Analog smoothing using a weighted average filter
 by Tom Igoe

 Based on notes by David Schultz, at


 This program reads an analog input and smoothes out the result using a
 weighted average  filter. It works by taking a weighted average of the
 current reading  and the average of the previous readings.

 This example uses a TMP36 temperature sensor on Analog pin 1.  A0 is the
 TMP36's ground, and A2 is its power pin.

 In this example, a second analog reading,
 attached to a trimmer potentiometer, sets the weight.  When the trimmer pot
 is set high, the average is weighted in favor of the current reading, and
 almost no smoothing is done.  When the trimmer pot value is low, the average
 is weighted in favor of the previous readings, and the current reading
 affects the average very little.

 n.b. the variable "lastEstimate" needs to be a global, since it's modified
 each time a new filtering is done.  So if you want to use this for multiple
 inouts, you'll need a "lastEstimate" variable for each input.

 Created 17 October 2005
 Updated 6 Feb 2012


float lastEstimate = 0;      // previous result

void setup() {
  // use pins A0 and A2 to power the TMP36 sensor:
  pinMode(A0, OUTPUT);
  pinMode(A2, OUTPUT);
  // set A0 low, and A2 high:
  digitalWrite(A0, LOW);
  digitalWrite(A2, HIGH);

void loop() {
  // read the sensor:
  int sensorVal = analogRead(A1);
  // convert to voltage:
  float voltage = 5.0 * sensorVal / 1024.0;
  // convert to degrees celsius:
  float temperature = (voltage - 0.5) * 100;
  // read the trim pot:
  float trimPotValue = analogRead(A3)/1023.0;

  // filter the sensor's result:
  float currentEstimate = filter(temperature, trimPotValue, lastEstimate);
  // print the result:
  // save the current result for future use:
  lastEstimate = currentEstimate;

// filter the current result using a weighted average filter:
float filter(float rawValue, float weight, float lastValue) {
  // run the filter:
  float result = weight * rawValue + (1.0-weight)*lastValue;
  // return the result:
  return result;