Some_RPI-Pico_stuff VS rp2040pio

You don't need to sum anything if you don't want to! If you use DMA, it can do the summing for you. Here is an example.

The problem is very similar to a combination of the button-debouncer with the pio/C++ IRQ mechanism from this rotary encoder.

There's an example PWM input implementation here: https://github.com/GitJer/Some_RPI-Pico_stuff/blob/main/PwmIn/PwmIn.pio which uses one state machine per input.

Just realised you were talking about PWM inputs, I had outputs on my mind. For testing I have just been counting the pulses by adding an interrupt on falling edges, but if I can treat it like a PWM input that simplifies things. Is that what you're doing here with a Hall-effect sensor?

I needed a way to get sensor-data from an LMT-01. The pulses were too fast so Interrupts were out of the question. After some help and some fiddling around we came up with this.

The pico datasheet does a reasonably good job. And maybe you can look at some examples and explanations here.

I have implemented pio code to measure the pulses in the pulse train as you described it (I do not have a LMT01, so I generated the pulses using another MCU). Since I do not use micropython I have made it using the C++ SDK, but if you look at the .pio file here, it should be no problem to implement it. The problem with the wrong readings only happen the very first time if the state machine starts in a pulse train.

Where Paul's code is rather simple and elegant, mine is rather convoluted. Why? Because I wanted to play with the DMA sniffer and a repeated timer! The DMA sniffer allows summing of all values that pass through the DMA. So, there is no need to sum the values yourself afterwards.

PIO emulator: https://github.com/soundpaint/rp2040pio

Be sure to check out the RP2040 datasheet for more in-depth learning and this PIO emulator for more advanced development and debugging scenarios.