Crossing the bridge from dispensing to mass imaging

Two technologies that have pursued separate development paths to date are fluid dispensing and mass imaging (stencil printing). While each technology offers advantages, one of them has achieved a critical stage. Through continuing development, mass imaging is now poised to broaden its scope, encompassing an increasing number of traditional dispensing applications and, in fact, surpassing them in newer process technologies. A variety of factors - those that drive productivity, lead to the exploration of new processes and help build bridges from one industry to another - have all contributed to this evolution.

Productivity drivers

In any given industry, certain issues and drivers make equipment and vendor selection a key factor for success. The issues relate to a supplier's performance over time. It is easy to look at past trends and understand present-day capabilities - but what about the future? Will the needs be greater than the current capabilities, leading users to make do or even lower their expectations? Do the companies and the technologies they develop instil confidence in their ability to deliver next-generation needs? As for drivers, in the most general terms they are throughput, capacity (output per real estate), first-pass yield, final yield and the cost to achieve the desired results.

The requirements of different industries and the needs of their particular business models and timeframes may cause differences in productivity drivers. For instance, today's economic climate is causing some manufacturing businesses to focus away from pure throughput (as in units per hour) in favour of emphasising cost to produce as the prime consideration. These types of performance indicators can and should take into account all aspects of equipment cost, from initial purchase price to long-term cost of ownership.

The earth is not flat

When the early explorers set about trying to convince their peers that they would not fall off the edge of the earth, it took many generations to sway conventional wisdom and the belief that it was suicide to venture towards the horizon. However, as the visionaries kept testing the waters and inching farther out, it soon became apparent to all that the earth was indeed round.

In a sense, the same can be said of mass imaging. Wise investments in the exploration of new technology and innovation have pushed its capabilities into areas that many people, even today, cannot imagine. Statements such as, "printing is not repeatable at wafer-level resolutions" or "it's not viable to print onto singulated substrates" are but a few of the echoes reverberating around inside the box of conventional thought.

Before taking a radical step into new territory, manufacturers require a significant reason to change to something that has the appearance of the unknown: the opportunity to triple output, for example. They also need a measure of security that certain things - such as materials that have gone through a lengthy qualification process - will not change, to ensure a seamless transition to a new process technology. But those who do cross the bridge can explore new and exciting technologies for improved manufacturing and assembly.

There is no doubt that, over the past 20 years, fluid-dispense technology has evolved into a highly successful and precise technique for depositing various fluid materials. It has made great strides to meet ever-increasing dispensing demands.

Through in-depth understanding of the various fluid dynamics needed to meet manufacturing needs, technological development has given rise to major breakthroughs in material control. Having X, Y and Z control over a traversing needle has allowed dispensing to deliver big rewards for many industries, for example in terms of yield, throughput and yearly cost savings.

In today's semiconductor packaging sector, dispensing companies are striving to achieve more flexible platforms and more precise material control in order to meet rapidly evolving requirements. Thinner ICs and packages require material to flow into more constricted spaces, for example. And the unique commercial realities of global semiconductor markets force vendors to seek to maximise throughput as well as reducing the cost to produce.

The flexibility to overcome emerging process obstacles - while delivering significantly increased throughput - is the benefit that mass imaging is poised to contribute to any material deposition process in any segment of the electronics industry and other industrial applications. Basically, today, virtually any material that can be dispensed is capable of being imaged onto any substrate, even those that are partially populated, in any type of carrier. A confluence of technological advances - in tooling, deposition, cleaning and stencil design - has enabled mass imaging to achieve this dramatic breakthrough into new territory. All of these advances have enabled one manufacturer - AMD - to make a seamless transition, accompanied by a tripling of throughput, from dispensing to mass imaging in its semiconductor assembly operation.

Historically, three issues had led the company to dispense every material used in its process: flux, underfill, thermal interface, lid seal and even solder paste, which is most commonly applied by stencil printing. The issues were the need for precise alignment of singulated substrates within a carrier, the need to maintain extreme cleanliness and the challenge of depositing materials onto partially populated substrates.

Being able to use the same material, without having to go through reformulation and requalification, was also an obstacle to change. However, faced with a pressing need to increase throughput and hesitating to make a major capital investment in additional dispensers, the company took the opportunity to explore a new technology.

Attacking the bottleneck

Because the primary process bottleneck was determined to be solder paste deposition for passive attach, that step was the first to be addressed. To resolve the alignment accuracy issue, a new tooling solution that allowed for precise, automated alignment of substrates within carriers was introduced and proved to function successfully. Working closely with AMD, the carriers were enhanced so as to take full advantage of DEK's Edge Reference Virtual Panel Tooling (ERVPT) system.

ERVPT consists of a unique precision pin design and a matched vacuum tower that centres each substrate within the JEDEC-compliant carrier and then lifts them to the printing position. This achieves the effect of having ten individualised substrates aligned with precision to each other, creating the effect of a panellised pattern (Figures 1 and 2).

Fig.1: ERVPT combines a precision pin design with matched vacuum towers that centre each substrate and lift them to the printing position. This achieves the effect of aligning individualised substrates to each other, creating the effect of a panellised pattern

Fig.2: Locked in place by vacuum in an ERVPT system, substrates are aligned with a high degree of accuracy as a virtual panel that may be imaged in a single cycle to achieve high throughput

The cleanliness issue was addressed with two new technologies. First, an enclosed print head, one that kept the material fully contained in a cassette, eliminated the need for squeegees, stencil wiping, handling of the material and its exposure to the environment (Figure 3). Second, a new paperless cleaning system ensured stencil cleanliness without the need to introduce paper or cloth, and the lint these materials create, into the facility's ultra-clean environment (Figure 4).

Fig.3: An enclosed print head enhances material management. The fully contained cassette reduces handling of the material and minimises its exposure to the environment, while virtually eliminating stencil surface cleaning

Fig.4: A paperless under-stencil cleaning system ensures cleanliness without the need to introduce paper or cloth, and the lint these materials create, into an ultra-clean production environment

Working in combination, these materials deposition technologies resolved the process issues, allowed for a seamless transition to mass imaging and, most important, enabled AMD to achieve a significant improvement in high-volume passive attach throughput, tripling its units per hour, while enhancing quality and consistency. Having experienced one successful transition to mass imaging, the company identified additional process bottlenecks - thermal interface, lid seal, underfill - and trials began to determine feasibility.

Additional benefits of such a transition include flexibility and lower costs, both initially and over time. Mass imaging uses a stencil printing platform, which costs significantly less than a dedicated dispensing system and occupies less real estate on the factory floor.

By changing the stencil and, in some cases, adapting the enclosed transfer head, the basic platform may be used to image an ever-lengthening list of materials, including solder paste, adhesives, underfills, encapsulants, lid seals, thermal interfaces and solder spheres. These low-cost adaptations enable any user to respond rapidly and cost-effectively to changing production requirements.

Tip of the iceberg

So, how does a manufacturer know what can or cannot be printed? An iceberg analogy - the proportion of what is visible to what is not - holds the answer. Although some applications require a lateral or "out of the box" thought process, they still have clearly defined boundaries. Indeed, most people address only the visible part of a challenge. Rather, what is needed is a "what if" rather than an "it can't" mind set and not getting caught in the conventional trap of looking only at what is visible.

Taking the iceberg concept one step further, if nine-tenths of it is not visible, imagine what possibilities lie beneath the surface. Without exploration, those possibilities are very likely to remain invisible. In the words of Edd Alcid, engineering manager of flip chip assembly at AMD, it was "interdependent partnering, with both parties working collaboratively on our overlapping missions, that enabled us to overcome the obstacles, explore an unknown region and achieve a successful outcome."

The clear path to success, then, requires teaming up with the right partner, one that has a good track record of producing technology and innovation that crosses barriers and, more importantly, possesses the knowledge to bring its ideas and concepts to reality in a timely manner. This approach enables partners to overcome barriers, reveal the invisible, cross inter-industry bridges - and improve both performance and productivity.