Skip to main content
IronOS-powered soldering irons use a modified N-type thermocouple integrated into the tip for temperature measurement. This design creates a temperature sensor at minimal cost but introduces unique measurement challenges.

Thermocouple Design

The thermocouple is constructed by using different metal types to join one of the rings to the heating coil, effectively creating a free temperature sensor with very low construction difficulty.

Advantages

  • Extremely low cost
  • Simple construction
  • Integrated into the tip design

Limitations

  1. Non-optimal metals: Uses available materials rather than calibrated thermocouple alloys
  2. Non-constant response: Temperature response varies across the range
  3. Shared connections: Same connections used for heating current, preventing simultaneous heating and measurement

How Thermocouples Work

Thermocouples use a junction of two dissimilar metals to generate a small voltage (microvolts). This voltage can be measured and converted to temperature using a known transfer function.

Cold Junction Compensation

The copper terminations where thermocouple wires connect form additional junctions that also generate voltages. These are called the cold junction.
  • If held at 0°C, cold junction effects are negligible
  • At room temperature, measured voltage must be compensated
  • Every metal-to-metal junction in the circuit acts as a small thermocouple

Temperature Measurement Implementation

The hardware design (Pinecil, TS100, TS80) uses a consistent measurement approach:
  1. Op-amp connected directly across heating connections to the tip
  2. Handle temperature sensor for reference measurement
  3. ADC sampling when heating is off

Measurement Process

When the iron is not heating:
  1. Microcontroller reads op-amp output via ADC
  2. Voltage represents temperature difference (tip - handle)
  3. Apply offset compensation (removes ADC and op-amp offsets)
  4. Convert to temperature delta in °C/K
  5. Add handle temperature to derive tip temperature
Tip Compatibility: TS100 and Pinecil tips are approximately equivalent to Hakko T12 tips (within measurement error). This compatibility exists because the TS100 design mimicked the excellent T12 tip construction.

Temperature Control Implications

Heating vs Reading Tradeoff

Since temperature can only be measured when not heating, there’s a critical tradeoff:
  • More frequent measurements = Better temperature control, less heating time
  • Fewer measurements = More heating power, slower response
IronOS Solution: Implements a two-speed controller loop:
  1. Heat-up mode: Fewer temperature measurements, maximum power delivery
  2. Maintenance mode: Double measurement rate for faster reaction times

Internal vs External Temperature

The thermocouple measures the inside of the tip (near the heater coil), not the outside surface where soldering occurs.
Thermal Lag: The internal temperature reading lags behind the external tip temperature during rapid changes.

PID Tuning Philosophy

The PID controller is intentionally tuned to be slightly underdamped (more “jumpy” than typical): Rationale: When internal temperature drops, the external temperature has dropped more significantly. Overcompensation helps the system catch up faster. Result: Temperature display may fluctuate during use, but the thermal mass of the tip smooths these variations for stable soldering performance.
IronOS displays the actual measured temperature rather than hiding variations within a tolerance band. Some commercial controllers (e.g., Weller) hide ±3-5°C fluctuations to create an illusion of perfect stability.

Temperature Stabilization

With no external cooling, internal and external tip temperatures equalize in 3-5 seconds.
When testing temperature accuracy, allow 3-5 seconds for thermal stabilization.

Measurement Complexity and Accuracy

Voltage Reference Limitations

The firmware performs best-effort temperature calculation within hardware constraints:
  • Voltage reference accuracy: ±1% (affects all measurements)
  • No factory calibration on low-cost MCUs
  • Cannot compensate for reference drift
Accuracy Expectation: Don’t expect better than 1% accuracy due to fundamental hardware limitations.

Compensation Model

The firmware accounts for:
  1. Cold junction compensation (implemented)
  2. Constant offset (implemented)
Not modeled: Linear offset variation with handle temperature

Handle Temperature Effects

The unmodeled linear offset with handle temperature is generally acceptable because:
  • During use, handle temperature stabilizes within ~10°C
  • Error becomes relatively constant during normal operation
  • Target use case is primarily above 150°C
Low Temperature Anomaly: When tip temperature approaches room temperature (sub-100°C), you may observe:
  • Jumping and unstable readings
  • Control issues at low setpoints
This is a known tradeoff since the iron’s intended use is primarily above 150°C where these errors are not dominant.

Technical Specifications

Measurement Timing

  1. FET switches off (stops heating)
  2. Recovery time for tip capacitance discharge
  3. Op-amp exits saturation
  4. ADC samples 8 times quickly
  5. PWM output resumes
This creates a small dead time in the output signal during each measurement cycle.

Performance Optimization

The firmware balances:
  • Frequent sampling: Stable temperature control
  • Infrequent sampling: Maximum power delivery
See Power Management for details on power delivery optimization.

Resources

Build docs developers (and LLMs) love