Issue link: https://iconnect007.uberflip.com/i/474043
50 The PCB Magazine • March 2015 350, 500, 700, 1000 & 1200mA). The results were then compared to similar materials cur- rently being used, and although this was not exactly a scientific approach, it was observed that TDS information was not always accu- rate when comparing materials (based on junction temperature). Note: The differences observed with TDS performance figures may be due to the various test procedures used and some figures maybe based on theoretical calculation. Another test programme included the build- ing of equipment capable of accurately and consistently testing the thermal performance of IMS materials. It was decided to adopt a similar technique used by one of the well-established IMS manufacturers and this involved using a TO-220 device under a controlled load as the heat source. A standard sample tile of the IMS is located onto a temperature controlled heat-sink (which maintains a constant temperature dur- ing the test cycle); the TO-220 is clamped onto the IMS and the temperature of the top and bottom of the IMS monitored until the TO-220 junction temperature remains constant (Figure 3). The thermal impedance for each material could be calculated using these test results and these results are now used to select the best ma- terial for a particular application based on per- formance and price (always a key element for LED lighting products). The equipment can be used to test the effectiveness of TIMs, heat sinks, expected operated temperature of a luminaire, etc., as well as being a standard test method for testing new IMS materials. As IMS materials become more widely used and understood, alternative ways of incorporat- ing this technology into other, non-standard constructions has increased and these include metal-core and hybrid builds (i.e., PTH or mul- tilayer boards bonded to a metal heat-sink using ceramic-loaded prepregs). THERMAL MANAGEMENT FOR LED LIGHTING APPLICATIONS continues Feature figure 3: Thermal impedance test equipment. (Source: Spirit Circuits)