Counting pulses 200us interval time

RE: Counting pulses 200us interval time

Håkon Alseth — Wed, 09 Mar 2022 13:05:15 GMT

Hi,

The nRF devices GPIOs are not guaranteed to detect such a low pulse (ie. 10 ns).

[quote user="Robertzh"]This is also higher than the given pulse width from the Matrix Mcry12 of 10ns. With your calculation, this can not work, too.[/quote]

I would recommend that you contact Matrix and ask how this is to be implemented.

Kind regards,

Håkon

RE: Counting pulses 200us interval time

Robertzh — Tue, 08 Mar 2022 15:28:45 GMT

Hi Hakon

[quote userid="2115" url="~/f/nordic-q-a/85491/counting-pulses-200us-interval-time/356690#356690"]Unforuntately, we cannot handle such a low timing interval.[/quote]

I am not quite sure if you are correct. I took a look with some colleagues of this problem. Matrix Industries implemented on their Eval board the MCU RL78/G1D from Renesus. And they connect the ADC_DATA signal to a Timer array. The maximum frequency is 32Mhz from the RL78 which results in a period of 31.25ns. I do not see in their datasheet any difference of the timer architecture.

This is also higher than the given pulse width from the Matrix Mcry12 of 10ns. With your calculation, this can not work, too.

I have not tested the EvalKit, but I hardly believe it cannot work as you describe.

May it be, that we missunderstood each other?

Best regards

Robert

RE: Counting pulses 200us interval time

Håkon Alseth — Mon, 07 Mar 2022 16:04:06 GMT

Hi Robert,

[quote user="Robertzh"]

And just for interests: Where do I find the information with the 62.5ns minmal pulse width detection?

[/quote]

This is not explicitly stated in the GPIO/GPIOTE chapter, but it is given by the clock input for the peripherals, which is 16 MHz (1/16 = 62.5ns):

https://infocenter.nordicsemi.com/topic/com.nordic.infocenter.nrf52832.ps.v1.1/clock.html?cp=4_2_0_18#frontpage_clock

[quote user="Robertzh"]You are correct, it is for 10ns low. Anyway, I set up a GPIOTE on a falling edge, therefore I assumed, the length of the pulse width is irrelevant. Do I miss something?[/quote]

Unforuntately, we cannot handle such a low timing interval.

Kind regards,

Håkon

RE: Counting pulses 200us interval time

Robertzh — Mon, 07 Mar 2022 15:49:15 GMT

Hi Hakon,

Thank you very much for your response.

You are correct, it is for 10ns low. Anyway, I set up a GPIOTE on a falling edge, therefore I assumed, the length of the pulse width is irrelevant. Do I miss something?

And just for interests: Where do I find the information with the 62.5ns minmal pulse width detection?

Best regards

Robert

RE: Counting pulses 200us interval time

Håkon Alseth — Mon, 07 Mar 2022 15:31:52 GMT

Hi Robert,

I just wanted to confirm this parameter:

[quote user=""] tLow = 10ns.[/quote]

Is the pulse low for 10 ns for each bit?

If that is the case, you will not be able to detect this pulse using any nRF device. The lowest we can detect is 62.5 ns.

Kind regards,

Håkon