This Espressif ESP32-Powered Internet of Things Meat Thermometer Gets Your Food Done Just Right
A clever calibration step with a Maxim DS18B20 and a glass of boiling water lets this cooking tool make use of low-cost probes.
Pseudonymous maker "moeburn" has put together a low-cost Internet of Things (IoT) meat thermometer, powered by an Espressif ESP32 and with graphing capabilities — and the ability to use low-cost sensor probes without sacrificing accuracy.
"[The thermometer] features a graphing web UI [User Interface], estimates how many minutes until your meat is done, uses [a] super precise 16-bit ADC [Analog to Digital Converter, and can calibrate any cheap thermistor probe against a DS18B20 digital sensor," moeburn explains of the project. "All for about $30 in parts."
The heart of the device is an Espressif ESP32 microcontroller, but it's not alone. To improve precision, moeburn added a Texas Instruments ADS1115 16-bit analog to digital converter (ADC) — while support for low-cost probes was added by introducing a clever calibration step which compares the readings to a Maxim DS18B20 digital temperature sensor.
"[They] are very accurate and factory-calibrated by Maxim," moeburn explains, "but unfortunately are too thick to use as a meat thermometer on their own. "Placing both in a glass of boiling water then recording analog readings at 75, 50, and 25°C, we can derive the Steinhart-Hart equation coefficients, so the temperature probe is perfectly accurate across all temperature ranges."
The cooking accessory makes use of the Wi-Fi connectivity available on the ESP32 microcontroller to provide a web interface with a temperature graph and estimated-time-to-cook to take the guesswork away — "with, moeburn jokes, "Windows 98 File Copy Dialog levels of accuracy." Assistance on this part was provided by OpenAI's ChatGPT, the maker notes, "because I do not understand JavaScript at all."
There's even support for local alerts via a built-in OLED display and a piezoelectric speaker — with moeburn providing code for playing a 10-second snippet of Johnny Cash's Ring of Fire when the food's cooked.
The proof of the project, of course, is in the eating. "Just gave it a test on a piece of trout," moeburn writes of the thermometer's first field test, "and the fish was cooked so perfectly it literally fell off the skin. Matches up perfectly with my store-bought meat thermometer. The calibrator helped me use cheap probes. The ETA helped me time it with the vegetables. And the web app helped me watch it all from another room. Useful build."
More information is available in moeburn's Reddit post, while the source code for the project has been published to GitHub under an unspecified open source license.
