Pin1.09 is initialized high at the beginning of code execution

[Giannis Anastasopoulos]

Dear All,

I have noticed the following issue:

In the beginning of my code execution pin 1.09 is high, before I can even call

nrf_drv_gpiote_init();

.

From what I found in the datasheet pin p1.09 is used also for trace data, so I am assuming that the pin is configured for that in the beginning and then when my code executes it initializes to low, but before that it is set to high.

Is my assumption right? Is there any way to make the pin start low in the beginning of the execution? Am including the measurement from a logic analyser.

I am using NRF5SDK v16.0.0, softdevice: s113, version 7.

[Kenneth]

The pin should not by default be setup for TRACE no, if you don't call nrf_drv_gpiote_init() does it stay high? Have you tried to pull it low with for instance a 1Mohm pull down, just to check by what amount it is pulled high? Can it be something nearby or connected to P1.09 pulling it high? All pins should by default be setup with floating/disconnect on power on reset, and if the pin was configured with TRACE, you would not be able to override the behavior by using nrf_drv_gpiote_init(). You can possible check the content of https://infocenter.nordicsemi.com/topic/ps_nrf52833/clock.html#register.TRACECONFIG 

Kenneth

[Giannis Anastasopoulos]

Dear Kenneth,

Based on what I see on the nRF52833DK, pin 1.09 is high by default. If I do not touch it the pin remains high, after the initialization of the gpio device, so I think that there is something special with this pin forcing it to behave differently.

I was able to verify this with my logic analyser (CHRG is the pin in question).

[Giannis Anastasopoulos]

The issue was with the bootloader. There was some custom GPIO initialization there, that was setting the pin high

[Kenneth]

See your update now, good to hear you found the problem!

Kenneth