CSC Armbian for RK322x TV box boards

[nokirunner]

well .. my journey was intense but short, with your advice I managed to backup the rom, using the method suggested by @MFaust
so sure that everything went smoothly, I started armbian from sdcard and I tried an installation on mmcblk2 ... I restarted and end of the race, the device does not start anymore, neither from nand that from sdcard.

the strange thing is that it fooled me is that since the sdcard turned out to be like mmcblk1  and  the internal memory as mmcblk2  I didn't think it would even ruin the boot from sdcard.

after which I spent all afternoon trying to debrick the device via mask rom, unfortunately my device is quite unknown and there are no examples on how it can be restored ... and those that seemed more similar to me, trying to create a short circuit between clk and grd on the nand memory, they did not work.

my tv box is an  andowl q-m6 

Spoiler

 

 

 

the nand memory is a 16gb
Samsung 641
KVMTU000LM-B503
GS062C9Z8EM

 

I searched the in the internet, but I couldn't find the clk pin of this memory ..


looking better is an identical to this:
https://forum.freaktab.com/forum/tv-player-support/rockchip-based-tv-players/rk3229-devices/750159-scishion-model-x-android-8-1-rk-3229-2-16gb-5g-wifi-fast-lan

https://forum.freaktab.com/forum/tv-player-support/rockchip-based-tv-players/rk3229-devices/firmware-roms-tools-bg/754403-need-to-identify-my-android-box-and-find-firmware

[jock]

@nokirunner

I'm really sorry this happened

Everything seems odd to me, beginning from the unusual USB device ID, sdcard allocated to mmcblk1 (it should be mmcblk0) and now this unexpected brick.

 

1. Did you use the images provided by @Maker39 or those provided by me on first page? If you used one from Maker39, try to burn an image provided by me, put the sdcard into and boot again.

2. Did you run armbian-config and followed the instructions to install in internal eMMC or you did install in another way?

3. Did you run the hexdump command I suggested you in the previous post (here) ? It was useful to understand if you really got an rk3229, because that USB device ID is suspect.

[nokirunner]

23 minutes ago, jock said:

@nokirunner

I'm really sorry this happened

Everything seems odd to me, beginning from the unusual USB device ID, sdcard allocated to mmcblk1 (it should be mmcblk0) and now this unexpected brick.

maybe it was mmcblk0, I don't remember well

I installed from armbian-config
I prepared the rom from your first page with the method of starting directly from sdcard suggested in the comments below.

 I had performed the hexdump before launching the installation but nothing had come out, it could not find the path of the file, there were subfolders, I promised myself to investigate better after the installation and instead I found the surprise.

what do you think prevents the sdcard from starting now?

@jock
I got an idea, I have another identical engine, which I have for the other TV ... I'm thinking about it now, I try to start armbian on sdcard on that other machine and I tell you all the precise information, (of course I don't install anything on that tvbox)
tell me if you need other precise information

[jock]

@nokirunner

 

I got an identical board to yours here, thanks to @fabiobassa I can show you the eMMC clock pin. There is a small white circle in the upper left angle of the eMMC chips, that's the pin 1 of the eMMC. The eMMC clock pin is the 7th starting from the top on the other side.

Now in this photo @fabiobassa soldered it for me, because he is much more experienced.

 

Before proceeding any further please answer the question on the previous post and burn an image provided in the first page of this thread on your sdcard and check if it boots.

 

If does not boot remove the power, and try to short the clock pin using a cross-headed screwdriver: push the head of the screwdriver with a bit of force (but not too much) between the 7th and 8th pins to connect the two pins. While you're shorting the two pins, attach the power cord, then remove the screwdriver.

If the procedure went well, the red led should become very bright after a couple of seconds and after 15 seconds the blue led should start blinking. If this does not happen, power off the board and try again.

I just tested this on my board and it works.

 

[fabiobassa]

@nokirunner

Since this is the very first board on which me and @jock worked , we know it veeeeeeryyy well

you will unbrick, you bricked because you flashed a idbloader for ddr3 and this board has ddr2, that is all

PLEASE do not MIX methods unless you know what you doing... hte booting methods AREN'T compatible .

IF you use jock's use jock's if you use hexdump use hexdump ;-)

[nokirunner]

fantastico...  
I presume by the name that you are Italian  aren't you?
Well thank you ... now I see to resurrect ..
from stables to stars alle stalle  alle stelle again

the fact is that this is also the first arm board for me .. and i didn't know that i had to be more careful ... the world of x86 and  android recovery of the phones got me used to it

[fabiobassa]

@nokirunner
you very welcome and glad to help you and happy the board is back to life

Yes, I am italian and translate of "from stables to stars"  is Dalle stalle  alle stelle , just a missing D and ty for this italian quote 

This is one of the worst nightmare in bringing to linux  those tvboxes , sooo many combinations bootloader ddr and real processor (3228a 3228b 3229). Proprietary things work well, mainline adapted things work well the same, mixting thing up to today is a very talented skill, but luckly  we have now a growing statistic of working boards

[jock]

@nokirunner

Very glad to hear you brought the board to life again

 

 

[Ruan Marins]

Isn't there a Desktop version to the 5.5.16 linux kernel? Why?

[nokirunner]

@jock @fabiobassa

now i am in mask rom, but when i try to flash the backups i had done it tells me that matck device type failed ... so i am stopped .. i am not sure what i have to do from now ..

[fabiobassa]

mmmm can you give me back again the usb pid and vid ?

and then you should tick even parameters.txt box

anyway original firmware should be this:

rk3229_8.1_MX1VR_SSV6256_zhongx_20180925_v002_Spectek.zip

[jock]

@nokirunner sorry but I don't have ever used that tool so can't be of any help.

[nokirunner]

@jock
apparently I can't flash from windows with that tool, it was the tool I used to make backups, and for some reason even if it see the mask rom mode it doesn't work as it should,  (I saw some demonstration videos) I'm afraid I did some hardware damage yesterday making random attempts to find the clock pin ... now I'm trying from linux with your instructions ....
from linux the first thing i noticed, is that now the usb port is seen as 2207: 320b and it is like all the other rk2329 ... while before it was seen as 2207: 0002 ... now i am crossing my fingers hoping to have more luck ...

[jock]

3 minutes ago, nokirunner said:

@jock
apparently I can't flash from windows with that tool, it was the tool I used to make backups, and for some reason even if it see the mask rom mode it doesn't work as it should,  (I saw some demonstration videos) I'm afraid I did some hardware damage yesterday making random attempts to find the clock pin ... now I'm trying from linux with your instructions ....
from linux the first thing i noticed, is that now the usb port is seen as 2207: 320b and it is like all the other rk2329 ... while before it was seen as 2207: 0002 ... now i am crossing my fingers hoping to have more luck ...

 

I hope you didn't damage the eMMC. It could be that the eMMC is not available because you shorted the eMMC clock pin at boot. Shorting the eMMC to ground makes it unavailable to first stage bootloader.

Keep us informed about progress.

[nokirunner]

@jock

./rkdeveloptool ef
Creating Comm Object failed!


./rkdeveloptool wl 0x0  Armbian_20.05.0-trunk_Rk322x-box_focal_legacy_4.4.194_desktop.img 
Write LBA quit, creating comm object failed!

my suspicions are at risk of being confirmed, I think I have done some damage

what do you say or do you think it remains to do ??

 

[jock]

14 minutes ago, nokirunner said:

@jock

./rkdeveloptool ef
Creating Comm Object failed!


./rkdeveloptool wl 0x0  Armbian_20.05.0-trunk_Rk322x-box_focal_legacy_4.4.194_desktop.img 
Write LBA quit, creating comm object failed!

my suspicions are at risk of being confirmed, I think I have done some damage

what do you say or do you think it remains to do ??

 

 

in maskrom mode, before doing executing any command, you must upload the loader. rkdeveloptool db command may come in handy, you can see how to use it on the first page in the "Restore the firmware" section.

 

But I would say it is far easier to just boot into linux. Just plug the sdcard into and boot with the eMMC shorted. Remember to unshort as soon as you see the red led.

 

 

 

 

[fabiobassa]

@nokirunner

As Jock said you cannot erase something UNKNOWN to the processor 

 

The very early stage at the point where YOU are  is just load in ram at least a idbloader ( consider it as a bios) and this is done either in Linux and in Windows.

The very first box of rkdeveloptool under Windows in fact is loader, then comes parameters.txt

But probabily you haven't this loader usually called miniloader or rk322x_loder_2.47 or maybe 2.51.

And again if you flash the WRONG loader maybe made for DDR3 and you have DDR2 you will brick again the box

So the easiest way at this moment is 

1) reflash the whole android ( I gave you the rom name)

2) boot from a sd ( it will boot for sure  NOW!!!) and let have a look inside

 

[nokirunner]

@jock @fabiobassa Success!
First of all I want to thank the two of you guys, and all the others too ..
without you, I would not have experienced this new adventure, I certainly would not have got bogged down and I would not have come out of the sand afterwards .. certainly I have acquired a lot of important experience.
now I will tell you how I managed to resurrect something apparently dead ..
 You already know the first part of the story ..
I was about to give up, until I read the last suggestion of fabiobassa, to flash the original rom ... which in reality I had already downloaded and tried to flash without success with the first tool on windows, the rkandroid tool ... then going to check if the rom I had already found and downloaded was the same as the one suggested, I realized that inside the rar file there was another tool called factoryTool ... so, since I had tried them I have tried all this too .. this tool has the particularity of being a little more aggressive, it mounts and unmount the USB port where the device is attached until it can upload the various binary files and every time it used "wait mask rom" and I had to unplug the usb port and redo the clock short operation and I noticed that some times it failed and a couple of times for the first two small binary files, now I don't remember well it did "the ibl" and the loader I don't know .. but then the flash attempts kept failing, and asked me to unplug and replug the usb and do the mask mode ... until seeing that something was progressing, I decided not to un-plug on the usb 3 port that I had used so far, and to plug on the usb 2 port, and just I inserted the plug in usb 2 the rom flash started and it ended successfully ..

Now what is the synthesis of all this? that after all the failures that I have had so far, just that it showed no signs of life beyond the mask mode, trying with this tool and changing using a usb 2 port I got immediate success, now I have no idea if the problem was due to the usb 3 or some "out of the ordinary" problems that really didn't allow me to restore the device had emerged ..
Of course I was lucky and unfortunate at the same time, not that it was too serious .. in the end I would have lost 25 buks was something that I could let go ..

the master of the tools that saved me

[fabiobassa]

@nokirunner
well I never experimented programming throught a USB 3 port so I have no feedback, but I can say for sure that UNLESS physical damages those boards are quite unrickable.
The tool you have is similar to rkdeveloptool, even rkdevelop tool mounts and remounts usb but ONLY because once the usb is in mask rom and then after it goes in loader mode so physically it must unmount and remount usb
BTW you can try this tool , yours is a bit old


@Ruan Marins
studies on main kernel are still WIP so we just stay on what is well known and half-tested and half-finished
But you are absolutely welcome to compile it by yourself and share with us your impressions and opinions 
 

AndroidTool_Release_v2.67.zip

[nokirunner]

@fabiobassa
Probably it was usb 3, i don't know, for me this whole process was relatively new ..
The strange thing to say, that the first backup I had done without any problems on a usb 3, so this is triggered me, it was also quite fortuitous to guess to try a usb 2.

Back to normality , I can't activate the wi-fi on armbian, I removed it from the black list, and it shows the 2 famous wlan0 and p2p0 not activated
Is there a possibility to bypass this problem or do i have to give up for the internal wifi  and using a usb wi-fi ??

Do you think this problem is permanent, or is there a good chance that it will be fixed in future driver releases fot this wi-fi?

[Maker39]

@nokirunnerThere was a warning about USB3 in some instructions on flashing devices on RKxxxx. Highly recommended to use USB2.

 

 

Edited April 12, 2020 by Maker39

[Maker39]

PC_Firmware_upgrade.pdf

[jock]

@nokirunner @Maker39 As far as I know, the only programming USB port should be the OTG port. Sometimes it is labeled on the PCB, sometimes the enclosure has the OTG label, sometimes there is nothing telling you which port is the OTG one.

 

For example, the Scishion Model X (the same owned by @nokirunner) I got here has 2 USB ports, none of them is labeled OTG in any form and they are both labeled as "USB" on the chassis without any number on them.

 

Other devices, like the MXQ Pro 4K has the OTG signature on the PCB corresponding to the USB 4 port labeled on the chassis.

 

So it's a bit of hit and miss, but finding first the OTG port looking at the board can avoid many headaches later

[Maker39]

I , and probably @nokirunner , meant the version of USB3.0 and USB2.0 on the PC, but not the port on the board

As far as I know, there is only one OTG (on those boards that I have seen there are MOSFETs nearby) .

[jock]

1 hour ago, Maker39 said:

I , and probably @nokirunner , meant the version of USB3.0 and USB2.0 on the PC, but not the port on the board

As far as I know, there is only one OTG (on those boards that I have seen there are MOSFETs nearby) .

Yep I totally misunderstood

[fabiobassa]

@nokirunner

please when you report bugs don't forget ARMBIAN is just a way to define a rootfs of a linux kernel, an operating system but not the hardware specs.
Basically it means quite nothing in terms of hardware spec ; the relevant infos of course are
uname -r ( kernel version  this because of loading right modules and kernel objects *.ko)
lsusb
lshw 
when those utilities are installed

Anyway, going straight to your problem, I remember that board has dual band wifi chip and we have ONLY drivers and firmware for the ssv6051 and your board is using ssv6256

No kernel sources for that wifi chip on legacy kernel 4.4.xxx and even less support on main kernel ( as far as I know and until today but I guess we will never have )
So maybe the sv6051 driver in some manner recognizes the wifi chip but then no pairing occurs.
And even the working chip ssv6051 is so CRAPPY that I prefer plug in a few dollars usb wifi adapter if I really need wifi

Thank you so much for intensive testing , you confirm our sensations and as you said dalle stalle alIe stelle with this tvbox
I have the same board and I use for server applications and I am quite happy with it
 

[nokirunner]

I answer a little to everyone to simplify ..
Yes I was referring to the usb 3 of the pc,
It is a pity that there are still companies reluctant to release the open source drivers and firmware of some hardware, yet they would all benefit us, the users who would have well-functioning hardware, and the farm maintainers who would have less time and money to invest to debug and operate at best their hardware ...
stupidity take me away. 😁

my only desire to have this tvbox operating with armbian is to set up an octoprint server so that I can remotely control my 3d printer while printing, so it is essential for me that wifi works, I only have two free usb ports , and in one I want to put a webcam and the other is connected to the 3d printer ...
it is essential for me that it is always on, that  consume very little energy and so on ...
 

[jock]

13 hours ago, nokirunner said:

I answer a little to everyone to simplify ..
Yes I was referring to the usb 3 of the pc,
It is a pity that there are still companies reluctant to release the open source drivers and firmware of some hardware, yet they would all benefit us, the users who would have well-functioning hardware, and the farm maintainers who would have less time and money to invest to debug and operate at best their hardware ...
stupidity take me away. 😁

my only desire to have this tvbox operating with armbian is to set up an octoprint server so that I can remotely control my 3d printer while printing, so it is essential for me that wifi works, I only have two free usb ports , and in one I want to put a webcam and the other is connected to the 3d printer ...
it is essential for me that it is always on, that  consume very little energy and so on ...
 

Seems a pretty interesting project. Good luck for that!

 

Unfortunately the keep costs down, tv box makers use chips from unknown and unreliable sources, then happens that the vendor failed (at least rumors say so) leaving us with a very badly written driver. That is one of the reasons why I avoid cheap tv boxes when doing serious pet projects, a proper single-board-computer is much more suited.

 

As @fabiobassa said, ssv6051p driver is suitable for ssv6256p also, but for some reason it does not work.

[kruzer]

On 2/24/2020 at 9:36 PM, hexdump said:

@Alessandro - you may try the following on a linux system, assuming your sdcard is in /dev/mysdcard (whichever device this might be in your case):

"zcat working Librelec-xyz.img.gz | dd of=/dev/mysdcard bs=512 count=32768 status=progress"

then use "fdisk /dev/mysdcard" to create a partition starting at sector 32768 (something like o, n, p, 1, 32768, enter, w, q - might be wrong - all off mind) and then dump the armbian image to it:

"xzcat Armbian-xyz.img.xz | dd of=/dev/mysdcard bs=512 skip=8192 seek=32768 status=progress"

...

 

Thank you guys for your work, ideas and instructions,

Just wanted to report, that using this steps i managed to successfully run Armbian on my "HK1 mini" box (RK3299 chip 2GB ram)

 

[nokirunner]

2 hours ago, jock said:

Unfortunately the keep costs down, tv box makers use chips from unknown and unreliable sources, then happens that the vendor failed (at least rumors say so) leaving us with a very badly written driver. That is one of the reasons why I avoid cheap tv boxes when doing serious pet projects, a proper single-board-computer is much more suited.

I can understand this reasoning ...
But this makes me believe more firmly about the validity of the open source, if you have hardware working with opensoruce, even if the owner company is no longer available for any reason, the hardware always has the opportunity to improve ... I have become from my experience more convinced that it is better to invest in open source, I mean both in time and in money, especially if you are an end user, my experience is in the field of 3D in general, years ago there were many powerful and also expensive 3D applications, and people have invested a lot of time to grow on these applications .. over the years these have either failed or acquired from other competing houses and have been suppressed, some with very large communities that used that software, overnight the people who they invested in time and money on their growth around that software they found themselves "homeless".

Fortunately today things are starting to change, because people are now accustomed to the traps of closed source .. I give a very obvious example, which I know well because of my field, and it is that of Blender , which in recent years, through online donations has grown a lot, because people, but also companies have understood that it is much more convenient to pay engineers who develop on opensoruce ...

Moreover I am particularly enthusiastic because significant investments are also coming on the open-light hardware architecture ... in particular I am referring to the Risc - V ...
the future will be interesting ..

I apologize for the offtopic, now i return in the ranks.

I expect in the future, projects with single board computer open source architecture, developed by the communities, where then the various farms print and distribute the hardware ... in the wake of how it happened with arduino and the whole world of electronics projects that are evolved more or less ..