Boot problems

[Jakub Bandola]

Hello,

 

sorry but my english is not best.

I tried to search forum for these problems, but I do not found the answers

First problem is that serial console do not work for me, just random symbols, I already tried 3 USB to RS converters with 3.3 V logic, hercules_3-2-8 and putty as SW, with multiple setting, mostly 115200, 8 data bits, no handshake, no parity.

Second problem the HDMI do not show useful data either, at least for me there was nothing I can understand as problem.

I have 2 orange Pi lite available, so it is probably not a board problem, I tried 2 SD cards and multiple images (desktop legacy kernel 3.4.y and mainline kernel 4.14.y from https://www.armbian.com/orange-pi-lite/) I tried multiple power supplies.

I record one try, can someone point me to what I can have wrong?

 

Thanks in advance for your answers.

[chwe]

Known working putty configuration (com port might differ), with a 2$ most likely fake FDTI from 'untrusted' sources.. 

 

https://docs.armbian.com/User-Guide_Getting-Started/

for being sure you do it right (e.g. check download authenticity, test your SD-Card first use etcher).

 

59 minutes ago, Jakub Bandola said:

I tried multiple power supplies.

You only need one, but it should be a good one (and a good cable too)...  Hard to say cause you didn't gave any hints (rule of thumb: if it's rated @2A but was bough for 2$ from aliexpress it might not be the one you look for. We collect such threads: https://forum.armbian.com/forum/31-sd-card-and-power-supply/

 

What you see is most likely that your kernel fails before it can properly initialize HDMI, without a working UART-USB dongle it's blind flying to help you (only donald trumps 'repeal and replace' approach is appropriate, means replace all 'known problematic parts' e.g. SD-card PSU and hope that it comes up).  

 

Describe your PSU, check your SD-Cards and try to bring up a serial console which works or start with donalds approach. 

 

p.s. orange Pi lite in your title would make sense.

[Jakub Bandola]

I am not home currently but I try to answer most question.

I use etcher to write and verify SD card, I also tried 2 other programs, but without success and etcher just look best.

For the PSUs hobby PSU from local store rated at 2.1A, tablet charger rated at 2A, Laboratory PSU rated at 30A, and mostly PC USB charging port rated at 3A by motherboard manufacturer, no cheap PSU from alliexpress.

I checked the output of these PSUs with oscilloscope to be sure the power is stable, and there are no drops in voltage.

I buy one SD card from the local store and the other is bit older (originally for a phone), but both are Class 10.

I try the configuration you recommended later today.

Thanks for your answer.

[Jakub Bandola]

I tried with same settings, but the result is same

[chwe]

on 3.3V with RX -->TX, TX --> RX and GND -->GND?

[Tido]

https://linux-sunxi.org/UART#Attaching_the_UART

[Jakub Bandola]

I know how to connect uart, in case of wrong connection there would be nothing coming at all, because if rx and gnd produce no data to read. I rechecked everything multiple times before I post there. These "random symbols appear at about same time as text on screen. 

[Igor]

On 5/29/2018 at 7:42 PM, Jakub Bandola said:

2 orange Pi lite available, so it is probably not a board problem, I tried 2 SD cards and multiple images

Strange. You probably did it already but what about powering from USB2/3 port on the computer where you are attaching serial console? To have the same GND. Or just power via OTG port with mUSB?

[Jakub Bandola]

mostly PC USB charging port rated at 3A by motherboard manufacturer, it is USB port on same computer as the serial console, so GND should be the same. Orange Pi Lite do not support powering from mUSB port, or at least it is written, I do not try that.

[zador.blood.stained]

On 5/30/2018 at 7:46 AM, Jakub Bandola said:

I checked the output of these PSUs with oscilloscope to be sure the power is stable, and there are no drops in voltage.

So if you do have an oscilloscope, can you capture the signal from UART TX (from the board side) and calculate its baud rate? (reference, just in case).

 

9 minutes ago, Jakub Bandola said:

mostly PC USB charging port rated at 3A by motherboard manufacturer,

But they may need a special condition (i.e. a specific resistance between the USB data lines) to activate the high current limit mode.

[Jakub Bandola]

I already tried to figure out the baud rate, but I was trying to identify whole byte, without success, I will try it with the shortest pulse method, that is simple and clever.

I am pretty sure that USB on my PC provide more than 2A without any special conditions, because it can charge tablet faster than 2A charger, over two wire cable.

And it behaved the same when connected to laboratory PSU with limit 30A.

[Jakub Bandola]

baudrate is 115200, based on shortest pulse method that show 8 us

[chwe]

10 hours ago, Jakub Bandola said:

I know how to connect uart, in case of wrong connection there would be nothing coming at all, because if rx and gnd produce no data to read.

Nothing against you..  it's just to kill potential variables of errors. Errors can happen even when they shouldn't (e.g. me being in a hassle connect a 5V analog signal to an ESP8266 ADC frying it and wondering why the shit doesn't work anymore  most pins on a wemos are 'fine' with 5V, expect the ADC... )

 

2 hours ago, Jakub Bandola said:

baudrate is 115200,

so either UART from both boards are defective (sounds unlikely, but you never know... ) or something is wrong with the USB-UART dongles you use..  Assuming that there's no 'different ground' error and no 'bad PSU' error. Did you test those UART dongles at the same baudrate for other use-cases and did they work smoothly then?

 

7 hours ago, Jakub Bandola said:

I am pretty sure that USB on my PC provide more than 2A without any special conditions, because it can charge tablet faster than 2A charger, over two wire cable.

And it behaved the same when connected to laboratory PSU with limit 30A.

Current is mostly irrelevant for those cheap SBCs (as long as it is enough.. ), voltage drop is a way more important (depending on PSU, cable and charing port, in case of microUSB those are mostly rated to 1.8A if they fulfill the official specs - for your charging example of a tablet). 

On 5/30/2018 at 6:46 AM, Jakub Bandola said:

I checked the output of these PSUs with oscilloscope to be sure the power is stable, and there are no drops in voltage.

Did you also check it on the SBC side (e.g. on pinheader) to be sure it is sufficient on this side too? There's no 'undervoltage protection' on those cheap boards. 

[Jakub Bandola]

I test just now these dongles against each other and it work.

I was checking for voltage drops/stability on the pin header of the SBC.