Reading time ( words)
Sequential lamination, as it is used today in high density interconnect (HDI) and derivative technologies, is constrained by the fact that one cannot plate a blind hole deeper than the diameter of the hole. A larger hole allows processes to plate deeper. In fact, this manufacturing constraint has made it a challenge even to reliably plate and process blind holes up to a 1:1 aspect ratio (AR).
NextGIn Technology, a technology company based in Helmond, the Netherlands, took up the challenge to redesign PCB lamination techniques to be easier to fabricate, to increase performance, and to lower fabrication cost in comparison to current technologies. The constraint set by NextGIn Technology was to use only current fabrication processes and tools available in the board shops. Using no new equipment, NextGIn set out to develop new processes for existing facilities. To do this, NextGIn needed to rethink the possibilities for what can be done with the capital equipment and processes. NextGIn has named this new process “vertical conductive structure” or VeCS.
Traditional manufacturing constraints stipulate that to plate deeper, a larger diameter hole is required. And yet, there is often no additional space in the board design for bigger holes. Perhaps the shape of the hole can be rethought. An oblong hole or slot, for example, would allow the hole to be cut up in multiple structures. The limit to plating is the size of the holes. One can plate a blind hole as long as you respect the AR of 1:1. Even the 1:1 AR can be a challenge to plate reliably, but to fit it into the current design footprint is not an option.
Figure 1: Slot dimensions used to conduct initial “throw” testing.
Cutting holes into multiple sections has been on the research and development agenda of the interconnect industry for some decades but work to turn the technique into a process has not been successful to date. NextGIn started by modifying the shape of the hole. An oblong shape structure was created that was broken into multiple sections to form the contacts. In the initial plating experiments, this oblong shape showed good to average “throw” in the blind slots (Figures 1 and 2).
Figure 2: Test results, demonstrating that the more a slot resembles a circle, the shorter the plating “throw” capability.
Our results showed that the longer the slot, the easier it is to plate. Shorter and deeper slots tend to exhibit a threshold beyond which they experience a lack of plating. The threshold is defined by the depth and length of the slot. Presently, we target an AR threshold of maximum 4:1 in our designs with a ratio of slot length to slot width at a minimum of 3:1.
An interesting development from this initial work is that the AR definition for a blind slot now requires an additional dimensional variable. Along with slot width, depth, and the new parameter, length. Ultimately, the objective is to create deep slots up to 2 mm for regular circuit boards and even 3 mm for more advanced products. In addition to the slot depth, NextGIn’s experiments concentrated on slot diameters in the region of 0.2–0.5 mm. NextGIn selected this range because larger slot widths are not as useful with respect to BGA component footprints, and smaller width slots—width of interest—are difficult to form reliably for production due to the stability and useful life of the mechanical drill bits.
To read the full article, which appeared in the April 2019 issue of PCB007 Magazine, click here.
