For my research on the topic of power cycling capability of IGBT modules, I need to measure the junction temperature of the IGBT dies inside the module. To acquire this temperature measurement, I found that the measurement of VCEon caused by a small reference current is commonly used as a temperature sensitive electronic parameter (TSEP). However, I can’t find useful information on the topic of repeatability of this measurement. In most literature where the topic is discussed, it is stated that calibration of a module is required and that the need for this calibration is due to the tolerances of the used materials, but nothing is stated about the extent of this calibration. Does the variation between modules (and the dies inside the modules) consist solely of a shift in the amplitude of the whole graph, or does the sensitivity also vary?
Up to a large number of cycles there are small changes in VCEon. What is the extent of these small changes? Is there a need to recalibrate the module after a certain number of cycles?
Dear Bram,
calibration is necessary because of deviations in producing power semiconductors and refers to every module within your test. The influence is not an offset in the graph Vce=f(T)@I=const. but may lead to a slight change in the graph's slope. Every module has to be calibrated individually and only once!
A further Recalibration after a certain period of testing would lead to misleading results of the final test.
Minor deviations during the measurement, usually in a range of millivolts, can be influences from aging thermal compounds, a real killer criteria is a growth in forward voltage by 25% representing hundreds of millivolt.
Cheers - Martin
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering