IEC 61010
 
IEC 61010

Company News

Thermocouple Attachment for Temperature Rise: The Cement Trap

A temperature rise test produces a number that looks like a property of the product and is partly a property of the laboratory. IEC 61010-1 sets the limits and the steady-state criterion; the thermocouple, its wire, and the way it is fixed to the surface all filter the temperature being measured, and the filter is not small: attachment alone can account for several kelvin on a measurement whose limit has a margin of ten.

What the measurement is supposed to represent

The test asks how hot a surface or a winding becomes in service, and compares that against a limit that protects against insulation degradation, burns to the user, and fire. The thermocouple is a proxy. Three physical effects stand between the true temperature and the reading:

  • Thermal resistance of the attachment. Cement, adhesive or tape between the junction and the surface drops part of the temperature difference.
  • Heat conducted away by the wires. A thick thermocouple wire is a fin that carries heat out of the junction and reads low.
  • Radiant exchange with the surroundings. A junction exposed to a cooler enclosure wall loses heat by radiation and reads low, even when it is in contact.

Wire gauge: the fin effect in numbers

Wire diameterTypical responsePractical effect
0.1 mmUnder a secondMinimal fin loss; fragile and hard to attach
0.25 mm1 to 2 sGood compromise for surface work
0.5 mm2 to 4 sNoticeable fin loss on small parts; robust
1.0 mm5 to 10 sReads several kelvin low on small features

The fin effect scales with the wire cross-section and inversely with the length in contact with the hot surface. Short, thick wires read low; long, thin wires read closer to true. Where the standard does not specify a wire size, 0.25 mm is a defensible default for surface measurement and 0.1 mm for small windings.

The cement trap

Attaching a thermocouple with a blob of cement is the most common method and the most common source of error. The cement is an insulator, and it sits exactly where the measurement needs to be:

  • A thick blob adds thermal resistance between the surface and the junction, so the reading lags and reads low during a ramp and low at steady state.
  • A thin layer is better, but only if the junction is actually in contact with the surface. A junction sitting in a cement shell 0.5 mm above the surface is measuring the cement, not the part.
  • Cement that shrinks on curing pulls away from the junction and creates an air gap that grows over time.

The alternative methods and their trade-offs:

  1. Welded or brazed to a metal surface. The most accurate and the most invasive. Acceptable on a test sample, not on a production unit.
  2. Adhesive tape over a fine wire. Fast, repeatable, and good on flat surfaces. The tape thickness matters; a single layer of thin tape is roughly 0.05 mm and adds little.
  3. Spring-loaded contact probe. No adhesive, good repeatability, and reads the surface rather than the air. Useful for flat metal surfaces.
  4. Embedded in a winding. For windings the junction must be between turns, not on the outside of the coil, or it reads the surface of the insulation rather than the hottest spot.

Placement rules that matter

  • Measure the hottest point, which is rarely the most convenient. On a winding it is inside; on a heatsink it is at the device case; on a terminal it is under the screw.
  • Keep the wire run away from hot surfaces other than the measurement point, or the wire conducts heat into the junction and reads high.
  • Shield the junction from radiant exchange with a cooler wall when the surface is small and exposed.
  • Record the attachment method in the report. Two laboratories measuring the same product with different methods will differ, and the method is the explanation.

Steady state, and how long to wait

Temperature rise is measured at steady state, defined as a change of less than a defined amount over a defined period — typically 1 K per hour on the standard's wording, or 2 K per hour for some product standards. Two practical issues:

  1. A large thermal mass takes hours. A cast enclosure with a transformer inside can take four hours to reach steady state. Extrapolating from the first hour is a common shortcut that understates the final rise.
  2. Ambient drift invalidates the run. A test room that warms by 3 K during the run changes the delta. Record ambient at the start, during and at the end; if it moves more than 2 K, the run is questionable.

Reporting

The number alone is not the result. Report the measured temperature, the ambient at the time, the attachment method, the wire diameter and the location. With those five items, a second laboratory can reproduce the measurement; without them, it can only reproduce the conclusion.

The test equipment range includes thermocouple sets and surface probes for thermal testing; the site index links the data loggers and reference thermometers used to calibrate them.

    Standard:       IEC 61032 IEC 60061 IEC 60335 IEC 60529 IEC 60068 IEC 60950 IEC 61010 IEC62368 IEC 60601
    TAG:
Test Probe Go Not Go Gauges Spring Impact Hammer Test Pin Probe Test Probe B Test Probe 18 IP Test Probe
 
Lampholder Gauges Lampcap Gauge Test Probe 11 UL Test Finger Probe Test Vessel Rigid Test Finger Impact Testing
test probe