|
| 1 | +#include <Arduino.h> |
| 2 | + |
| 3 | +/* |
| 4 | + * A class to interpret IR remotes with NEC encoding. NEC encoding sends four bytes: |
| 5 | + * |
| 6 | + * [device ID, ~divice ID, key code, ~key code] |
| 7 | + * |
| 8 | + * Sending the inverse allow for easy error checking (and reduces saturation in the receiver). |
| 9 | + * |
| 10 | + * Codes are send in little endian; this library reverses upon reception, so the first bit received |
| 11 | + * is in the LSB of currCode. That means that the key code is found in bits [23..16] of currCode |
| 12 | + * |
| 13 | + * https://techdocs.altium.com/display/FPGA/NEC+Infrared+Transmission+Protocol |
| 14 | + * |
| 15 | + * This does not interpret the codes into which key was pressed. That needs to be |
| 16 | + * mapped on a remote by remote basis. |
| 17 | + */ |
| 18 | + |
| 19 | +/* |
| 20 | + * TO DO: |
| 21 | + * - (MED) Possibly indicate repeat code instead of just "refreshing" the latest code? |
| 22 | + * - (LOW) Consider a better way to indicate that a valid code has been entered; |
| 23 | + * returning -1 is OK if we return a 16-bit number (so 255 is a valid code), |
| 24 | + * but there may be a more elegant solution. |
| 25 | + * - (MED) Add ability to use PCINT library to this generic library. Use a derived class? |
| 26 | + */ |
| 27 | + |
| 28 | +class IRDecoder |
| 29 | +{ |
| 30 | +private: |
| 31 | + enum IR_STATE |
| 32 | + { |
| 33 | + IR_READY, //idle, returns to this state after you request a code |
| 34 | + IR_PREAMBLE, //received the start burst, waiting for first bit |
| 35 | + IR_REPEAT, //received repeat code (part of NEC protocol); last code will be returned |
| 36 | + IR_ACTIVE, //have some bits, but not yet complete |
| 37 | + IR_COMPLETE, //a valid code has been received |
| 38 | + IR_ERROR |
| 39 | + }; //an error occurred; won't return a valid code |
| 40 | + |
| 41 | + uint8_t irPin; |
| 42 | + |
| 43 | + IR_STATE state = IR_READY; //a simple state machine for managing reception |
| 44 | + |
| 45 | + volatile uint32_t lastReceiveTime = 0; //not really used -- could be used to sunset codes |
| 46 | + |
| 47 | + volatile uint32_t currCode = 0; //the most recently received valid code |
| 48 | + volatile uint8_t index = 0; //for tracking which bit we're on |
| 49 | + |
| 50 | + volatile uint32_t fallingEdge = 0; |
| 51 | + volatile uint32_t risingEdge = 0; |
| 52 | + |
| 53 | + volatile uint32_t lastRisingEdge = 0; //used for tracking spacing between rising edges, i.e., bit value |
| 54 | + |
| 55 | +public: |
| 56 | + IRDecoder(uint8_t pin) : irPin(pin) {} |
| 57 | + void init(void); //call this in the setup() |
| 58 | + void handleIRsensor(void); //ISR |
| 59 | + |
| 60 | + uint32_t getCode(void) //returns the most recent valid code; returns zero if there was an error or nothing new |
| 61 | + { |
| 62 | + if (state == IR_COMPLETE || state == IR_REPEAT) |
| 63 | + { |
| 64 | + state = IR_READY; |
| 65 | + return currCode; |
| 66 | + } |
| 67 | + else |
| 68 | + return 0; |
| 69 | + } |
| 70 | + |
| 71 | + int16_t getKeyCode(bool acceptRepeat = false) //returns the most recent key code; returns -1 on error (not sure if 0 can be a code or not!!!) |
| 72 | + { |
| 73 | + if (state == IR_COMPLETE || (acceptRepeat == true && state == IR_REPEAT)) |
| 74 | + { |
| 75 | + state = IR_READY; |
| 76 | + return (currCode >> 16) & 0x0ff; |
| 77 | + } |
| 78 | + //else if(state == IR_ERROR) return currCode; //for debugging, if needed |
| 79 | + else |
| 80 | + return -1; |
| 81 | + } |
| 82 | +}; |
| 83 | + |
| 84 | +class IRDecoder32U4 : public IRDecoder |
| 85 | +{ |
| 86 | + IRDecoder32U4(uint8_t pin) : IRDecoder(pin) {} |
| 87 | +}; |
| 88 | + |
| 89 | +extern IRDecoder decoder; |
0 commit comments