Author Topic: Trying to troubleshoot a main board with Rockchip RK3288 microprocessor  (Read 744 times)

0 Members and 1 Guest are viewing this topic.

Offline Yamin

  • Regular Contributor
  • *
  • Posts: 114
  • Country: mv
Hi there everybody,
Sometime back I received a faulty guitar effects pedal board, it quite famous among the guitarist apparently. The unit does not boot up, goes through a start up sequence and just turns off. The main board is made up of Rockchip RK3288. Doing an internet search showed that a lot of people was going though the same situation. Nothing seems to get the unit to start back up - resetting etc.
The manufacture seems to not acknowledge this known issue and weren't very helpful. However I was able to source a spare board from the party which manufactures the main board.
Installing the new board rectified the problem, but after about three weeks the same thing happened. This time I was able to get it back to life by pressing the reset button, after couple of months it strikes again. This time however I was not able to get it back to life with the reset button. The guitarist it not really interested in reviving it anymore. So I thought I will just mess with just for my personal reference.
So as I said the microprocessor used Rockchip RK3288, the power management IC used is the ACT8846. I tried to find any obviously faulty components, couldn't find it. Going through the datasheet for the ACT8846 I found out that there was a way to manually reset the processor by shorting nPBIN to ground. I tried this but no difference. Start up sequence seems to work I even get 5V at the PWRHLD pin for a moment and it goes back to 0V. The nPBSTAT is used to fire an interrupt and it seems to be doing that as well. But the power doesn't latch. The button used to turn on the unit is one of those momentary push buttons which I assume fires some kind of an interrupt. If I press and hold this button, the voltage tends to remain at the pins for as long as I hold it.
I am quite new to troubleshooting circuits including microprocessors. Not having the schematic makes it that much harder. Just trying to 'solidly' figure out what the issue could be, even though I am not able to fix it. Any tips/help troubleshooting this would be greatly appreciated.
At this moment I am using the old mainboard with nothing else attached (buttons etc) only the 12V power supply.


Rockchip RK3288 datasheet -

http://webcache.googleusercontent.com/search?q=cache:http://opensource.rock-chips.com/images/4/4e/Rockchip_RK3288_Datasheet_V2.2-20170301.pdf

 

Offline coromonadalix

  • Super Contributor
  • ***
  • Posts: 2459
  • Country: ca
Re: Trying to troubleshoot a main board with Rockchip RK3288 microprocessor
« Reply #1 on: February 13, 2020, 06:15:28 pm »
Check the battery cell, on a different system with a dead battery it simply refused to boot ???

my 2 cents.


Is it booting thru an sd card ?  maybe the card is corrupted,  i have a board with an imx25 cpu with an winCE 6.0 OS   who did that symptom, the card had too many read and writes cycles ...  changed it for an sandisk brand  it was a kingston before, everything is now fine.
 
The following users thanked this post: Yamin

Offline Yamin

  • Regular Contributor
  • *
  • Posts: 114
  • Country: mv
Re: Trying to troubleshoot a main board with Rockchip RK3288 microprocessor
« Reply #2 on: February 14, 2020, 07:33:24 pm »
I did check the battery it seems to have charge.

There is no SD card, I had a similar case with another effects pedal where the SD card contact was a bit tarnished. I cleaned it and it started working.

Any suggestion on  reading materials for troubleshooting microprocessor circuitry?

Thanks so much for your help.
 

Online fzabkar

  • Frequent Contributor
  • **
  • Posts: 590
  • Country: au
Re: Trying to troubleshoot a main board with Rockchip RK3288 microprocessor
« Reply #3 on: February 15, 2020, 12:03:07 am »
I would measure the output voltages of each of the switchmode supplies and LDO regulators.

Given the symptoms, I'm wondering whether there is something wrong with the power-up sequence. The PMIC is programmed to power up each supply in a predefined order, so I'm thinking that the uP might get hung up if each supply doesn't come up at the right time.
 
The following users thanked this post: Yamin

Offline rsjsouza

  • Super Contributor
  • ***
  • Posts: 3875
  • Country: us
  • Eternally curious
    • Vbe - vídeo blog eletrônico
Re: Trying to troubleshoot a main board with Rockchip RK3288 microprocessor
« Reply #4 on: February 15, 2020, 02:50:48 am »
If this product is running embedded Linux and it is powered on and off with a mechanical switch, there is a chance the filesystem is getting corrupt (if it is poorly designed). I have seen this in certain products.
Vbe - vídeo blog eletrônico http://videos.vbeletronico.com

Oh, the "whys" of the datasheets... The information is there not to be an axiomatic truth, but instead each speck of data must be slowly inhaled while carefully performing a deep search inside oneself to find the true metaphysical sense...
 
The following users thanked this post: Yamin

Offline Yamin

  • Regular Contributor
  • *
  • Posts: 114
  • Country: mv
Re: Trying to troubleshoot a main board with Rockchip RK3288 microprocessor
« Reply #5 on: February 15, 2020, 09:32:25 am »
I would measure the output voltages of each of the switchmode supplies and LDO regulators.

Given the symptoms, I'm wondering whether there is something wrong with the power-up sequence. The PMIC is programmed to power up each supply in a predefined order, so I'm thinking that the uP might get hung up if each supply doesn't come up at the right time.

Probably a very newbie question, but what is LDO? I came across that on the datasheet as well.

Make sense with the point you made about the power up sequence - would there be any way to know what the power up sequence should be or is it only known for the manufactures?
 

Offline Yamin

  • Regular Contributor
  • *
  • Posts: 114
  • Country: mv
Re: Trying to troubleshoot a main board with Rockchip RK3288 microprocessor
« Reply #6 on: February 15, 2020, 09:41:40 am »
If this product is running embedded Linux and it is powered on and off with a mechanical switch, there is a chance the filesystem is getting corrupt (if it is poorly designed). I have seen this in certain products.
This was my thought as well, and that's what a lot of people are speculating for this particular unit. However would there be any way to conclusively find whether that is the problem.
 

Offline rsjsouza

  • Super Contributor
  • ***
  • Posts: 3875
  • Country: us
  • Eternally curious
    • Vbe - vídeo blog eletrônico
Re: Trying to troubleshoot a main board with Rockchip RK3288 microprocessor
« Reply #7 on: February 15, 2020, 11:29:01 am »
If this product is running embedded Linux and it is powered on and off with a mechanical switch, there is a chance the filesystem is getting corrupt (if it is poorly designed). I have seen this in certain products.
This was my thought as well, and that's what a lot of people are speculating for this particular unit. However would there be any way to conclusively find whether that is the problem.
In this case, I would try to find three pins that would correspond to a UART port (maybe marked on the silk as TX, RX and GND) and connect a serial cable to a terminal program on a computer, configuring it to 115200 baud, 8 bits, no parity, 1 stop bit and no handshake (these are the most common settings for such ports in the marketplace).

Turning on the equipment, you should see a MLO/U-boot prompt, which will follow a sequence of operations with diagnostic messages.

If they have this filesystem indeed in flash, they should have furnished the board with enough RAM to hold the filesystem there. U-boot allows uncompressing the kernel and filesystem to RAM, which has the benefit that any corruption caused by the mechanical on/off would be simply vanished.

Another issue I have seen is the simple corruption of an ailing Flash device. This can be caused by either the processor itself (I have seen earlier Linux-based processors have bad NAND or NOR Flash controllers) or the Flash IC itself (either a  ad devices or simply second-hand ones in the most inescrupulous manufacturers).
Vbe - vídeo blog eletrônico http://videos.vbeletronico.com

Oh, the "whys" of the datasheets... The information is there not to be an axiomatic truth, but instead each speck of data must be slowly inhaled while carefully performing a deep search inside oneself to find the true metaphysical sense...
 

Offline Haenk

  • Regular Contributor
  • *
  • Posts: 157
  • Country: de
Re: Trying to troubleshoot a main board with Rockchip RK3288 microprocessor
« Reply #8 on: February 15, 2020, 01:04:50 pm »
In this case, I would try to find three pins that would correspond to a UART port (maybe marked on the silk as TX, RX and GND) and connect a serial cable to a terminal program on a computer, configuring it to 115200 baud, 8 bits, no parity, 1 stop bit and no handshake (these are the most common settings for such ports in the marketplace).

There *is* a debug port, even noted as such and Comes with a header - even though it has four pins, that might be the way to go, probably just one pin not connected...
 

Offline rsjsouza

  • Super Contributor
  • ***
  • Posts: 3875
  • Country: us
  • Eternally curious
    • Vbe - vídeo blog eletrônico
Re: Trying to troubleshoot a main board with Rockchip RK3288 microprocessor
« Reply #9 on: February 15, 2020, 01:18:06 pm »
In this case, I would try to find three pins that would correspond to a UART port (maybe marked on the silk as TX, RX and GND) and connect a serial cable to a terminal program on a computer, configuring it to 115200 baud, 8 bits, no parity, 1 stop bit and no handshake (these are the most common settings for such ports in the marketplace).

There *is* a debug port, even noted as such and Comes with a header - even though it has four pins, that might be the way to go, probably just one pin not connected...
Yeah, I just saw that on the photograph. Although four pins could also be JTAG (TCK, TDI, TDO, TMS, with GND being somewhere else). Check the connections with a continuity tester.
Vbe - vídeo blog eletrônico http://videos.vbeletronico.com

Oh, the "whys" of the datasheets... The information is there not to be an axiomatic truth, but instead each speck of data must be slowly inhaled while carefully performing a deep search inside oneself to find the true metaphysical sense...
 

Offline coromonadalix

  • Super Contributor
  • ***
  • Posts: 2459
  • Country: ca
Re: Trying to troubleshoot a main board with Rockchip RK3288 microprocessor
« Reply #10 on: February 15, 2020, 03:26:10 pm »
In this case, I would try to find three pins that would correspond to a UART port (maybe marked on the silk as TX, RX and GND) and connect a serial cable to a terminal program on a computer, configuring it to 115200 baud, 8 bits, no parity, 1 stop bit and no handshake (these are the most common settings for such ports in the marketplace).

There *is* a debug port, even noted as such and Comes with a header - even though it has four pins, that might be the way to go, probably just one pin not connected...


And the 5 pins white connector near the 4 pins header ??
Yeah, I just saw that on the photograph. Although four pins could also be JTAG (TCK, TDI, TDO, TMS, with GND being somewhere else). Check the connections with a continuity tester.

Could be an serial or jtag port, as written  check for a ground pin and an 5v or 3.3vdc  pin .... and do a continuity check to the voltages regulators  once you find the main ones ...

And the 5 pins white header near the 4 pins  ??? what does it do ?
« Last Edit: February 15, 2020, 03:28:46 pm by coromonadalix »
 

Offline Haenk

  • Regular Contributor
  • *
  • Posts: 157
  • Country: de
Re: Trying to troubleshoot a main board with Rockchip RK3288 microprocessor
« Reply #11 on: February 15, 2020, 05:35:11 pm »
And the 5 pins white header near the 4 pins  ??? what does it do ?

My guesstimate is "USB header" - a better picture would help (it's stated on the PCB, but only "SB" is readable).
 

Online fzabkar

  • Frequent Contributor
  • **
  • Posts: 590
  • Country: au
Re: Trying to troubleshoot a main board with Rockchip RK3288 microprocessor
« Reply #12 on: February 15, 2020, 07:57:44 pm »
LDO = low dropout, used to describe a linear regulator which can work in environments where there is a low differential between input and output.
 

Offline Yamin

  • Regular Contributor
  • *
  • Posts: 114
  • Country: mv
Re: Trying to troubleshoot a main board with Rockchip RK3288 microprocessor
« Reply #13 on: February 15, 2020, 11:25:49 pm »
Wow, thanks so much guys! Its getting very exciting. I shall take a picture first thing in the morning when I get into the shop and upload it here.

 

Offline Yamin

  • Regular Contributor
  • *
  • Posts: 114
  • Country: mv
Re: Trying to troubleshoot a main board with Rockchip RK3288 microprocessor
« Reply #14 on: February 16, 2020, 07:42:41 pm »
Here are some close up photos of the connectors.
 

Online fzabkar

  • Frequent Contributor
  • **
  • Posts: 590
  • Country: au
Re: Trying to troubleshoot a main board with Rockchip RK3288 microprocessor
« Reply #15 on: February 17, 2020, 12:57:26 am »
I'm just curious. What are the complete markings on the ACT8846 chip? Is there a "490" marking anywhere?
 

Offline Yamin

  • Regular Contributor
  • *
  • Posts: 114
  • Country: mv
Re: Trying to troubleshoot a main board with Rockchip RK3288 microprocessor
« Reply #16 on: February 17, 2020, 03:15:23 pm »
I'm just curious. What are the complete markings on the ACT8846 chip? Is there a "490" marking anywhere?
There is no marking "490" may I ask what that means?
 

Online fzabkar

  • Frequent Contributor
  • **
  • Posts: 590
  • Country: au
Re: Trying to troubleshoot a main board with Rockchip RK3288 microprocessor
« Reply #17 on: February 17, 2020, 07:04:54 pm »
There are different versions of the ACT8846 which are preprogrammed with different voltages and different power-up sequences for each specific SoC. The RK3288 requires the ACT8846QM490 version (see pages 3 & 34 of datasheet).
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf