A robot cook machine's thermal control system is a closed-loop feedback circuit that uses microcontroller-driven PID algorithms to maintain precise cooking temperatures by dynamically adjusting power to a heating element based on real-time sensor data. By replacing a basic bimetallic thermostat with an ESP32-based proportional controller, you change a reactive, high-overshoot heating circuit into a predictive power modulation system that holds temperatures within ±0.5°C. Beginners commonly confuse this PID (Proportional-Integral-Derivative) tuning with simple hysteresis (bang-bang) control, or they mistakenly assume the microcontroller's high-frequency PWM signal can directly drive an AC Solid State Relay (SSR) without a zero-crossing detector.
Component Architecture and Hardware Spec Sheet
To achieve sub-degree thermal stability in a robot cook machine, you cannot rely on cheap thermistors or mechanical relays. The physical plant (the heater and the food) has significant thermal mass and lag. You need a high-resolution sensor, a dedicated analog-to-digital front end, and a zero-crossing SSR to switch AC mains without generating massive electromagnetic interference (EMI) that will reset your microcontroller.
| Component | Model / Part Number | Key Specification | Role in Control Loop |
|---|---|---|---|
| Microcontroller | ESP32-WROOM-32U | Dual-core 240MHz, 520KB SRAM | Executes PID math at 10Hz, handles WiFi telemetry |
| RTD Amplifier | MAX31865 | 15-bit ADC, SPI interface, 3-wire support | Converts PT100 resistance to digital temperature data |
| Temp Sensor | PT100 Class A (3-wire) | ±0.15°C at 0°C, 3850 ppm/°C | Provides linear, high-accuracy process variable (PV) feedback |
| AC Switch | Crydom D2440 (Zero-Cross) | 40A RMS, 24-280VAC, 8.3ms max turn-on | Switches 1500W heating element via time-proportioning control |
The choice of a 3-wire PT100 over a standard 10k NTC thermistor is critical here. NTC thermistors suffer from self-heating and non-linear resistance curves at high temperatures. The PT100 remains highly linear up to 300°C, and the 3-wire configuration allows the MAX31865 to cancel out lead wire resistance, which is vital when the ESP32 control board is mounted 50cm away from the cooking vessel.
PID Theory vs. Bang-Bang Control (With Numeric Example)
Most commercial slow-cookers use bang-bang control: if the temperature is below the setpoint, the heater runs at 100%; if it is above, it turns off. This creates a continuous sawtooth temperature wave. In a robot cook machine executing delicate tasks like tempering chocolate or sous-vide cooking, a 5°C swing is unacceptable.
PID control solves this by calculating three distinct terms:
- Proportional (P): Reacts to the current error (distance from setpoint).
- Integral (I): Reacts to the accumulation of past errors (eliminates steady-state offset).
- Derivative (D): Reacts to the rate of change (predicts future error and applies the brakes).
Worked Numeric Example: Thermal Overshoot
Imagine heating 2 liters of cooking oil (mass = 1.84 kg, specific heat capacity = 2.0 J/g°C) from 20°C to a target of 180°C using a 1500W cartridge heater.
Energy Required: Q = mcΔT = 1840g × 2.0 J/g°C × 160°C = 588,800 Joules.
Time at 1500W: 588,800 J / 1500W ≈ 392 seconds (6.5 minutes).
The Bang-Bang Failure: Your PT100 sensor has a thermal lag of 5 seconds relative to the bulk oil. When the sensor finally reads 180°C, the oil is actually hotter, and the heater has been dumping 1500W into the system for 5 extra seconds. That is 7,500 Joules of uncommanded energy. Divided by the thermal mass (3680 J/°C), this causes a 2.04°C overshoot before the heater even turns off. If the heating element itself retains 15,000 Joules of residual heat, your overshoot easily exceeds 6°C, burning the food.
The PID Solution: The Derivative term detects the rapid rate of temperature rise at 172°C. The PID algorithm drops the SSR duty cycle to 15%, feeding only 225W into the system for the final approach, allowing the thermal mass to settle exactly at 180.0°C with zero overshoot.
For implementation, the Arduino PID Library by Brett Beauregard remains the gold standard for embedded C++ environments, handling the complex millis() timing and derivative-on-measurement calculations required to prevent derivative kick.
Where You Meet This In Practice: Mains Wiring and SSR Drive
Theory falls apart if the hardware execution is flawed. The most common point of failure in DIY robot cook machines is the interface between the 3.3V ESP32 GPIO and the 240V AC heating circuit.
⚠️ MAINS VOLTAGE HAZARD: Working with 120V/240V AC heating elements carries a lethal shock and fire risk. Always de-energize the circuit at the breaker panel, apply a lockout/tagout device, and verify the circuit is dead with a CAT III rated multimeter before touching any terminals. Local electrical codes (NEC Article 424 for fixed electric heating) may require this installation to be performed by a licensed electrician.
You must use a Zero-Crossing SSR (like the Crydom D2440) rather than a random-fire SSR. A zero-crossing SSR only switches the AC load when the sine wave crosses 0V. This minimizes inrush currents and prevents the massive voltage spikes that would otherwise couple back into your ESP32's power supply, causing brownouts and watchdog resets.
SSR Heat Sinking Calculation:
A zero-crossing SSR has an internal voltage drop of roughly 1.2V. At a continuous 15A load (typical for a 1500W heater on a 120V circuit), the SSR dissipates 18W of heat (1.2V × 15A). Without a heat sink, a standard panel-mount SSR will reach its maximum junction temperature and fail open or short in minutes. You need a heat sink with a thermal resistance of at least
Tuning Edge Cases and Embedded FAQs
Even with perfect hardware, a poorly tuned PID loop will oscillate wildly. When tuning your robot cook machine, use the Ziegler-Nichols method: set I and D to zero, increase P until the temperature oscillates evenly, note that ultimate gain (Ku) and period (Pu), and calculate your final PID values from there.
Frequently Asked Questions
Why does my heater click rapidly and my ESP32 keep rebooting?
You are likely using a mechanical relay or a random-fire SSR with high-frequency PWM. The ESP32 is experiencing EMI-induced brownouts. Switch to a zero-crossing SSR and use time-proportioning control (switching the SSR on and off in 1-second to 5-second windows) rather than hardware PWM.
What is Integral Windup and how do I stop it?
Integral windup happens when the heater is maxed out at 100% duty cycle for a long time (like heating cold water from the tap). The 'I' term accumulates a massive error value. Once the water finally reaches the setpoint, the 'I' term is so large it keeps the heater on, causing massive overshoot. Fix this in your code by setting the PID library's SetOutputLimits(0, 255) and enabling anti-windup logic, which stops integrating when the output is saturated.
Can I use a thermocouple (Type K) instead of a PT100?
You can, using a MAX6675 or MAX31855 amplifier, but thermocouples are better suited for extreme heat (ovens, kilns >400°C). For a robot cook machine operating between 40°C and 250°C, the PT100 provides vastly superior resolution and stability, which makes the PID derivative math much smoother and prevents erratic heater toggling.
Designing the thermal loop for a robot cook machine bridges the gap between abstract control theory and high-power electrical reality. By respecting the thermal mass of your ingredients, selecting linear RTD sensors, and properly isolating your AC switching hardware, you build a system that doesn't just heat food, but actively understands and manages the thermodynamics of the cooking process.
