|
|
|
|
|
Selective Reflow Rework Process
Production Floor |
|
Authored By:Omar Garcia, Enrique Avelar, C. Sanchez, M. Carrillo, O. Mendoza, J. Medina, Zhen (Jane) Feng,Ph.D., Murad Kurwa Flextronics International Inc. Zapopan Jalisco, Mexico SummaryIn the rework environment, most equipment and procedures are designed for low volume repair/rework process. When a high volume rework is needed, the challenges begin. For example, a long cycle time is required to perform ball grid array (BGA) rework. When we need to remove material, do pad dressing, pad inspection, paste printing and place a new BGA, those steps increase the amount of dedicated rework equipment. Some machines are used to remove material, others are used to do pad dressing and others to place a new BGA. This results in hundreds of rework tools and equipment on the production floor. That volume of rework consumes enormous amounts of resources, requiring process controls such as daily profiling and maintenance using excessive hours of human resources. In addition, the standard rework process has low yield and high scrap rates. The Selective Reflow Rework Process is an approach to improving the high volume rework process, increasing process capabilities and process repeatability by using a standard reflow oven of 12 zones, pick and place machinery, semi-automated printing gear and Solder Paste Inspection (SPI) implementations. This approach was able to reduce the amount of rework equipment by more than half. Our human resource requirements (indirect and direct labor) were cut by more than 50% and our rolled throughput yield increased from 68.9% to 84.14%. The Selective Reflow Rework Process is less reliant upon operators and has become a repeatable, stable rework process. To obtain this advantage and have a successful implementation of this technology, the process requires new controls for printing, and check points before proceeding to the next process step. The printing process has a major impact on the HiP reduction, optimizing solder paste transfer efficiency (TE) and establishing a real SPC that gives real time warnings of anomalies. By identifying challenging process key parameters, including paste height, printing technique, pallets design and thermal barrier protection of TH parts, this paper will discuss some aspects of the process optimization and changes made to improve the quality of the rework process. ConclusionsWhile several factors affected the yield in our field failure rework units, the main contributor to yield loss was HiP (head in pillow). From this research, the following conclusions can be made: BGA placement and reflow become a steady process using the SMT machinery - little human interaction affected the process One of the main variables that became significant was the profile (affecting the BGA warpage). Profiles with peak temperatures close to 253 degrees C produced less HiP defects. Manual paste printing was the major variable that aided or instigated the HiP effect Automation of the process minimized variability, increasing the yields Paste height lower than 0.16 mm (6.3 mils) is more likely to produce HiP, while paste height above 0.191 mm (7.5 mils) increases the risk of solder bridges Component warpage under thermal stress combined with solder ball sitting plane height created a large gap, and printed pastes became too sensitive to promote or reduce the HiP The HiP phenomenon is not a metallurgical issue, but rather a mechanical/thermal problem Initially Published in the IPC Proceedings |
|
Comments
|
|
|
|
|
|
|
|
|