Encoder

/*

   This sketch shows how to use the Encoder IP in MKRVidor 4000

   Quadrature encoders are a very common way to detect position (and speed) of a moving part, like a DC motor.

   Normally, in the microcontroller world, decoding a quadrature encoder has serious limitations since all egdes must be counted, otherwise the final number will be wrong. This is usually accomplished using interrupts, but over a certain revolution speed the intinsic overhead of servicing an interrupt destroys the count reliability.

   Using the FPGA to perform decoding allows:

    - not to lose any edge until the revolution speed is less than some million RPM :)

    - we can "read" the encoder at any time, because the FPGA is counting independently

    Circuit:

      connect ENC_A to A0 and ENC_B to A1

*/

#include "VidorGraphics.h"
#include "VidorEncoder.h"

// Initialize Encoder #0 (connected to A0 and A1)
// Refer to the online documentation to find which pins correspond to a given index
// This assignment may change between bitstreams

VidorEncoder encoder(0);

void setup() {

  Serial.begin(9600);

  // wait for the serial monitor to open,

  // if you are powering the board from a USB charger remove the next line

  while (!Serial);

  // Let's start the FPGA

  if (!FPGA.begin()) {

    Serial.println("Initialization failed!");

    while (1) {}

  }
}

void loop() {

  // Read the encoder

  int value = encoder.read();

  Serial.print("Encoder value: ");

  Serial.println(value);

#if 1

  // Wait one second with interrupts disabled

  noInterrupts();

  // We can't call delay() since it uses interrupts, so use busy loops

  for (int i = 0; i < F_CPU / 10; i++) {

    asm ("nop");

  }

  interrupts();

#else

  delay(200);
#endif

  if (value >= 200 || value <= -200) {

    // Reset the count

    encoder.write(0);

  }
}