Issue link: https://iconnect007.uberflip.com/i/663184
70 The PCB Magazine • April 2016 Dielectric breakdown voltage varied by ma- terial and test region. Laminate A breakdown voltages remained steady up to T3 (185°C), where the voltages in all regions sharply dropped. Laminate D breakdown voltages re- mained fairly steady in all aging temperatures and test regions. Laminate B breakdown volt- ages remained fairly steady in all aging temper- atures and test regions. Laminate C breakdown voltages demonstrated a decline in strength, particularly at higher temperatures. Laminate E breakdown voltages declined steadily, but unex- pectedly recovered at T4 (280°C). The increase in Laminate E breakdown volt- ages at T4 corresponds with a sharp increase in weight loss at that temperature, as well in- creased delamination and general degradation of the board condition. The higher breakdown voltages could be caused by a number of fac- tors, and do not necessarily indicate a higher dielectric strength in the material at that tem- perature. Heavy delamination may have re- sulted in the copper components of the board being exposed to more air, causing formation of copper oxides and degrading the test circuit's ability to conduct electricity. The markedly in- creased weight loss, -9% at 280°C compared to -1% at 260°C, indicates the higher temperature may be causing certain compounds in the ma- terial to decompose or react in ways that aren't possible at lower temperatures. The high degree of delamination, degradation of copper compo- nents, and general changes in the physical ge- ometry and condition of the board may have altered the way the voltage is applied during the test. It was later determined after discussions with Laminate E supplier that the temperatures selected for Laminate E were too high. This was taken into consideration when selecting tem- peratures for the 1000-hour test. It was theorized here that the influence of the PWB heterogeneous stack up vs. the lami- nate manufacturer's homogeneous stack up will have a significant influence on the change in robustness of the materials at elevated tempera- tures. This difference highlights the need and value of performing thermal endurance testing on a manufactured PWB. By doing so, the pos- sible influence of different manufacturers and process sets are also taken into account. Dielectric strength was calculated by divid- ing dielectric breakdown voltage by thickness tested. % retention was calculated using the fol- lowing equation: Where DS T0 = Dielectric Strength Time Zero and DS T552 = Dielectric Strength at 552 hours. A plot was then created of % retention (x) vs. temperature (y) for each laminate which is presented in Figure 4a through 4e below. Data is fitted to a line or curve. In the data present- ed below, regression was used to determine the equation that fits the data best. % retention value desired (y) is substituted into the equa- tion to find the corresponding temperature (x) to predict the highest temperature to be used in the 1,000-hour aging test. Data can be fit mul- tiple ways, linear, polynomial, or logarithmic. long-term thermal reliability of pCb materials Figure 6: Average % change in weight for 500-hour test vehicles. Figure 7: Average % change in thickness for 500-hour test vehicles.