DSP_and_Math

The idea behind this library is provide some useful DSP and Math functions, focusing on embedded systems (multi platform).

To enable multiple instances (similar to C++) various functions receive and return values via structs containing the related parameters. With this is possible, for eg, to run two IIR High Pass in diferent signals or channel. But this approach creates a little overhead. To overcome this is possible to select the prototype of functions using "__inline" in the "DSP_and_Math.h" and also use compiler optimization.

To be microcontroller friendly, some functions were implemented entirely in fixed point, other functions use the main calculus in fixed math and return the result in float, allowing a trade off between accuracy and efficiency.

Plot Examples

High pass Filter

Low pass Filter

Implemented Math Functions

Square Root (SQRT)

Square Root of a Integer value (int32_t)

int32_t sqrt_Int32(int32_t x);

RMS value

RMS value of an array of N samples

float rmsValueArray_Float_StdMath(const float * arrayIn, uint_fast16_t size, float dcLevel);
float rmsValueArray_Int16_StdMath(const int16_t * arrayIn, uint_fast16_t size, int16_t dcLevel);

RMS sample by sample

Enable to add samples on demand and calculate the value after N samples. After add all samples calculation is done with a separate function

void rmsValueAddSample_Float(rms_float_t * inputStruct, float sample);
void rmsValueAddSample_Int16(rms_int16_t * inputStruct, int16_t sample);

void rmsClearStruct_Float(rms_float_t * inputStruct);
void rmsClearStruct_Int16(rms_int16_t * inputStruct);

void rmsValueCalcRmsStdMath_Float(rms_float_t * inputStruct);
void rmsValueCalcRmsStdMath_Int16(rms_int16_t * inputStruct);

Sine wave generator

Applying the same function N times is possible to generate complex waves with harmonics (see examples)

Array version

Generate the sine wave in a defined float array

void sineWaveGen_Array_Float(float * outputArray, float freq, float phase_rad, float amplitude, float V_offset, uint_fast16_t points, uint_fast8_t doClean);

Sample-by-sample

Allow to generate each sample iteratively

void sineWaveGen_bySample_Init(sine_wave_parameters *inputParameters, float freq, float phase, float amp, float v_off, uint_fast16_t points, uint_fast8_t doClean);
float sineWaveGen_GetSample(sine_wave_parameters *inputParameters);

Implemented DSP Functions

IIR Single Pole High Pass

IIR Single Pole High Pass Filter - Float Point Version

First function initialize the struct with parameters based in the cuttof frequency (normalized from 0 to 1). It's possible to select between clean or not the variables during the inittialization. Can be useful if you need to change the cutoff frequency withou reset entire filter state.

void iir_SinglePoleHighPass_Float_Init(iirHighPassFloat_t * structInput, float cutoffFreq, uint_fast8_t doClean);