has this board and has the original firmware or a backup of the original firmware, it would be great if he/she could share to let us study it. The board seems to be one of the best in terms of performance for rk322x, so it is a pity if it could not be supported well enough. Thanks! ipb900_1 ipb900_2