EVALUATION OF QFN TECHNOLOGY WITH OPTICALLY INSPECT ABLE SOLDER CONNECTIONS

摘要

QFN or Quad Flat No-Lead packages have steadily been gaining popularity since being first introduced and then registered as a JEDEC package almost twenty years ago [1]. This leadframe based, leadless package can offer many advantages over leaded (i.e., QFP, SOIC, TSSOP, etc..) and ball grid array (BGA) packages including lower overall mounted height, reduced footprint, excellent electrical and thermal performance and all at a cost typically lower than substrate-based and leaded packages of similar I/O [2]. Currently, QFN is typically limited to up to 12x12 mm body size and up to the low 100's in I/O depending on lead pitch for single row QFNs [3-4]. QFN originally exclusively incorporated gold and then copper wire-bond interconnects from die to leadframe, but has evolved to include flip chip on lead. Other developments in QFN configurations that have occurred in recent years include cavity QFN with a pre-molded cavity instead of conventional transfer molded, stacked die, multi row or array QFN, half etching of leadframes to increase interlocking and therefore package integrity, thermal enhancements, and pitch reductions from the original 1.0 mm pitch down to 0.4 and even 0.3 mm pitch. Most QFNs incorporate either NiPd/NiPdAu pre-plated leadframes (PPF) or matte Sn electroplating following the molding process. Another area of considerable development has been with the QFN package assembly process itself. QFNs have always been molded on a multi-up (matrix) leadframes. For QFNs where each individual unit has its own mold cavity, a punching process is typically used to singulate units from the leadframe. For QFNs where multiple units are molded in a single cavity, a saw process is used to singulate individual units. Both of these processes can result in leadframe bare copper on the exterior perimeter of the part being exposed. One of the perceived disadvantages of QFN had been the fact that on packages with this exposed copper, which generally gets oxidized prior to SMT reflow, it does not consistently wet with solder in typical SMT processes. Lack of a QFN side fillet results in only the joints under the package being inspectable by X-ray in a similar manner to BGA. However, BGA does not rely as much as QFN on a very consistent solder paste printing process since most of the volume of a BGA solder joint is supplied by the volume of BGA sphere. This paper will go over leadframe and package manufacturing processes that can be used to potentially make the portion of the leadframe that is exposed on the side of the package to be solderable. This can lead to consistent solder joint fillet formation on the side of the QFN package that is easily inspectable including with Automated Optical Inspection (AOI). The two most common ways to achieve at least a partially wettable lead on the sides of the QFN are the use of "dimpled" leadframes or by step-cutting the package so that part of the edge of the lead gets electroplated with Sn. QFNs that promote the formation of side fillets are referred to as QFNs with "inspectable joints" (IJ) or "wettable flanks" (WF). This paper will present a case study on step-cut QFNs that are able to form these optically inspectable joints. PCB footprint design, stencil design, SMT processes, inspection data along with reliability data will be discussed.