Two BeagleBone black boards with 1GHZ processors do not make boot sequece after some unsucceseful attemts to format eMMC.
It looks like they do not make PORz sequence. Both ocillators inacvtive. NRESET_INOUT is low. PORZn input high. All power voltages are OK.
I use simple code wich make BBB to act as USB flash drive to monitor eMMC with my PC. Code based on TI usb_dev_msc example in StarterWare.
It begin happens only with new boards. Old revision boards works OK.
Any body can help ?
I going to use JTAG to debug this processors but not shure that JTAG will work in this case.
If the BBB oscillator is not running, then I suspect the processor is bad. If you have two boards with the same issue, then I suspect the power supply you are using is bad. One of the big issues with the BBB is it doesn’t like to power down. Look at sheet 2 of the OSD335x schematics
You will see some protection circuitry at the bottom center of the sheet. This ensures the voltage difference between the 3v3 and 1v8 is never greater than 2V. The BBB does not have any such circuitry so when the 5V drops quickly, it is possible for the 1v8 to drop quicker than the 3v3 and hence killing the processor. This happened to me and I had to replace the processor.
The Beagleboard-x15 does have the same protection circuitry (BeagleBoard-X15_RevA2.pdf sheet 4).
Perhaps. But how it is connected with the recording in the flash memory. eMMC powered from different supply VDD_3V3B. According to documentation it takes about 50mA in write mode. So why VDD_1V8 drops down ? Bad 5V power from USB ? I will examine this issue. But in all other cases BBB works well. I have to find a workarownd.
Externally it looks like some exception or power fail after some time of extensive write to eMMC. I should kill some more BBB to find an answer
This is strange. I have several hundreds of BBB-based devices out there that do not have any of these protections. None of them died because of such a problem.
What does “old” and “new” mean - which revision? I know the BeagleBone Green no longer uses a FAT32 partition with a MLO-file but the bootloader-code located in or behind the MBR. May be this is the case on some BBB’s too and you simply killed all software?
I know about boot. But both oscillators is not woking !!! And processor stopped in PORz sequence. I can change boot sequence to boot from UART but processor should leave PORz sequence at first and remove reset from RESET pin.
On Fri, 28 Oct 2016 14:03:51 +0300, ??? ???
<nkurochkin@gmail.com> declaimed the
following:
"New" is revision C with 1GHz processor. I have some of them original and
embest clone. "Old" with 800MHz processor.
I was under the impression that ALL /black/ had 1GHz processors.
All "official" black's had 1GHz "tested" processors..
Even if they didn't have TI's efuse "set"... They had been speed bin'ed...
from our patchset:
/*
* All PG 2.0 silicon may not support 1GHz but some of the early
* BeagleBone Blacks have PG 2.0 silicon which is guaranteed
* to support 1GHz OPP so enable it for PG 2.0 on this board.
*/
It was the original /white/ that had the slower clock speed.
You will see some protection circuitry at the bottom center of the sheet. This ensures the voltage difference between the 3v3 and 1v8 is never greater than 2V. The BBB does not have any such circuitry so when the 5V drops quickly, it is possible for the 1v8 to drop quicker than the 3v3 and hence killing the processor.
This is strange. I have several hundreds of BBB-based devices out there that do not have any of these protections. None of them died because of such a problem.
It is just a matter of time. I had a A5A that I power up and down for years and then the processor pop’d. While investigating the problem, Gerald and I had some discussions and he pointed me to the circuitry that was added in the OSD335x and BeagleBoard-xM boards. I replace the processor and everything is working just fine. Gerald’s advise was to never power down the board. Even powering down the board with the power button won’t prevent this problem.
I would think BBB does not work because the oscillator is not running. check for power suuplies . If you have a logic probe(O’scope would be better) check oscillator output for frequency and amplitude close to its power supply. Also check reset line(s) to be inactive, not in active reset .