The spec says usb C but also makes it sound like it will take voltage directly.
Can I power rock5B with 12v Dc battery?
Rock5 12v power
Yes as long as you wire them to the correct pins of the USB-C. I’m powering one of mine Rock5 using direct 12v dc at the moment and it works well.
Awesome thanks for sharing, by chance any reference a schematic or data on your hookup, did you wire directly to port pins? Did you need a PD sink negotiation chip?
I’m using an off-the-shelf adapter like this one. You don’t need a PD sink negotiation chip, just connect the pins.
4 Likes
I can confirm that these USB-C leads work well hardwired to a 12v power source. That’s how I’m powering my Rock 5b (running Armbian).
You can also get the voltage currently connected with this bit of magic.
printf "%.1f V\n" $(bc -l <<< "$(</sys/devices/iio_sysfs_trigger/subsystem/devices/iio\:device0/in_voltage6_raw) / 172.45")
The 172.45 part is from ROCK 5B Debug Party Invitation
1 Like
May I also add that people need to use a 100W type c cable or a 4k type c cable capable of powering and not just any shitty type c cable underpowered for the overpower of the charger?
Keep that in mind also.
Did that happen on a non-PD source, or was that the result of an incorrect power delivery negotiation at a much higher rate?
The cable I’m using is connected to 12v and there’s no PD negotiation to mess things up.
You can see if there was a PD negotiation with this…
cat /sys/class/typec/port0/port0-partner/supports_usb_power_delivery
If you are using a non-PD connection it should return no.
1 Like
Cant see cause it didnt happened to me and to the person who did… well… lets just say the board burnt.
Yeah… I totally believe it could happen if someone used a data cable for a higher voltage PD rate.
I think the world would be a better place if power over USB had never become a thing. This is an area where I think ODROID has a better design. Just put barrel connectors on the boards and keep data and power separate.
The thread was about powering with things like large 12v batteries. It’s totally possible to do with non-PD, two-wire cables like the ones posted. The cable I’m using has two 22 AWG conductors in it… guessing that’s probably larger than the power conductors bundled in a regular USB-C cable like the one that got hot in the picture. 
This is clearly not an 22AWG (0.64516) thick cable, the most common in renewable energy/12V or 24V wiring. Here is a table of relation for safty usage. The limit of the wire is in Ampers. So for the same W (W=VA) in 12V you can use 12V7A=84W without problems.
After the product brief rev1.1, p7 (datasheets are all non-available 
):
The recommended power source should be able to produce, at least, 24W without a M.2 SSD or 30W with a M.2 SSD.
So ar 12, with M.2 SSD:
30/12=2.5A a 24AWG (0.51054mm, 3.5A max, is still really safe).
I don’t know if it can relay more, if you plug lot of devices by USB+PCI+GPIO ports.
1 Like
BTW you can even simplify it to avoid a fork-exec if needed, and report the voltage in millivolts:
$ echo $(($(</sys/devices/iio_sysfs_trigger/subsystem/devices/iio\:device0/in_voltage6_raw) * 100000 / 17245))
12003
1 Like
Just fyi, I found that directly powering my Rock 5B from my RV’s 12V nominal electrical system worked but the Rock 5B rebooted every day or so.
I suspect it was because the voltage of my 12V nominal system varied due to solar and alternator charging plus battery level. Voltage could vary between 14.4V down to almost 12V, and could swing more during the day from my solar panels and MPTT controllers and the sun.
Since I added a 12V regulator to the power input of the Rock 5B, it has been much more stable. I used the DROK DC-DC Stabilizer, 9V-36V to 12V Boost Buck Converter 5A 60W:
2 Likes
I was having some issues using a PD charger - under high CPU or IO load the board would reboot. My PD charger supplies 20V and should be capable. After some filesystem corruption from the reboots I attached a 4TB disk to try to get a backup, while backing up to it I can hear it resetting, meaning 5v line is unstable. I have a Samsung nvme too,
I have made a barrel jack to usb-c adapter and use a 12v power supply and that resolved those issues. To test under high load I ran 7zip benchmark while reading my nvme disk to /dev/null and the watts consumed rose to about 17W. The 4TB disk is attached and working without issue. The 12v supply I used is the one from my previous rk3399 board rated at 2.3A 12V so should be enough for my needs.
I think there may be issues with an input of 20V and high watts, maybe the buck converter has issues where it doesn’t at 12V?