Why compare with a similar component
The absolute temperature of a component depends on load, ambient conditions, ventilation and emissivity. Comparing against an identical component subjected to the same conditions cancels almost all of that and leaves visible only what is different: the defect.
How it is done
- Choose the reference: the typical case is the three phases at the same connection point, or twin units with similar load.
- Ensure comparable conditions: same current, same exposure to air and same geometry. A phase with less load will always look colder.
- Use the same parameters: similar emissivity, distance and angle for both points, ideally in the same image.
- Measure at the equivalent point on each component, not the maximum of one and the average of the other.
- Calculate the difference between the two readings and record it together with the temperatures.
Difference versus rise over ambient
Rise over ambient is the difference between the component and the air around it. It is useful when no comparable counterpart exists, but it depends more on load and on room conditions. Comparison between similar components is more sensitive to a point fault.
When it does not apply
- If all the components compared suffer from the same problem, for example a general overload, the difference will be small and misleading.
- If the reference is not truly healthy, the result underestimates the defect.
- If the loads are very different, the difference reflects the load and not the condition.
How to interpret the result
Use both comparisons when possible, note the load during measurement and remember that the heat produced increases with the square of the current. The action levels for each difference are set by the applicable procedures or standards, in their current edition.