SWD not working after using ONOFF button or setting SNVS->LPCR TOP on i.MXRT 1064

[gabrielharrison]

Hi,

We are testing, and would like to use, the PMIC in the i.MXRT 1064 using our own design PCB which has a coin cell battery powering the SNVS.

When we press the ONOFF button for 5 seconds the device turns off as expected however power cycling the main 3v3 supply or using SRC.POR_B reset does not get the device into a state where we can connection the SWD debugger. 

The debugger can see power is present but reports "JTAG test failed - the TDO/SWDIO pin held permanently high."

The only way to restore the MCU is be pressing the ONOFF again.

Is this expected behavior? 

Gabriel

[gabrielharrison]

Can anyone confirm that,

1. POR_B will not restore the CPU to full power

2. It is correct that the SWD does not work when in SNVS shutdown?

If both these statements are true, what is the correct way regain debug access if the device has used the SNVS->LPCR TOP register to gracefully shutdown?

[FelipeGarcia]

Hi Gabriel,

It is expected behavior that SWD does not work when in SNVS mode. It is the same behavior as if you disconnect the board while debugging. You will need to restart debug session or attach to a running target after that.

Best regards,

Felipe

[gabrielharrison]

Thank you Felipe and apologies for not replying earlier. I write something which never got posted.

Could you confirm if doing a reset using POR_B from SNVS will return the device to SNVS or full power mode?

Regards,

Gabriel

[gabrielharrison]

Hi,

The power cycle was only the 3v3 so SNVS remains powered with the coin cell - thank you for clarifying that.

We would like to use the ONOFF pin full PMIC functionality, and have TEST_MODE tied low.

The reset I tested was using pin M7 POR_B, not the ONOFF pin. Should this bring the chip back to full power?

Thanks,

Gabriel

[tomula]

I don't know that for sure but that's something I would expect from battery backup that it will retain its state even after its main power cycles. And in that context I don't believe that POR reset should wake it up.

I would tie down TEST_MODE to ground through pull-down so I can connect it to high when programming and test it with ONOFF user reset.

[tomula]

By cycling power you mean SNVS power too? If not then thats why. If you want to do a user reset you have to set TEST_MODE pin high and then you can use ONOFF pin as classic reset pin.