I also have two of the JetBot kits and have similar issues to mcinnovation. In my case, however, the nano and motor driver shut off regardless of which battery is plugged in.
Setup:
stock SD card image (5W mode)
battery powering the nano using the orange USB port, motor driver using the black USB port
battery at full charge before test
JetBot assembled as per the provided Assembly Guide (camera plugged in, I2C connected, WiFi module inserted, etc.), using no third party components.
Behavior:
Motor driver USB plugged in first, followed by Nano’s USB: Nano powers off during boot. Connecting to the Nano’s UART serial console shows that it dies soon after the kernel begins to boot (i.e. the line "Starting kernel… ")
Nano’s USB plugged in first, followed by motor driver: Nano boots as expected, before the motor driver USB is connected. Once the motor driver is plugged in, the Nano dies within a couple of seconds (but not immediately).
Note: once the Nano has died while running off of the batteries, the batteries will sometimes, but not always, need to be plugged into a charger before they will power the Nano back on again.
Trobleshooting:
Swap batteries (from both kits): both batteries die in the same manner described above.
Unplug pHat and all I2C/Qwiic connections: same behavior as above. This rules out the possibilities of the I2C ground being used as the motor driver’s return path or a faulty OLED board causing issues.
Power Nano from the barrel jack, motor driver from the battery: works as expected, can run jupyter notebook demos
Power Nano from one battery pack, motor driver from the other battery pack: works as expected, can run jupyter notebook demos
Note: while both battery packs were originally able to power the Nano alone (sans motor driver) without issue, one of them is now behaving erratically and will sometimes die even without the motor driver connected.
Any ideas other than RMA?