SMC3 code update using ResponsiveAnalogRead library to mitigate cogging effect for Flight Sim

Good day everyone,


I was reading/ researching a bit on motor cogging - while i feel it a little bit when hand flying (in msfs), it is more pronounced when on autopilot. I guess the changes are more gradual.

I, unfortunately, am not a programmer nor a specialist on motor control, hence was engaging with ChatGPT on possible improvement of the step size.

It suggested to use the library ResponsiveAnalogRead in the SMC code. I thought to post it here and it even wrote part of the code (yet to see if it has errors). I'll see when i have time to check it, i also use Flypt for the interface.








// SMC3.ino with ResponsiveAnalogRead support for Arduino UNO
// Only modified sections shown; rest of SMC3 code remains unchanged

#include <EEPROM.h>
#include <ResponsiveAnalogRead.h>

// MotorData structure remains unchanged

// Create ResponsiveAnalogRead objects for each motor
ResponsiveAnalogRead feedbackReader[] = {
ResponsiveAnalogRead(A0, true), // Motor 1
ResponsiveAnalogRead(A1, true), // Motor 2
ResponsiveAnalogRead(A2, true) // Motor 3
};

// Optional: tune the responsiveness and noise filtering
void ConfigureResponsiveAnalogRead() {
for (int i = 0; i < 3; i++) {
feedbackReader.setSnapMultiplier(0.01); // Lower = smoother
}
}

// Helper to get raw analog input
int GetRawAnalogValue(int pin) {
return analogRead(pin);
}

// Modified InitMotor to prime the filter
void InitMotor(MotorData &m, int Index) {
m.feedback = 0;
m.target = NEUTRAL_POS;
m.idle = 0;
m.pwmOut = 0;
m.disabled = true;
m.cumError = 0;
m.dTerm_x100 = 0;
m.lastPos = 0;
m.kdCounter = 0;
m.deadzone = 0;
m.cutoffMin = 50;
m.cutoffMax = INV(m.cutoffMin);
m.clipMin = 190;
m.clipMax = INV(m.clipMin);
m.centerOffset = 0;
m.kp_x100 = 120;
m.ki_x100 = 1;
m.kd_x100 = 10;
m.ks = 5;
m.pwmMin = 0;
m.pwmMax = 180;
m.pwmRev = 200;
m.number = Index + 1;
m.reverse = Reverse[Index];
m.invert = Invert[Index];
m.scale = Scale[Index];
m.feedbackPin = FeedbackPin[Index];
m.enaPinA = EnaPinA[Index];
m.enaPinB = EnaPinB[Index];
m.pwmPin = PwmPin[Index];

pinMode(m.feedbackPin, INPUT);
pinMode(m.enaPinA, OUTPUT);
pinMode(m.enaPinB, OUTPUT);
pinMode(m.pwmPin, OUTPUT);

feedbackReader[Index].update();
feedbackReader[Index].setValue(GetRawAnalogValue(m.feedbackPin));

m.idle = m.lastPos = GetFeedback(m);
}

// Modified GetFeedback to use filtered signal
int GetFeedback(MotorData &m) {
int index = m.number - 1;
feedbackReader[index].update();
int value = feedbackReader[index].getValue();
if (m.reverse) {
value = INV(value);
}
return value;
}

// Call this in setup()
void setup() {
InitMotor(Motor1, 0);
InitMotor(Motor2, 1);
InitMotor(Motor3, 2);

ConfigureResponsiveAnalogRead(); // NEW LINE to configure filters

// ... rest of setup code from original SMC3
}

SMC already uses special fast analog input. There are some odd instructions in the setup() routine that set it up. This library might do something more but I am skeptical.

Right now in SMCUtils there is a number on the bottom if the screen which will be around 4000. This is the frequency of the control loop. So every 250 micro seconds it updates the data. So if you get this code running look at that number and see if it will run faster. The max it will run is 4000 because there is code which delays if it’s less than 250us. Remove that to see a max number.

That being said there is a bug in the PID calculation. It is likely to be the cause of your issues. This is a long standing bug I suggest you try it.

https://www.xsimulator.net/communit...or-driver-and-windows-utilities.4957/page-134

there is a download on this page go back a page to look at the discussion. Maybe post your analog input question there to get more views.