Effects when the VDD_IN cannot keep longer than 20ms in an uncontrolled powerdown

The “Jetson TX2 OEM Product Design Guide” states the power on USB0_OTG_ID pin should stay above 2.9v for longer than 20ms because TX2 module needs 20ms to do a controlled shutdown. That basically means a big capacitor needs to be connected to the VDD_IN line. The capacitor needs to be hundreds uF or about 1000uF. That will occupy a large space on the carrier design. To avoid using a large capacitor, I have two groups of questions:

If VDD_IN cannot be kept long enough for TX2 to finish the controlled shutdown, what would be the effect exactly? Would it be just a Linux user filesystem corruption? Is there a way to repair from the corruption after the module comes back up?

When VIN_POWER_BAD# goes low, that triggers a controlled shutdown. In this controlled shutdown, which part/parts of the system is/are involved? Is that just the kernel part, or is it also other processors involved? Is there a way to modify the code so that the system can perform the controlled shutdown faster by throwing away some data that normally should be flushed?


Hi minghua,

Module has such capacitors already, no need big one on carrier board. This controlled shutdown process is fixed, please just keep it as default status.

Hi, Trumany, Thanks for the information. Good to know the answer is simpler than I expected.

Another question:

In the new “Jetson TX2/TX2i OEM Product Design Guide” of 20180306, the VDD_MOD or VDD_IN to TX2 also gets discharged in power-down as shown in Figure 2 on page 10. This was not in earlier version of the document named “Jetson TX2 OEM Design Guide”. Is this new changed required for TX2 module too? What is the effect or what problem would be caused if this is not implemented on the carrier board?


No effect to TX2, that design is backward compatible. You can check doc of Jetson TX2-TX2i Interface Comparison and Migration for more.


Thank you!