nrf_sdh_enable_request Doesn't Return on a Custom PCBA

[WestCoastDaz]

I have a custom board that won't return from the call to nrf_sdh_enable_request(). It's very similar to this post https://devzone.nordicsemi.com/f/nordic-q-a/42477/doesn-t-return-from-sd_softdevice_enable except that I have an external crystal mounted on the custom PCBA.  When I set NRF_SDH_LRF_SRC 0, NRF_SDH_CLOCK_LF_RC_CTIV 16 and NRF_SDH_CLOCK_LF_RC_TEMP_CTIV 2 the call nrf_sdh_enable_request()returns fine.

This suggest to me that there is a problem with the XTAL2 but when I try the RTC peripheral example code it seems to work fine.  Looking at the sdk_config.h file for the example it seems that it's running from the LF XTAL.  Is there a method to confirm the crystal is oscillating?  I uses my Saleae Logic Analyzer on the crystal load caps but I didn't see any oscillations.

SD 7.0.1, SDK 16

Thanks,
Darren

[hmolesworth]

"If the LFXO is selected as the clock source, the LFCLK will initially start running from the 32.768 kHz LFRC while the LFXO is starting up and automatically switch to using the LFXO once this oscillator is running. The LFCLKSTARTED event will be generated when the LFXO has been started."

I think the SoftDevice monitors LFCLKSTARTED whereas RTC does not; this suggests as the RTC works that the 32kHz crystal never starts oscillating, usually a load cap issue.

[haakonsh]

What Hugh already said is true.

Unless you've specifically start the LFXO through the clock driver/HAL the RTC will use the LFRC as the source. 

Is there a method to confirm the crystal is oscillating?

1.  Use the CLOCK driver to start the LFXO:
   
   nrfx_clock_init() -> nrfx_clock_lfclk_start() -> while(!(nrfx_clock_lfclk_is_running())){ //Do nothing, block until LFXO starts}
2. Scope the crystal's pins with an oscilloscope.

[WestCoastDaz]

hmolesworth

Thanks for the help!