Part I: manufacturing with Increasingly Smaller Components
December 31, 1969 |Estimated reading time: 5 minutes
By Dave Kalen
Despite the benefits inherent in 0201 components, their use can complicate the assembly process. Understanding smaller component spacing requirements can create flexible, cost-effective products.
Always aggressive, electronics manufacturers continue accelerating the drive toward the twin goals of product miniaturization and functional diversity. To get more function and performance into existing products and create even smaller, lighter products, manufacturers are looking toward increasingly smaller components. The latest passive components reduce board space by 66 percent. In every form factor consumers want more for instance the trend is to equip cellular phones with Internet access. Unfortunately, smaller component package sizes complicate the board and module assembly process, e.g., by introducing the need for three-axis correction in both pickup and placement. These issues will become more critical as manufacturers turn to 01005 sized chips, which are scheduled for use in advanced placement by 2005.
The two key areas for most assembly technical issues lie in solder paste printing and physically placing components on the motherboard. This first article in a two-part series addresses the printing issues: How they affect component density and how manufacturer requirements can be met. The next article discusses the placement issues involved with 0201 components.
Minimizing the Component GapOne key consideration in using 0201 components in new product designs is that of component spacing. The current standard of about 0.2 mm between components is inadequate for 0201 components. After all, manufacturers using 0201 technology face significantly higher component and rework costs than with 0402 technology. Therefore, it is clear that using conventional component spacing will not provide the component density to allow cost-effective product manufacture. More importantly, the products requiring the smallest components are those that are portable, sometimes wearable, and often designed in sizes and shapes that suit the whims of fashion rather than standard product package design. Component space is at a premium; therefore, extra space between components is unacceptable. (This explains why 0201 placement often is referred to as "extreme adjacent mounting.")
So how extreme do we need to be? From a practical perspective, how much space needs to be reduced? More importantly, what are the technical issues that must be addressed? Research indicates that the optimum spacing for 0201 components is 0.1 mm, or half the space for 0402 components.
To meet this gap requirement, it is necessary to change the way we think about component pads in two areas. The first is with typical solder fillet. For mounting components larger than 0201, component pads typically are printed slightly larger than the components themselves, resulting in a solder fillet that bridges between the end of the component and the exposed portion of the pad. This, however, wastes board space. The extra pad, the portion not covered by the component is unnecessary. To eliminate the fillet, a pad smaller than the actual component needs to be printed (Figure 1). Reducing pad size prevents fillets from forming.
Figure 1. Solder fillets protrude from the component sides, consuming valuable space and making solder bridges more easily possible. Filletless mounting, while conserving board space and decreasing assembly defects, presents an inspection problem.
Besides eliminating the fillet, a smaller pad offers another important benefit: For any given component gap, it increases the physical space between adjacent pad areas, which can reduce the incidence of bridging.
There are practical limits to how far this pad-size reduction strategy can be taken. If the pad is too small, printability can deteriorate, causing problems rather than solving them. There are finite limits to all improvements. In this case, the screen aperture dimensions need to be the same as the pad dimensions. Screen thickness also limits aperture size. If the stencil is too thin, it will reduce the solder paste volume available for other surface mount devices on the board. For 0201 components, however, the smaller pad size works effectively.
Figure 2. Constructing the pad and resist at the same height (normal resists) is the optimum approach. The solder mask defined pad will determine the pattern dimensions.
The second printing concern is solder mask thickness relative to that of the component pad (Figure 2). There are two choices with the resist: In printing normal solder mask, the solder mask opening is slightly larger than the pad dimensions. With solder mask defined pads, the resist opening is smaller than the pad dimensions.
Figure 3. Making resist thicker than the pad reduces solderability with normal resist (a). In this case, it is better use over resist to maintain the contact between the screen and the resist (b).
In cases in which the solder mask is taller than the pad, the extruded solder paste (through the stencil aperture) may not come in contact with the pad sufficiently and paste transfer to the pad will be inadequate (Figures 3a and 3b), reducing solder volume significantly. In this case, over resist works best because the resist and screen are in close contact, allowing the bottom of the solder paste to touch the top of the pad.
When the pad is as thick as the solder mask, the solder paste prints fine. If, on the other hand, there are solder mask defined pads and a thicker deposit, there may be too much solder paste volume, causing bridging or solder mask chip out.
Solder Paste SelectionWith pad sizes becoming smaller, aperture size and stencil printing thickness will shrink as well.
One area that significantly affects the printed pad quality is the solder particle size. It seems obvious that smaller particles will print with higher resolution, but how small is small enough? On today's production floor, Type 3 powder, with solder particles in the 38 to 45 μm range, typically is used. Smaller particles called "fines" are shaped irregularly and are difficult to classify and separate. The need for Type 5 or 6 powder is critical to future solder paste needs.
To define a best-case scenario for filletless 0201 placement and reflow, one company* experimented with various particle sizes. Using various solder paste particle sizes, a single pattern was printed and tight control was maintained over as many variables as possible. The test parameters in the testing involved a component gap of 0.1 mm, stencil thickness of 0.1 mm, a squeegee speed of 15.0 mm/sec and squeegee pressure of 5.5 kg. Four tests were run, printing with normal solder mask and solder mask defined pads.
The four sizes used were less than 20 μm, 15 to 25 μm, 5 to 15 μm and 18 to 37 μm. Generally, as expected, smaller particles produced better printability. Interestingly, when particle size is over a more narrow range, normal solder mask prints consistently better than over solder mask defined pads. When particle size is less than 25 μm, satisfactory printability for filletless patterns can be obtained.
Based on these results, it is believed that filletless patterns will enable parts to be brickwalled closer together; thus, obtaining higher component densities will require smaller solder particle sizes and thinner stencils.
*Hitachi High Technologies America
Dave Kalen, central regional sales manager, may be contacted at Hitachi High Technologies America, 240 McArthur Court, Anderson, IN 46012; (765) 649-1066; Fax: (765) 649-1140; E-mail: dave.kalen@hitachi-hhta.com.