Overview

This project tests and compares the ADC performance of three popular platforms:

• Arduino Uno — 10-bit internal ADC
• ESP32 — 12-bit internal ADC
• ADS1115 — 16-bit external I²C ADC

The main focus is linearity, noise, and measurement stability using identical input voltage for all ADCs.

Test Setup

The test setup ensures fair comparison and accurate measurements:

• Input Voltage: 0 V–3.3 V adjustable using a multi-turn potentiometer (fine control down to 1 mV)
• Buffering: Voltage follower op-amp prevents loading effects
• Noise Reduction: RC low-pass filter minimizes input noise
• Power Supply: Lead-acid battery to avoid SMPS switching noise
• ADC Timing: Minimum 20 ms between consecutive readings ensures proper settling

Circuit Diagram:

Data Collection Method

• ESP32 collects ADC data from Arduino Uno (UART), ADS1115 (I²C), and its own ADC.
• All readings are sent to PC via USB and logged in a CSV file using Python.
• Two main experiments:

1️⃣ Linearity Test

• 88 voltage points, 100 samples per point
• Plots show ADC linearity and allow ESP32 calibration

2️⃣ Noise Test

• 5 fixed voltage points, 1000 samples per point
• Histograms and standard deviation calculated for stability evaluation

Source Codes

All codes are available on GitHub:

• Arduino Uno: adc_test_arduino_uno.ino
• ESP32: adc_test_esp32.ino
• Python Logger: adc_test_N_samples.py

Linearity Results

Linearity Comparison:

• Arduino Uno & ADS1115 almost perfectly linear
• ESP32 shows non-linearity with dead zone <0.13 V and saturation >3.18 V

ESP32 2-Point Calibration (0.501 V & 2.000 V):

• Improves usable range to ~0.15 V–2.5 V

Noise Results (Noise + Histogram)

ESP32: Highest fluctuation, wide histogram — filtering & calibration required

Arduino Uno: Stable but quantization steps visible due to 10-bit resolution

ADS1115: Most stable, lowest noise

Standard Deviation Comparison

Numeric comparison of noise level: lower SD = better stability