Hello!
Sometimes my BBB Rev C dont boot, when I connect the 5V.
The Power LED is on but the 4 USR LED are off.
This happen not really often … So I have no debug information.
But maybe someone has the same problem and know where the problem is?
Thank you!
As posted here a few times.
If your power supply does not have a fast enough rise time getting up to
voltage
your BBB will not boot reliable.
You have not said what specs your power supply is or where you got it from.
I have the same problem.
I often connect BBB with a PCB as power source and don’t have such problem. But when I power it with USB of my MAC, it does that some time. (I also attach a WiFi USB dongle to the BBB).
Powered by "usb" with a "usb" dongled plugged = no power left for the cpu...
Regards,
Powered by “usb” with a “usb” dongled plugged = no power left for the cpu…
Regards,
I have found that you can power by USB and use ethernet, but that is it. So when we tried powering by USB, and use ethernet + CAN . . . the board sometimes works, the board sometimes does not work. The take away from this is that if you power by USB, you can maybe power one peripheral / IO interface, but when trying to power more than one peripheral / IO interface. You should use the barrel jack, with a 1-2A good power supply.
Also, when flashing the eMMC, it is also recommended to power by barrel jack + 1A power supply( minimum ).
It also depends on the usb host/hub, on power up we get 100mA, and in
u-boot we request "1.3A" over usb... If we actually get that "1.3A" we
can do a lot, but most often then not we only get 500mA's..
I don't have the exact numbers at my desk, but i thinking a booting
v3.8.x needs a peak around 350mA's for the core & mmc..
Regards,
It also depends on the usb host/hub, on power up we get 100mA, and in
u-boot we request “1.3A” over usb… If we actually get that “1.3A” we
can do a lot, but most often then not we only get 500mA’s…I don’t have the exact numbers at my desk, but i thinking a booting
v3.8.x needs a peak around 350mA’s for the core & mmc…
So that ust be some “special” USB power adapter, that is not plugged into a PC per se. I’ve been powering over USB 3.0 port to the RevC. Which is from a “high end” laptop, which should be able to give the board at least 800ma. I do not recall if I had problems with that setup using ethernet + CAN. But the one where we have the logic supply serial / can cape connected to. Used a 1A wall-wart to USB adapter. It worked fine for a little while, then start exhibiting strange behavior at system up. So on a hunch, wulf and I just replaced that power supply with a 1A barrel jack supply, and the problem went away . . .
I’m still confused how you can get more than 500ma over USB2.0 though . . . as the spec says 500ma max.
It's in the battery charging spec.. So for most devices users plug
this in, it doesn't work..
Regards,
rise time spec
1000ms setup 50ms rise time
maybe gerald can give his thumbs up on that spec
seems fast enough can you put a scope on it to be sure ?
maybe the power supply is out of spec some of the time
i have been using a cheap 5 dollar chinese supply for 2 years with no issues
I forgot:
I power up the power supply ... So not only connected the BBB on the Running power supply.
Trank you
You definitely need a supply that can generate enough current to have a decent ramp-up of 5V or the PMIC gets angry. But, there’s also the issue of u-boot stopping if it sees a “phantom character” from the UART. So, if you need high reliability start-up you should rebuild u-boot for “keyed entry” as well.
-Ron
Hello Ron,
what did you mean with “keyed entry”?
Maybe you have a link or a instruction for this?
Thank you!
Hi Fohnbit,
Keyed entry just means that u-boot will be looking for a specific stop character instead of “any key”. Details for the changes are here with a link to RCN’s instructions & patches for building u-boot.
-Ron
I have also seen this occur on the order of 1/20 boots using 5V power jack. I’m guessing it could be due to a ‘glitch’ in the power supply, or possibly to noise entering some of the GPIO pins that effect boot. It looks to me that in these cases, the BBB does boot, but just doesn’t turn on LED’s as holding the RESET pin (i.e S3) down for 10s turns BBB off; this indicates to me OS is up and timers are working.