Booting Error
Back ground :
UP Board was located in inside of instrument. Thus we can’t push power button.
So, we change BIOS setting to boot automatically when 5V power applied.
But randomly UP board didn’t boot even if power was applied.
So, we have to disassemble an instrument and push power button for booting.
Trouble :
1)FAN & LED is normally workable but booting doesn’t proceed (NO BIOS logo) when input power.
2)If we push power button when 1), booting proceed normally.
3)After 2), there was no booting error anymore.
Test Condition:
1)BIOS Version : UPC1BMOT / UPC1BMOS / UPC1DM17
2)BIOS setting : Default (State After G3 [S0 State])
3)Test method : 5V Power ON/OFF in a cycle and then monitor booting process (25sec On / 2sec Off)
This is critical problem for the instrument manufacturer....
Comments
-
@Jim_Kang
Do you mean the UP doesn't start, booting when 5V power applied? And easy to reproduce the issue with power on/off test (25sec On / 2sec Off)?
Please provide us with the following information
1. Steps to reproduce this issue, or Video
2. UP board serial number -
Same issue here (UP 4Gb/32Gb, Ubilinux Kernel, Ubuntu 16.04, No graphical interface, No monitor, No keyboard). It is not easy to reproduce the error since it happens randomly every 100 boots (more or less). We guess it should be something related to the voltage profile during the power loss. When this happens you may cycle power as many times as you wish but until you do not push the power button again the UP does not come up.
Once booted manually everything is normal, bios configuration are correct (State After G3 [S0 State]).Since we have several of these devices (actually about 100) in embedded prototypes (which cannot be opened and manually restarted all the times) we hope there could be some setting to force Autostart at every power cycle. Could it be?
Thank you for your reply in advance
-
@Carlo Alberto
Give me a list of serial numbers, or any
I need to check which version of UP are you currently using -
MAC Address is 0007 3242 D69D
Firmware was updated to last available version at March 2019 -
This happened so far on at least 4 different models (The others are at customer site at the moment), during an
experimental phase of 36 prototypes booting on average 8 times daily since February 2019. It is a transportation
related application and power is attached to the engine so it can be loss at any time and reattached at any time.
Power was adapted from battery to 5V PS using a 50W TRACO Power DC-DC converter (in 9-48V out 5V).The system is fully embedded (no monitor, no disk, no keyboard) and having a readonly filesystem (overlayroot-fs)
we have no trace in the log of what happened before the event. The only log we collected was CPU temperature
which however was below 55°C.Half of the UP (60 of them from ) were bought early 2017 from Digimax, while a second half (about 50) were bought
from different Italian resellers (RS-Components, Digimax, Mouser). In a similar application, some of them (6) reboot
daily since April 2018 without any issue.If this behaviour is model related I would like to know that since we are planning a much larger exploitation and
we need to be sure (at least) that this issue is solved and it will not happen on next and on sale models. -
@Carlo Alberto
I can't say for sure if the UP is broken or components has failed.
However, for better troubleshooting, I would suggest you to contact your UP supplier for any RMA support -
My hypothesis is that there is no major failure but just some wrongly sized (capacitor) component who can erratically bring the system to erroneous shutdown when power is lost.
For sure this is not a specific damage since the same phenomenon happens on several devices (almost all of them installed from scratch with just purchased boards). The defect only appears once every several boots (about 100), and once these devices start they continue working regularly for long periods before showing the same fuzzy behaviour.
My point is that I have to decide if suggesting or not them into a specification document for a large production, if there is a contact point to contact before taking this decision I will be happy to do that.
-
After two more months of use, out of 30 new UP-BOARD systems other three devices stopped restarting automatically after power loss.
Definitely something happens so in a particular condition these device do not start automatically but require the user to
manually push the power button.The restart miss only happen once out of about hundreds restart, so it is not easy to diagnose which is the real reason for that.
The last devices were produced in late 2018 and had no fault on the hardware.
Yet I had no answer form last post.
Should this lack of response means that these devices cannot be considered to be reliable for embedded-industrial applications?
Should I look some different board?
-
Hello @Carlo Alberto . What were your conclusions about these boards? We have a product on an early stage with these board installed and we are interested on knowing if it would be better to change. The customer won't have access to the power button just like the problem you are describing but havent had the problem you are describing. So far we have only 5 equipments installed
-
Pitching in. Have eight up boards installed on remote location and they just randomly don't start or freeze. It's not often. Maybe one of the eight computers do this every two weeks. Since it's fully embedded the result is that the computer will be off for a full day until next power cycle.
We're running ubuntu 16 with overlay-fs
-
When using the lockdown function, there was a 20% chance of a startup failure when the power was turned on.
-
I had the same problem. In my case, it was a faulty/short-circuited ethernet cable
