Immediate HardFault on custom nRF52832 board

[Lucas Magasweran]

I am trying to port blinky_blank (no SoftDevice) to my target and started by creating a custom_board.h header file with correct GPIO pins for the LEDs and duplicating blinky_blank_pca10040 armgcc Makefile project. The project flashes and runs just fine on an erased PCA10040 version 0.9.0 (granted, the LEDs don't blink since they're moved). However, when debugging with my custom target the debugger is not able to break on main() and when I pause, I see it's in the HardFault_Handler(). Breaking on SystemInit or _start does not work either so this is happening very early. I was sure to erase the entire chip where I had a working BLE application before.

The PCA10040 has a N52832 QFAABA and my target has a QFAAAC.

Any tips on how I can debug this? I don't think it's a crystal or linker file issue since I'm using a SoftDevice.

[Hung Bui]

Hi Lucas,

The QAAAC is the Engineering A chip, which is very early version and has lots of errata with it. Maybe some of the workaround is not applied when you test with newer SDK. Could you try to test with SDK v0.9.0 here ? Also you mentioned that you had a working BLE application before on your custom board, would you be able to debug with that application ? I mean to get to main loop ?

[Lucas Magasweran]

Thank you @HungBui ! SoC revision was my issue. I updated my board to a Engineering C SoC and ported my project to the latest nRF5 SDK v11.0.0. I did an nrfjprog --eraseall to remove the previous SoftDevice and now blinky works with debug support.

[Hung Bui]

Hi Lucas,

Good to know it worked now, and please be noted that if you want to use PCA10040 v0.9 with S132 v2.0.0 production you would to apply a PAN fix. Please refer to the release note.

[Lucas Magasweran]

Hi @hungbui, I am not sure what a PAN fix is. The release notes mention that to use the latest SoftDevices on Engineering A and B nRF52 chips you must implement the workaround for anomaly 73 (TIMER: Event lost, see infocenter.nordicsemi.com/.../errata.html ).

[Hung Bui]

Hi Lucas,

Yes it's the anomaly 73 as in this document: infocenter.nordicsemi.com/.../nRF52832_Engineering_B_Errata_v1.1.pdf

We called them PAN (Product Anomaly Notification) earlier.