I am working on making a custom carrier board for a Jetson TX2 (4GB) module. The design guide states clearly that discharge circuitry is required to meet power-down requirements. The discharge circuitry on the NVIDIA carrier board has FETs to discharge all of the voltage rails at the same time.
My carrier board, however, does not have all of these voltage rails, and does not need to support timing requirements for display/camera interfaces like the NVIDIA board does. So I have the following questions:
Is this discharge circuitry necessary still? What would happen if it was not included on a simple custom carrier board?
If so, what about the module itself requires it?
The following questions aren’t related to discharge:
I saw a single small note in the design guide that specifies GND should be connected before power. What exactly would go wrong if they were connected almost simultaneously? I know for hotswap applications, you always connect GND first but this module is definitely not a hotswap capable device.
What do I need to do on my carrier board with the RESET_OUT# signal of the module? The power-up sequence diagrams in the OEM DG state that it should be de-asserted after the carrier board has powered up, but I have seen elsewhere that the module itself de-asserts it.