Reflow Thermal Recipe Segment Optimization Model Based on Artificial Neural Network Approach

The temperature settings for the reflow oven chamber (i.e., recipe) are critical to the quality of the Printed Circuit Board (PCB) in the surface mount technology because solder joints are formed on the boards with the placed components during the reflow process. Inappropriate profiles cause various defects such as cracks, bridging, delamination, etc. Solder pastes manufacturers have generally provided the ideal thermal profile (i.e., target profile), and PCB manufacturers have attempted to meet the given profile by fine-tuning the oven’s recipe. The conventional method tunes the recipe to gather thermal data with a thermal measurement device and adjust the profile relying on the trial-and-error method. This method took a lot of time and effort, and it cannot guarantee consistent product quality because it’s so dependent on the engineers. We proposed (1) a stage-based (ramp, soak, and reflow) input data segmentation method for data preprocessing, (2) a model for predicting the zone temperature in the soldering reflow process (SRP) using a state-of-the-art machine learning, (3) an algorithm for generating the optimal recipe to reduce the gap between the actual processing profile and the target profile. Our method uses artificial intelligence, specifically a backpropagation neural network, to enable non-contact prediction using thermal data from a single experiment (BPNN). In the fully equipped in-house laboratory, the validity of the approach was tested. As a result, within 10 min of starting the experiment, the generated optimal recipe shows 99% fitness to the targeted profile.