A thermocouple produces a tiny voltage that varies with temperature, but that raw millivolt signal is far too small and too nonlinear to feed into an ordinary analog input. A thermocouple input module is the specialized card built to make sense of it - amplifying microvolt-level signals, correcting for the temperature at its own terminals, converting the reading through a lookup table, and watching for a broken sensor. This guide covers the TC input card as hardware, which is distinct from what a thermocouple sensor is: cold-junction compensation, millivolt sensing, linearization, and burnout detection.
Thermocouple Input Module in one line: A thermocouple input module (TC input card) is a specialized PLC or RTU module that reads the small millivolt signal a thermocouple generates and converts it into a temperature value. It performs cold-junction compensation to correct for the temperature at its own terminals, uses a linearization table to convert the nonlinear voltage into degrees, and includes burnout detection to flag an open or broken thermocouple.
A thermocouple works because two dissimilar metals joined at a measuring point produce a voltage that changes with the temperature difference between that point and the other end where it connects to the instrument. The signal is tiny - on the order of tens of microvolts per degree - so a thermocouple input module begins with a high-gain, low-noise front end able to resolve those microvolts without letting electrical noise swamp the reading. This is why you cannot simply wire a thermocouple to a general analog input expecting a current or a volt; the levels are orders of magnitude smaller.
The catch is that the thermocouple only reports a difference, so the module must know the temperature of its own terminals - the cold junction, or reference junction - to work out the absolute temperature at the tip. Cold-junction compensation solves this: the module measures its terminal temperature with a built-in sensor and mathematically adds back the offset the connection point introduces. Without it, the reading would drift as the enclosure warms and cools through the day. Good TC cards place the reference sensor right at the terminals and compensate continuously, which is a large part of why a dedicated card exists at all.
A thermocouple's voltage-to-temperature relationship is not a straight line - the millivolts per degree change across the range and differ for each thermocouple type, such as type K, J, or T. Turning the compensated millivolt reading into an accurate temperature therefore requires linearization: the module applies the standard reference table for the selected thermocouple type, converting voltage into degrees along the correct curve. This is why the card must be told which type of thermocouple is connected on each channel; select the wrong type and the linearization table is wrong, so the temperature is wrong even though the hardware is fine.
Because thermocouples are thin wires exposed to heat, vibration, and corrosion, they eventually fail open, and a card that ignored that could report a stale or ambiguous value. Burnout detection addresses this: the module drives a very small current through the sensor and watches for the open circuit that appears when the thermocouple breaks, then flags the channel and can drive the reading to a configured upper or lower limit so the fault is obvious. That behavior is configurable precisely so the failed reading fails toward the safe direction for the process - high for something that must trip on overtemperature, for instance - rather than sitting at a plausible mid-range value that hides the break.
The temperatures a thermocouple input module produces feed directly into the process picture a SCADA system presents. Heater-treater and line-heater temperatures, compressor and engine readings, and process-vessel measurements often start as thermocouple millivolts on a TC card before being scaled and polled up to a platform like Merobix for trending and alarming. When an operator watches a temperature climb toward a limit from a remote dashboard, a thermocouple input channel is doing the sensing behind that trend.
The card's diagnostics are what make those remote readings dependable. Burnout detection and out-of-range flags let the SCADA layer distinguish a genuinely hot process from a broken sensor, so an alarm reflects reality rather than a failed wire. On unmanned sites, that distinction is valuable - it decides whether an operator dispatches a crew for a real high-temperature event or ignores a nuisance reading, and it prevents a silently failed thermocouple from masking an overtemperature condition nobody is on site to catch.
A thermocouple only measures the temperature difference between its tip and the terminals where it connects, so the module must know the terminal temperature to find the absolute reading. Cold-junction compensation measures that terminal, or reference-junction, temperature and mathematically corrects for it. Without it, readings would drift as the enclosure temperature changed.
A thermocouple produces only tens of microvolts per degree - far too small for a standard 4-20 mA or voltage input to resolve accurately. It also needs cold-junction compensation and type-specific linearization. A thermocouple input module provides the high-gain millivolt front end, the reference-junction correction, and the linearization tables that a general analog input lacks.
Thermocouples eventually break and fail open. Burnout detection drives a tiny current through the sensor and watches for the open circuit that appears when the wire breaks, then flags the channel and drives the reading to a configured high or low limit. This makes a failed sensor obvious instead of leaving a misleading mid-range value on the screen.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.