Temperature rise measurements are reported to a tenth of a kelvin and are frequently wrong by several. The instrument is rarely the problem. The thermocouple is a small thermal system that has to come into equilibrium with a surface it is only touching, and every part of that system adds a lag or an offset.
Understanding the lag turns a mysterious disagreement between two laboratories into a predictable one.
The three time constants
| Element | Typical value | Effect on the reading |
| Wire diameter | 0.1 mm to 0.5 mm | Thicker wire conducts heat away from the junction |
| Bead diameter | 2 to 3 times the wire diameter | Bead mass sets the response time |
| Bare junction, still air | A few seconds to tens of seconds | Fast enough for most rise tests |
| Insulated junction | Two to three times slower | The insulation is a thermal resistor |
| Attachment | Welded, cemented, clamped or taped | Sets the contact resistance to the surface |
| Reading stability | Within 0.5 K over one minute | The practical criterion for a valid reading |
A 0.1 mm wire with a welded bead reaches 63 percent of a step change in about a second. A 0.5 mm wire with a taped bead takes ten times longer, and a temperature rise test that ends before equilibrium then reports a value that is low by however much the junction had left to climb.
Why the attachment dominates
Once the junction is fast enough, the attachment becomes the limiting factor. A thermocouple taped to a terminal with a piece of glass tape is separated from the metal by the tape thickness, and that layer is a thermal resistance in series with the measurement. The junction reads the tape surface, which is cooler than the terminal because it is losing heat to the air.
The error is not constant. It grows with the temperature difference between the terminal and the ambient, which is exactly the quantity being measured. A 10 K rise may read 9.2 K and a 45 K rise may read 40 K, and the second one fails a product that is compliant.
Attachment methods, ranked
- Welded or brazed to the surface: best thermal contact, but destroys the plating and is not always allowed on a production sample.
- Cemented with a filled adhesive: good contact, needs curing time, and the adhesive has a maximum temperature.
- Clamped under a screw or a spring: good and repeatable, provided the clamp does not change the current path.
- Taped: the worst option that is still common. Acceptable only for a rough indication.
Reading a rise test correctly
Wait for stability, not for a fixed time. The criterion that matters is a change of less than 0.5 K over one minute at constant load and constant ambient. Record the ambient at the same time and in the same place, because a rise is a difference between two measurements and either one can be wrong.
Then check the reference point. A terminal temperature is measured on the terminal, not on the wire insulation next to it, and not on the screw head if the standard names the conductor. Getting that wrong changes the number by more than any instrument error.
A temperature rise test system with a multi-channel recorder and fine-wire thermocouples is the practical setup, and the site index lists the thermocouple types and wire gauges by application.
Standards referenced: IEC 60584-1; IEC 61010-1.
