nrf51422 PPI and interrupt latency

[Vito]

I need to do some single-cycle precision time measurement on the RADIO events. I am currently using PPI, configuring the radio events to to trigger timer tasks. From what I have read about the M0 core, the interrupts have fixed predictable execution time. From what I see in chapter 9 "peripheral interface" of nRF51 Series Reference Manual v2.1, it looks like NVIC and PPI share the same inner working.

Is it correct to say then, that TIMER measurements triggered with PPI from events, will have as well a deterministic execution time? Where can I find more information on that?

Do the timer tasks START and CAPTURE have a fixed execution time?

I am not interested in the execution time itself, rather it's fundamental for me to have low or 0 jitter. Thanks.

[Asbjørn]

The PPI timing from event to task is one clock cycle. So yes, the PPI event will have a deterministic execution time.

[Vito]

Thanks. Is there a document where I can see the timings of various functions used in nrf51xxx, in general?

[Asbjørn]

Not more than what is already stated in the nRF51822 product specification and the nRF51 reference manual.