SmoothThermistor

A thermistor‑reading library using the Beta model or the Steinhart–Hart equation.
This version modernizes the original library while preserving its basic constructor and behavior.

Original library:
https://github.com/giannivh/SmoothThermistor/

---

Installation

Place the library folder into Arduino/libraries/.
The public API preserves the original constructor for basic sketches.

---

Circuit

              Analog pin 0
                    |
5V |-----/\/\/\-----+-----/\/\/\-----| GND

           ^                ^ 
    10K thermistor     10K resistor

Optional AREF circuit:

      AREF      Analog pin 0
       |              |
3.3V |-+---/\/\/\-----+-----/\/\/\-----| GND

             ^                ^ 
      10K thermistor     10K resistor

---

Usage

Include the library:

#include <SmoothThermistor.h>

Basic constructor (namespace required):

smooththermistor::SmoothThermistor therm(A0);

Configuration‑based constructor:

smooththermistor::ThermistorConfig cfg;
cfg.adc_resolution = smooththermistor::AdcResolution::bit12;
cfg.nominal_resistance = 10000;
cfg.series_resistance = 10000;
cfg.beta_coefficient = 3950;
cfg.samples = 16;
cfg.filter_mode = smooththermistor::FilterMode::moving_average;

smooththermistor::SmoothThermistor therm(A0, cfg);

Optional AREF scaling:

therm.use_aref(true);

Read temperature:

double t = therm.temperature();

---

API Overview

Core Types

Type	Description
smooththermistor::SmoothThermistor	Main thermistor reader
smooththermistor::ThermistorConfig	Configuration struct
smooththermistor::AdcResolution	ADC resolution enum
smooththermistor::FilterMode	Filtering mode enum

---

ThermistorConfig Fields

Field	Type	Meaning
adc_resolution	AdcResolution	ADC bit depth
samples	std::size_t	Number of ADC samples
oversample_factor	std::size_t	Additional averaging
nominal_resistance	double	Thermistor nominal value
series_resistance	double	Series resistor
beta_coefficient	double	Beta constant
use_full_steinhart	bool	Enable A/B/C coefficients
sh_a, sh_b, sh_c	double	Steinhart–Hart coefficients
filter_mode	FilterMode	Filtering method
enable_noise_estimation	bool	Track noise variance
stability_threshold	double	Stability detection threshold

---

SmoothThermistor Methods

Method	Returns	Description
temperature()	double	Current filtered temperature
tick()	bool	Non‑blocking update step
last_filtered()	double	Last filtered temperature
is_stable()	bool	Stability flag
last_noise_stddev()	double	Noise standard deviation
last_noise_variance()	double	Noise variance
use_aref(bool)	void	Enable AREF scaling

---

Non‑Blocking Engine

The library supports a non‑blocking mode using tick():

if (therm.tick()) {
    double t = therm.last_filtered();
}

tick() performs:

• oversampling
• ADC accumulation
• resistance calculation
• filtering
• noise estimation (optional)
• stability detection (optional)

without blocking the main loop.

The synchronous temperature() method remains available for drop‑in compatibility.

---

Filtering Modes

Moving average

Sliding window average over recent raw temperature values.

Median

Sorts the history window and selects the median. Useful for rejecting spikes.

Exponential

Weighted filter using exp_alpha. Faster response with less smoothing.

---

Steinhart–Hart Mode

cfg.use_full_steinhart = true;
cfg.sh_a = 1.009249522e-03;
cfg.sh_b = 2.378405444e-04;
cfg.sh_c = 2.019202697e-07;

---

Example Snippets

Moving average + noise estimation

smooththermistor::ThermistorConfig cfg;
cfg.filter_mode = smooththermistor::FilterMode::moving_average;
cfg.samples = 20;
cfg.enable_noise_estimation = true;

smooththermistor::SmoothThermistor therm(A0, cfg);

void loop() {
    Serial.println(therm.temperature());
    Serial.println(therm.last_noise_stddev());
}

Steinhart–Hart mode

smooththermistor::ThermistorConfig cfg;
cfg.use_full_steinhart = true;
cfg.sh_a = 1.009249522e-03;
cfg.sh_b = 2.378405444e-04;
cfg.sh_c = 2.019202697e-07;

smooththermistor::SmoothThermistor therm(A0, cfg);

Custom ADC reader

cfg.adc_reader = []() {
    return analogRead(A0);
};

---

Filtering Pipeline

ADC read
↓
Oversampling (optional)
↓
Resistance calculation (Beta or Steinhart–Hart)
↓
Filtering (moving average / median / exponential)
↓
Noise estimation (optional)
↓
Stability detection (optional)
↓
Temperature output

---

Compatibility

The original constructor remains:

smooththermistor::SmoothThermistor therm(A0);

Existing sketches only need to add the namespace.

---

ADC Resolution

smooththermistor::AdcResolution::bit8
smooththermistor::AdcResolution::bit10
smooththermistor::AdcResolution::bit12
smooththermistor::AdcResolution::bit16

Default: 10‑bit.