SAADC multiple channel buffer inconsistencies

[John Smith]

I am using nRF52 DK.

When I do 2 analog inputs using the SAADC, I noticed that, in the SAADC callback,  p_event->data.done.p_buffer[0] corresponds to the correct SAADC channel (channel 0) and p_event->data.done.p_buffer[1] corresponds to the correct SAADC channel (channel 1). After about 3 seconds of printing out the values from the channels, I notice that p_event->data.done.p_buffer[0] now corresponds to SAADC channel 1 and p_event->data.done.p_buffer[1] corresponds to channel 0. What is going on? 

Thanks

[haakonsh]

It's a known issue. 

The workaround is to use at least two buffers(ping-pong) and connect the END event to the START task via PPI. 

You need to update the RESULT.PTR and RESULT.MAXCOUNT registers immediately after you've gotten a STARTED event (nrfx_saadc_buffer_convert). If the RESULT registers are not updated before the START task is triggered you will overwrite the previous buffer, therefore you will also need to make sure that you use large enough buffers. 

[John Smith]

Thanks for the response.

Do you have example code that shows me the solution?

Also would be helpful to know how to edit the default saadc example to achieve what you are saying. Is there a file which shows saadc example with the changes you mentioned?