Heater Thermocouple Open Detection: Keep Test Current Out of Temperature Readings

System boundary

The thermocouple attached to a thick-film heater, its wiring, input protection/filter network, ADC configuration and controller sample handling. Cold-junction compensation and physical hot-junction attachment are separate established reviews.

Integration interfaces

Integration risks

Distinguish the continuity challenge from the temperature measurement

Identify how the actual converter or external circuit detects an open thermocouple. Some arrangements use continuous resistor biasing; others enable a small current source for a separate diagnostic conversion. Those arrangements produce different errors and timing requirements. Record the circuit and configuration rather than relying on a generic sensor-break feature name.

For a switched diagnostic, give each sample an explicit mode identity. A high diagnostic reading may be an expected response to an open input, not a hot-junction temperature. Conversely, a plausible normal reading does not establish continuity unless the selected topology makes that conclusion valid. Keep the electrical fault test distinct from the cold-junction compensation calculation.

Follow test current through the series resistances

Draw the diagnostic current path from its source to its return. Include the thermocouple leads, any series input-filter resistors, connector resistance and relevant multiplexer paths. The voltage added to the measured differential input depends on where the current flows and which nodes the converter observes. Two equal resistors do not cancel their drops when the test current drives through both in the same loop.

For a deliberately simplified loop with equal source and sink current, the added differential voltage is diagnostic current multiplied by total series loop resistance. An assumed 1-microampere current through 400 ohms adds 400 microvolts. With an illustrative local thermocouple sensitivity of 40 microvolts per kelvin, that voltage corresponds to a ten-kelvin apparent shift if incorrectly treated as temperature. These assumed values are not a sensor specification or recommended diagnostic setting.

Choose a diagnostic challenge without assuming one current fits every harness

A diagnostic current must produce a distinguishable open-input response within the allowed time while respecting the converter input behavior. Increasing it can speed charging of input capacitance but also increases the disturbance applied to an intact channel. Input clamps, leakage and common-mode limits can alter the result before a nominal full-scale code appears.

Test the longest relevant harness and the actual filter network, not only a short bench jumper. Include the input state at diagnostic entry because capacitors may retain charge from the preceding conversion or channel. Use the selected device's specifications to define valid input and diagnostic conditions; do not use an arbitrary current merely because it yields an obvious alarm in one setup.

Budget recovery after disabling the test current

Removing the current source does not instantly remove the voltage stored on every input capacitor. In a single-pole illustrative recovery, a disturbance decays as its initial value multiplied by exp(−t/tau). Reducing that disturbance to one percent takes approximately 4.61 time constants. If tau were 2 milliseconds, the analog portion alone would require about 9.21 milliseconds under that simple model.

The converter's digital filter may retain diagnostic history after the analog input has settled. Multiplexer changes, gain changes and synchronization also have device-specific latency. Establish the first valid conversion from the actual acquisition sequence, not by adding one nominal sample period. Save a recovery trace showing the diagnostic disturbance and the accepted-temperature boundary.

Specify the measurement state machine at the sample boundary

Make the transition sequence reviewable by both hardware and firmware engineers. The timing should describe when the current changes and which conversion integrates each condition. A function call returning does not prove that the analog pin or digital filter has reached the requested state.

Test fault signatures without claiming complete sensor supervision

Exercise the proposed detection with controlled opens at relevant connector and lead locations under an approved test method. Check whether the result differs from a saturated but intact input, a shorted pair and a plausible voltage source. Identify any topology in which leakage or another connected channel holds an open input inside the normal range.

An electrically continuous thermocouple can be detached from the heater or shorted at a location that senses the wrong temperature. The open test does not prove proper thermal attachment, correct alloy or a valid reference junction. List those uncovered conditions explicitly in the system validation plan. A single diagnostic bit must not be relabeled as comprehensive heater protection.

Handle the gap in fresh temperature data deliberately

While the channel is in diagnostic or recovery mode, the last valid temperature becomes older. Define the maximum interval, how the controller tracks sample age and what happens if the next valid conversion never arrives. A held last value may be a control strategy for a bounded interval, but it is not a fresh measurement and cannot silently continue forever.

Coordinate diagnostic scheduling with the equipment's protective response requirements. Do not hide a real overtemperature condition behind repeated retries, prolonged averaging or a diagnostic state that suppresses every alarm. If an independent protective sensor is required, retain its independent acquisition and interruption path during the test. Firmware verification can begin with injected data before approved powered checks.

Retain mode-tagged raw data with the acceptance result

The final record should include the electrical schematic, test-current settings, series resistance range, filter values, conversion timing and state transitions. Keep raw input-referred voltage, mode tags, timestamps and published temperature together. This allows a reviewer to determine whether a temperature step came from the thermocouple or from a change in diagnostic configuration.

Repeat the relevant checks after a harness, filter, ADC rate or firmware scheduling change. A resistor substitution can alter both noise filtering and diagnostic recovery even if a room-temperature check still passes. For the heater inquiry, supply the valid temperature stream and fault response boundary; the useful deliverable is a verified sensing sequence, not a new claim about the thick-film element's operating range.