I have a device( Linux embeded) that ,before it finishes booting process, restarts( and over and over again).
The device uses several voltage levels( for CPU, flash and so on ).It uses several step-down controllers for that. I suspect one of the voltage is faulty and as a result the device restarts.
I would need to find out which voltage is faulty that is which voltage has a very short drops. I tried 200 Mhz oscilloscope but did not find any changes. Does it mean the oscilloscope is rather slow to show such a voltage change? Or are there not any drops and the problem is somewhere else?
Thank you for your feedbacks.
Have you checked the reset pin? It sounds like there might be a problem.
It could mean there's no voltage problem. It could also mean that you didn't set up triggering properly to catch the event. I doubt a 200 MHz scope is too slow to catch it, with adequate decoupling the system should be able to ride through any voltage issues that narrow.
You have a symptom defined, reboots continuously. That does not automatically point to having a bad or wrong DC voltage source. You certainly should check them out with a good digital volt meter, but best to broaden your horizon on what the ultimate fix turns out to be. See if you can gain further hints in your symptom analysis.
When you say, before it finishes the boot process, do you mean it boots into Linux, and then resets? What is the furthest stage it gets to (booting a bootloader, booting the kernel,etc)?
Anything in the linux logs?
The frequency between rebooting is important as that can quickly indicate if it's software or hardware induced and where in the boot process it occurs.
Thank you all for the reply. It starts only the first stage ( during booting) and when it should start the seconds stage, it starts again the first stage( from the beginning).
@Shock:Can you please explain further how the frequency between rebooting can indicate if it is software or hardware problem? The frequency of my rebooting is about 1 second.
Thank you for help
It would help if you post which device it is.
E.g. Raspberry Pis can have problems if the PSU is too weak and you have USB peripherals connected - it starts booting, then turns USB on, the extra current draw causes the PSU voltage to sag and the whole thing resets, restarting the cycle again.
Can you flash it with a different firmware? Or load your firmware into a different unit?
Is there any wrong with the Mhz oscilloscope?
Some advices are really useful.
Thanks for your feedbacks
The device is a linux embeded satellite receiver.
I can not flash the device because only the first stage is executed, so no possibility to flash the device.
I use a tested power supply adaptor , so the fault is not in that adaptor.
I use the mainboard( with CPU, RAM, flash) only no other parts so the problem must be caused by an item on the mainboard. But which one ?
Thanks for your feedbacks
The device is a linux embeded satellite receiver.
I can not flash the device because only the first stage is executed, so no possibility to flash the device.
I use a tested power supply adaptor , so the fault is not in that adaptor.
I use the mainboard( with CPU, RAM, flash) only no other parts so the problem must be caused by an item on the mainboard. But which one ?
Well, that is impossible to say like this, without seeing the board, knowing which MCU/SoC are we talking about and ideally the schematics.
What you are seeing could be caused by many things (bad flash, corrupted flash image, short on the board, bad cap somewhere, misconfigured jumper/dip switch holding the CPU in the bootloader/service mode/some weird mode, cold joint somewhere, dead part ...), not only bad voltage. In fact, if some of the CPU power rails was wrong then it likely wouldn't even attempt to boot.
In short, we need more information to be able to help you.