Softdevice assertion failed error at PC 0x17826

[Mayank Panwar]

Hi, 

I am using softdevice s140_nrf52_7.0.1. I am trying to send the SAADC calculated voltage over NUS to a gateway. For like 20-30 seconds the program works fine and the saadc value gets successfully transferred to the gateway over NUS but then the softdevice asserts as PC 0x17826. For SAADC I am using RTC. I am not able to figure out the root cause of this assertion. Can you please help?

[run_ar]

Hi,

Seems this assert is related to timer0. Can you check that your application is not using/changing any of the restricted hw peripherals? See Hardware peripherals

[Houwer de Geus]

Hi,

I am using softdevice nRF5_SDK_17.0.2_d674dde and get the same fault address (0x17826). I also get fault address 0x15a80. I checked the restricted and blocked peripherals, but I am not using one of them. Sometimes it takes minutes, sometimes it takes hours to get the fault. Is there a way to find out what is causing this assert fault? And what may be the cause in my case?

Best regards,

Houwer de Geus

[run_ar]

Hi,
Can you check how you have configured your interrupts to make sure you are not using any interrupt priorities used by the SD?

Can you also confirm which SD version you are using? Seems you are using SDK 17.0.2, but I need to know if you are using S140, 132, 113 or 112 and which version?

[Houwer de Geus]

Hi,

The SDK is: nRF5_SDK_17.0.2_d674dde

The softdevice is s140_nrf52_7.0.1_softdevice.hex

I pin pointed the problem in the uart driver. I wanted to enable/disable the uart RX for the purpose of reading an RS485 bus. In the function that was disabling the RX of the Uart , the RXTO timeout was polled until timeout occurred. This happened while global interrups were disabled. The polling took long enough time to let the softdevice generate the assert. The interrupt disable is not done via de softdevice interrupt disable. So that may be the problem for the softdevice to plan his actions. I tried to use the softdevice function sd_nvic_critical_region_enter() and sd_nvic_critical_region_exit() and that solved the softdevice asserts, but that did cause other problems sending bytes to the uart. I have had no time to investigate that, but we will do that later on.

[run_ar]

ok, let us know how that goes. And please consider creating a new case if you need assistance with that issue. it seems it is not related to the original questions. Thanks