Timer 0 interface program

RE: Timer 0 interface program

Håkon Alseth — Fri, 30 Aug 2019 12:32:30 GMT

Hi,

Could you try this sample and see if it works on your end? It uses TIMER in 16M resolution.

https://github.com/Rallare/fw-nrfconnect-nrf/tree/nrf9160_samples/samples/nrf9160/nrfx_timed_signal

You might have to change the pins to match your use-case. Its GPIOs 6 and 7 now, which are active low at my end.

Note the spm diff file, where DPPIC is set to non-secure.

Kind regards,

Håkon

RE: Timer 0 interface program

hmdra — Thu, 29 Aug 2019 13:31:33 GMT

Hi Hakon,
Thank you for your valuable response,

i need micro second timer accuracy, ultrasonic need two GPIOs, one for tirgger and another is for echo, i configure trigger as output and echo pin as interrupt, as per the ultrasonic datasheet the echo pin returns the high state for a short period of time which is help to measure the distance. so that i start measure timing for initial high level and end of the high level, then i converts it as nanosecond.

the posted program is now working good to me, the only thing make my timing different is printing statement. it tooks more time so i have missed the original count.

as per your suggestion i have doubt how to configure GPIOTE as toggle. i didnt have an API for that.

Thanks

hmdra

RE: Timer 0 interface program

Håkon Alseth — Wed, 28 Aug 2019 12:14:28 GMT

Hi,

What is the required accuracy for the signal? Right now you're using the RTC (the kernel timer), which is 32.768 kHz base frequency, which might not be accurate enough for your application.

In order to do this as accurate as possible, it would require:

* NRF_TIMER

* 2 x GPIOTE IN (One on rising, one on falling edge)

* Two PPI channels (one to start, one to stop TIMER)

Connect the signal you want to time to two GPIOs (configured as IN), as "GPIOTE IN Toggle" will not be able to differentiate between the start and stop of the signal here.

This method will require either bare-metal or the usage of nrfx drivers directly.

Kind regards,

Håkon