The optimal solution doesn't mean ramping the cleaning frequency.

Understencil cleaning is essential if a process is to be repeatable. Excessive cleaning, on the other hand, will slow throughput unnecessarily. SPC tools can help optimize the frequency, but this frequency is increasing as aperture dimensions diminish and Pb-free pastes display poorer release characteristics. Hence, faster understencil cleaning techniques are required to maintain high throughput at cleaning rates optimal for the latest generation stencil-and-paste combinations.

Dominant factors in the future of electronics assembly include smaller component dimensions and interconnect pitches, and Pb-free solder use in the majority of products destined for developed markets. Numerous studies into the behavior of Pb-free screen-printing processes, using experimental stencils designed for ICs in CSPs, make clear that  both trends tend to reduce paste release efficiency when screen printing with stainless steel laser-cut stencils. Figure 1 is generated from results gained during a study into Pb-free screen printing1, and demonstrates how paste release is reduced for Pb-free pastes compared to an established SnPb paste formulation. Figure 2² demonstrates how paste release reduces with the ratio of aperture width to stencil thickness.


The stencil apertures used during these experiments met established design rules for 0.5 mm and 0.4 mm pitch QFPs, and 0.5 mm pitch CSPs. Aperture geometries were determined by applying the established area ratio formula:


Among the many undesirable effects of poor paste release efficiency, solder paste that remains in the stencil aperture tends to partially block that aperture. Successive print cycles add to the blockage, preventing a progressively greater proportion of the paste from transferring to the board. If this remnant paste is not removed at suitable intervals, the process will move out of control and end-of-line yield will suffer.

Paste transfer falls during the interval immediately after an understencil clean is performed. In an optimized process operating within desired limits, a cleaning operation will be invoked just before the process moves outside those limits. Some solder paste inspection systems permit viewing of process performance in relation to the understencil cleaning interval, and adjust this interval accordingly to minimize cleaning time overheads without permitting the process to enter an out-of-limits condition.

Remembering that the research^1,2 into Pb-free screen printing shows that paste transfer efficiency falls with Pb-free pastes and the small stencil apertures required when assembling CSPs and other components of comparable size, the rate of understencil cleaning must increase to maintain the process within limits. But more frequent understencil cleaning introduces extra time overheads that will reduce overall throughput. To address this, cleaning systems must become faster, achieving a reduced cycle time to minimize the impact that this increased cleaning frequency will have on overall throughput.

Understencil Cleaning with Paper

To examine how the time overhead associated with understencil cleaning may be reduced, analyze the cleaning routine execution, as well as the actions required to replenish exhausted cleaning paper.

A conventional USC mechanism may permit a cleaning routine to be configured according to the stencil’s characteristics or meet specific product requirements. This information is usually stored in the product file. It is possible to specify a cleaning routine comprising multiple sweeps performing dry, wet or vacuum cleaning in any combination or sequence. For most products, three cleaning sweeps are sufficient; for example, combining one wet, one vacuum and one dry sweep. The machine retrieves the specification for the cleaning routine from the product file when the cleaning routine is activated. The understencil cleaning rate, which defines the number of print cycles to be executed before the cleaning routine is next performed, is also stored in the product file.

When the cleaning routine is activated, the USC mechanism is moved into the active area of the machine beneath the stencil. Cleaning solvent is dispensed onto the paper, if selected. The mechanism is raised to contact the stencil, and the cleaner paper is wiped across the underside. In the case of a vacuum cleaning routine specification from the product file.

Throughput Impact

Cleaning routine duration depends heavily on the number of sweeps required and the time to complete each sweep. Simply increasing the speed of the mechanism to wipe the stencil more quickly risks reducing cleaning effectiveness and will not deliver a real improvement. Techniques now being introduced with new USC systems include the use of an oscillating mechanism that causes the paper to scrub the underside of the stencil, permitting more paste to be removed in a single sweep.

Other measures to combine certain phases of the wet, dry and vacuum cleaning sequence can reduce the overall cleaning cycle time without impairing the effectiveness of the cleaning action. For example, a multi-action cleaning head (Figure 3) is able to achieve effective cleaning with fewer sweeps. In a typical situation, only two sweeps are required to perform a complete wet, vacuum and dry cleaning routine.


Replenishment Time

Time spent replenishing the cleaning system also has an important bearing on overall productivity. Paper replenishment may take place before commencing a print run, at setup or changeover, or during the course of a print run. In any case, this is non-productive time that increases the cost of ownership of the machine. The replenishment process can take between five and 10 minutes, depending on such factors as operator skill or experience.

Setting the stencil cleaning rate too high in the product file will hasten roll replenishment, thereby increasing machine stoppages. SPC tools can help determine the optimum rate for a given product and associated process parameters.

It is also worth considering that inefficient cleaning paper utilization – for example, advancing the paper further than is strictly necessary during any given cleaning routine – will also lead to excessive roll replenishment. Some cleaners, for example, index the paper forward by the same distance, regardless of whether a wet, dry or vacuum clean has occurred. However, a shorter distance may be acceptable after a dry sweep, for example. Enhanced cleaning systems that precisely control paper advance after each type of sweep can deliver cumulative savings in paper consumption, resulting in a significant reduction of replenishment overheads.³

Finally, measures to simplify paper replacement when the machine does have to be stopped will reduce tthe downtime and thereby significantly improve productivity, saving a valuable portion of the five to 10 minutes’ typical replenishment time.³

Cassette-Type Understencil Cleaners

To replenish the paper roll in a conventional understencil cleaner, the empty paper roller must first be removed, followed by the take-up roller containing the used paper. The used paper is then removed for disposal, and the take-up roller is repositioned in the USC body. The new paper roller is then prepared for insertion. When the new paper roll is in place, the operator must then route the paper correctly across the cleaner body, ensuring that the edge of the paper is parallel to the axis of the roller. The paper is secured to the take-up roller, and must then be tensioned correctly. Finally, the operator must verify correct and accurate routing of the paper and check that the tension is set correctly, before priming the paper feed mechanism. The machine is then ready to print.

Redesigning the traditional dual-spool paper mechanism, for example to adopt a cassette-type mechanism that can be manipulated as a self-contained unit (Figure 4), reduces the number of steps required to remove the paper feed and take-up rollers. The need to manually route the new paper roll from the feed roller to the take-up roller can also be eliminated.


Some of the latest USC systems have a lift-off cassette that can be replenished in a matter of seconds, which translates into a significant saving in machine stoppage time.³ The benefits of easier handling and straightforward setup also include reduced operator training requirements. The correct paper tension is also preset, which saves time and eliminates human error.

Forthcoming screen-printing processes will be dominated by Pb-free paste adoption and progressive reductions in stencil aperture dimensions. These factors will demand more frequent understencil cleaning. As the effect on throughput becomes apparent, assemblers will require faster understencil cleaning techniques to maintain productivity.

References

1. Clive Ashmore, “Further Investigation Into Mass Imaging Lead Free Materials Using Enclosed Print Head Technology,” Global SMT & Packaging, October 2004.
2. Clive Ashmore and Rick Goldsmith, “Investigating Mass Imaging Lead Free Materials Using Enclosed Print Head Technology,” Nepcon West, December 2002.
3. Trevor Warren, Simon Clasper and Clive Ashmore, Investigation into the Parameters of Understencil Cleaning Requirements, DEK internal study, March 2006.

Trevor Warren is Americas product manager at DEK (dek.com); twarren@dek.com.