Checking nRF52832 SoC for hardware faults

[Marco]

Dear support,

I am working on some custom boards embedding the nRF52832-CIAA chip using SDK-v11.0 and s132-v2.0.1.

I have got 9 custom units and out of them only 4 work as expected with exactly the same firmware (SD+bootloader+application).

Four faulty units are sharing the same issues:

1. UART clock frequency, it requires frequency adjustment (RealTerm), no matters what frequency I set in my firmware

2. power source measurement via SAADC (reading 2.2V when feeding it with 2.9V)

3. FICR register, in fact I had to comment out the initial test in the bootloader:

   APP_ERROR_CHECK_BOOL(NRF_FICR->CODEPAGESIZE == CODE_PAGE_SIZE);

Reading the FICR register returns the following:

$  nrfjprog --family nRF52 --memrd 0x10000000 --n 0x80
0x10000000: FFFFFFFF FFFFFFFF FFFFFFFF FFFFFFFF   |................|
0x10000010: FFFFFFFF FFFFFFFF FFFFFFFF FFFFFFFF   |................|
0x10000020: FFFFFFFF FFFFFFFF FFFFFFFF FFFFFFFF   |................|
0x10000030: FFFFFFFF FFFFFFFF FFFFFFFF FFFFFFFF   |................|
0x10000040: FFFFFFFF FFFFFFFF FFFFFFFF FFFFFFFF   |................|
0x10000050: FFFFFFFF FFFFFFFF FFFFFFFF FFFFFFFF   |................|
0x10000060: FFFFFFFF FFFFFFFF FFFFFFFF FFFFFFFF   |................|
0x10000070: 464B4150 48013234 FFFFFF30 FFFFFFFF   |PAKF42.H0.......|

It looks like the FICR register has not been initialized properly at the factory.

Do you have any suggestion on how to investigate the matter?

Thank you, Marco

[Marco]

Could it be possible that the listed faults might be caused by me doing something wrong? Or maybe could they all have a root cause (e.g. crystal oscillators, shorted lines, etc.)? In any case I will open a case directly with Nordic. Thanks for the suggestion. Could you please have a look also at my other issue? Thanks devzone.nordicsemi.com/.../