This is a hardware question regarding powering the Beaglebone Black and BeagleBone Black Wireless from a custom add on board.
We have a custom board that is powered by a 12VDC battery and has regulators that supply 5VDC and 3.3VDC. Based on our understanding of the BB SRM Rev. C, we believe we can supply the BB with 5VDC by back feeding through P9_8.
We have assembled our first custom board and confirmed the voltages on the interface pins. As a safe starting point, we chose to jumper from our board’s 5VDC (P9_8) and GND (P9_1) to the Beagle’s P9_8 and P9_1 so that no other pins on our board that might be creating a load would be connected. In other words, in test step 1, we just want to prove the custom board can boot and power the Beagle.
Upon applying power to our board, the Beagle’s blue PWR led comes on and the boot sequence appears to run due to observing flashes of lEDs usr0-usr3. However, eventually, only the PWR led is on and it starts to flash (about 8 times usually) and then all the BB LEDs go off.
We observe that the current draw is fluctuating during this boot up process because we use a benchtop power supply to supply the 12VDC to our board and it shows the current draw.
I don’t see anything in the SRM that indicates what 8 flashes of the PWR led and then shutdown means.
I suspect that the PMIC sees something that it doesn’t like. Since we only jumpered 5VDC and GND from our board to the BBB and nothing else is connected to any other BBB pins, I don’t think anything else could be causing the BBB to shutdown. This BBB powers up with a USB connector just fine when not connected to our custom board.
If we need to route our 5VDC supply to the BBB a different way, is there a reference for how to do that? I haven’t found it in the SRM.
This is a hardware question regarding powering the Beaglebone Black and BeagleB
one Black Wireless from a custom add on board.
We have a custom board that is powered by a 12VDC battery and has regulators th
at supply 5VDC and 3.3VDC. Based on our understanding of the BB SRM Rev. C, w
e believe we can supply the BB with 5VDC by back feeding through P9_8.
we’re doing something like that in our product and our bbbw is also
powered via p9 but we’re feeding in only 5V and that on the VDD_5V line,
ie. pin P9_5. this works fine.
I suspect that the PMIC sees something that it doesn’t like.
the bbb schematic says P9_8 (SYS_5V) is connected to the tps65217 SYS signal,
and the datasheet for the tps65217 says that the SYS pins are output. your
setup treats it like an input.
I found the exact thing over the weekend. I will be testing the theory that feeding our board’s 5V into P9_5 or P9_6 will run the BBB ok. I’m confident we have clean 5VDC from our board and plenty of current to run the BBB. Stand by for results!
So here’s an update for those who might run in to this problem themselves some day.
I used a lab bench supply to power the BBB wireless through P9_5 (VDD_5V) and it worked flawlessly.
I then jumpered from the 5VDC supply on our custom PCB to P9_5 and the board powered up, started the wifi and then shutdown. I did not see any transient drops on the 5VDC with our scope so hhmmm…what on earth is going on here?
I then chose to disconnect everything downstream of the 5VDC regulator’s output rail on our PCB and just have the 5VDC regulator power the BBB wireless, again, through P9_5. It powered up, wifi came on and then it shutdown. Hmmm…Arghhh…
So I put together a 5VDC regulator circuit on a protoboard using a through-hole version of the same regulator (the one on the PCB is surface mount) and powered the BBB Wireless from that off-board regulator. The BBB wireless came up and ran like a champ.
So I removed the surface mount version of the regulator and soldered in the through-hole version and now the Beagle is being powered by our custom PCB through P9_5.
My conclusion is that although the through-hole and surface mount versions of the regulator are supposed to be the same, the surface mount version may be getting overheated on our board and shutting itself down. What looks like the Beaglebone shutting down first is probably the linear regulator shutting down first from overheating. I further reinforce my conclusion by noting that the through-hole regulator is getting pretty warm although it is drawing less than half the full rated output current (0.6A vs 1.5A) Thankfully we have room to use a through-hole regulator and probably will on our next rev of this PCB.