Failing to understand app_timer

[ImQ009]

Hello.

I cannot get my head around app_timers. After reading the docs, examples and answers on the devzone I'm still unable to get it running. The timeout callback isn't called after starting the timer. Any assistance, please?

app_timer_id_t timerFoo;

void foobarHandler(void *context)
{
	log("foobar\r\n");
}

int main()
{	
	uartInit();
	log("begin\r\n");
	
	APP_TIMER_INIT(0, 1, 4, NULL);
	if(app_timer_create(&timerFoo, APP_TIMER_MODE_SINGLE_SHOT, foobarHandler) != NRF_SUCCESS) log("creation failed\r\n");
	if(app_timer_start(timerFoo, APP_TIMER_TICKS(2000, 0), NULL) != NRF_SUCCESS) log("starting failed\r\n");
	
	return 0;
}

[ImQ009]

Thank you so much! That's really helpful. Now, what if I want delays shorter than a milliecond? Do I simply give APP_TIMER_TICKS a floating point value?

[RK]

definitely not as it takes an unsigned integer. Not much uses floats (which is a good thing)

There's really loads of example code on this plus lots of comments in the header files about how to set the divider, there are macros to get ticks for different milliseconds, microseconds etc etc. The documentation has a good intro to the whole library too.

I really would look at the examples, although you aren't going to get much less than 1/10 millisecond with this library.

[ImQ009]

I wouldn't ask if I hadn't already looked at the examples. I couldn't find any that actually used app_timer.. . I couldn't find any that would use fractions of a millisecond for timing either. I guess I have to look through all of the examples available, eh?

Either way, thank you for your assistance so far

[RK]

There's loads, just grep for APP_TIMER_INIT, I found 38 of the examples used it.

You're not going to get fractions of a millisecond with this library, RTC doesn't run fast enough.

[ImQ009]

Heh. This looks like it's going to be much more complicated than I thought. Essentially I'm just trying to implement a simple timeout system on UART for packet transmission... 1/10 millisecond accuracy sounds alright... What about "nrf_delay_us"?