In my thesis I have summarised the differences of the reliability of BGA and LGA packaged components focusing on the geometrical and printed circuit board design factors and prepared a heat shock test for the reliability comparison.
In my work I have designed a test board for daisy chain LGA and BGA components. On the board, there are measurement pads around the components allowing resistance measurements on the solder joints and mounting pads for NTC thermoresistors which are used for counting the thermal cycles during the test.
For real-time resistance measurements, I have designed a measurement instrument based on an Arduino development board. The device counts the elapsed thermal cycles and makes records when the resistance of one of the monitored daisy chain sections increases. A record consists of the names of the measurement points of the failed daisy chain section and the number of the elapsed thermal cycles.
For configuring the measuring device, I have written a program with graphical user interface in Pascal language using Lazarus. The program communicates with the device through USB interface.
I have soldered LGA and BGA components to the testboard. To obtain LGA components, I have removed the solder bumps of the BGA components. For the removal process, I have placed the components on a bottom-side heater plate and used flux, desoldering wick and a soldering iron. After the soldering, I have examined the solder joints with an X-ray microscope and measured the resistances with four wire resistance measurement.
Finally, I have made resistance measurements using the self-designed measurement device and communication software. I have simulated a heat shock test by heating the thermal resistor. The measuring device successfully counted the number of the elapsed thermal cycles and recorded the failed daisy chain segments. After the experiment, I was able to read the recorded data from the device with the communication software, thus the reliability comparison of the LGA and BGA components is possible.