NRF_ERROR_NOT_SUPPORTED returned by sd_power_system_off() "sometimes", nrf52

RE: NRF_ERROR_NOT_SUPPORTED returned by sd_power_system_off() "sometimes", nrf52

Vidar Berg — Wed, 14 Feb 2018 08:03:29 GMT

Thanks for the update. I think the un-init is a good way to ensure that DMA is disabled before entering sleep.

RE: NRF_ERROR_NOT_SUPPORTED returned by sd_power_system_off() "sometimes", nrf52

MattFL — Tue, 13 Feb 2018 20:56:11 GMT

Quick update; I now un-init the SPI and I2S drivers before going to sleep, and I'm not 100% sure that fixed the problem since it was intermittent, but I haven't seen the problem since. Thanks for the tips about DMA, that helped a lot!

RE: NRF_ERROR_NOT_SUPPORTED returned by sd_power_system_off() "sometimes", nrf52

MattFL — Mon, 12 Feb 2018 22:52:26 GMT

Thank you very much for the reply. At the time the problem occurs, there is no debugger attached, and DMA operations "should have" finished some time ago. Is there a way to forcefully stop the DMA, is it as easy as setting a register?

RE: NRF_ERROR_NOT_SUPPORTED returned by sd_power_system_off() "sometimes", nrf52

Vidar Berg — Sat, 10 Feb 2018 12:13:00 GMT

Hello,

sd_power_system_off() should not return unless the chip is in debug interface mode, then it is as expected. Copied from PS:

"NVMC — Non-volatile memory controller, CPU, Flash, and RAM. Since the CPU is kept on in an emulated System OFF mode, it is recommended to add an infinite loop directly after entering System OFF, to prevent the CPU from executing code that normally should not be executed."

It's also important that any DMA transactions are completed before entering system OFF mode. The Power Management library can be used to poll DMA status and delay shutdown until there are no pending transactions.