Low-Cost Non-Contact PCBs Temperature Monitoring and Control in a Hot Air Reflow Process Based on Multiple Thermocouples Data Fusion
Tin Lun Lam
Abstract
As the environmental concern is being raised over residues of lead, the trend of transferring from the conventional lead-based soldering to lead-free soldering is overwhelming. Lead-free solders require the peak temperature to be about 30 degrees Celsius higher than lead-based solders, which induce a narrower margin between the highest melting temperatures of lead-free solders and the heat-resistant temperatures of electronic components. As a result, the accuracy of temperature control of reflow systems needs to meet a higher standard to maintain the solder quality. Whereas, the conventional control process of the onboard temperature is open-loop, which cannot achieve the required accuracy. A closed-loop method by using an array of thermal image cameras for temperature monitoring is too expensive. In order to provide a low-cost and accurate temperature control solution for reflow systems, a cost-effective non-contact temperature approximation and control system is proposed in this article. The proposed temperature approximation is achieved based on a machine learning method with multiple-input single-output strategies to get a relationship between the temperatures near the PCBs and the onboard temperature. The proposed system controls the temperature in a real-time fuzzy logic algorithm to achieve a more accurate control result. The experiment results reveal the feasibility of the proposed temperature approximation and control system.