Printer Friendly

Analysis of stencil tensioning (elasticity): foil tension variations may create distortion and unpredictable print deposition. A novel method for foil characterization and verification is proposed.

Typical foil mounting methods involve stretching a woven fabric, typically polyester, across an aluminum frame, creating a taut screen (FIGURE 1). There appears to be no standard for mesh tensioning values, but different sources report a range of 25 to 35 N/cm (1). The thin metal foil is placed and bonded on the bottom of the frame, and the mesh covering the squeegee side is cut away. The mesh tension will suspend the bonded metal foil, holding it taut. IPC-7525B is informative regarding stencil design guidelines and various terminology. In this mounted state the tension of the actual foil is rarely verified due to lengthy verification and specialized tools.

[FIGURE 1 OMITTED]

Densely populated printed circuit boards and fine printing features make planarity and consistency from frame to frame important. Variation negates any precision established in the stencil printer design and manufacture.

Among the concerns for mesh mounted stencils:

* Aluminum frame: Planarity variability, bow under tension (different wall thickness).

* Mesh to frame: Variable tension, gluing variability.

* Foil to mesh: Foil flatness, gluing variability.

* Hand-assembled: Operator-dependent.

Alternative stencil mounting technology involves a reusable metal foil mounted to a reusable master frame, tensioning by either pneumatic bladder or a multitude of levers and springs. This technology is targeted at addressing the space and planarity drawbacks of mesh-mounted stencils. Tensioning of the foil in a master frame is created by either air bladders or springs attempting to emulate the mesh mount tensioning characteristics. This technology has been touted as having better tensioning characteristics than mesh-mounted stencils. However, as seen in FIGURE 2, a different tensioning trait can be seen.

[FIGURE 2 OMITTED]

Tension versus elasticity. Within this specific research, we found it difficult to comprehend the use of the term of "tension" in a system that now uses a steel plate or steel membrane for print deposition. Therefore, following definitions are highlighted for clarity and understanding of the correct terms to be used and how this paper will refer to them:

* Tension. Tension is the pulling force exerted by a string, cable, chain, or similar solid object on another object. In the case of today's stencils it is the lateral force pulling on the foil by either the tension of the mesh in a mesh mount stencil frame or the spring/pneumatic force derived by reusable tensioning frames.

* Elasticity. Elasticity is a physical property of materials that return to their original shape after they are deformed. How this applies to stencil systems of today is it is the characteristic of the foil property returning to its original position after a force has been applied and then removed. The tensioning system applied to the foil/steel plate maintains the plate in a state of planarity. Any force applied will deform the plate by a proportionate amount. This is explained by Hooke's law, a principle of physics that states the force needed to extend or compress a spring by some distance is proportional to that distance.

From these definitions, it is clear that the true property and characteristic of a stencil system should be measured by the spring constant k. This is a measure of elasticity of a spring that in our specific stencil systems is the stencil foil.

k = [DELTA]F/[DELTA]x

where

[DELTA]F = Force applied to the foil to displace

[DELTA]x = The distance moved by the force applied.

Area mapping of elasticity across the foil. To understand the characteristics of elasticity across the various stencil mounting technologies, a measurement methodology must be employed. A novel, patent-pending, fast real-time mapping system for foil characteristics utilizes complex resonance algorithms to determine regional characteristics of the foil. The tool used for our initial characterization is an RST Gage SMT Foil Analyzer (1). This system consists of a known weight, a floating dial indicator and a gantry mechanism. The gantry system rests on the outer frame, suspending the dial indicator in the selected region. The dial indicator is set to touch the foil and zeroed. Applying the known weight on to the dial indicator will deflect the foil; therefore, a deflection measurement will be displayed.

Each displacement reading is converted into N/cm by the formula k = [DELTA]F/[DELTA]x.

Nine points are taken across the foil area (2) (FIGURE 3), and for graph smoothing, additional data points are interpolated from the nine point reading. The types of stencils characterized are:

[FIGURE 3 OMITTED]

* Standard mesh mounted from different suppliers.

* Space saver mesh mounted frames.

* Mesh-less foil mounted system from manufacturer A.

* Mesh-less foil mounted system from manufacturer B.

Test Readings

The following surface plots represent the plotted regions of each stencil system. These graphs are a visual representation of how the foil is behaving in its relative mounting system. The narrower the lines, the greater the range/deviation.

Mesh mount--standard 29" x 29." The following outlines the format and characteristic of the mesh mounted stencils under test (FIGURE 4).

[FIGURE 4 OMITTED]

* Vendor A--Range across the foil = 10 N/cm, foil center = 40 N/cm.

* Vendor B--Range across the foil = 5 N/cm, foil center = 38 N/cm.

* Vendor C--Range across the foil = 9 N/cm, foil center = 39 N/cm.

As can be seen by the surface 4 plots, patterns are starting to appear. It is clear that the corners of the foils are undergoing higher stress than their center points, resulting in a dishing effect.

Mesh mount--space saver 29" x 29." FIGURE 5 shows surface plots of production stencils from the same vendor.

[FIGURE 5 OMITTED]

* Vendor D-a--Range across the foil = 15 N/cm, foil center = 4 1 N/cm.

* Vendor D-b--Range across the foil = 11 N/cm, foil center = 49 N/cm.

* Vendor D-c--Range across the foil = 22 N/cm, foil center = 23 N/cm.

Once again the dish-shaped patterns are occurring. Surface plot "D-c" was of a stencil that showed delamination around the foil. The data show clear indication of non-normal distribution and need for replacement.

Mesh-less tensioning systems. This test is directed at characterizing the two most popular tensioning systems. One test is to use one master frame with different foils, and one test with two master frames is from the same supplier with one foil. This was based on availability of master frames and foils at the time of the test.

Test 1: Same frame, different foils. With the mesh-less surface plots, pattern characteristics are appearing similar to the mesh-mounted stencils, but clearly the stress intensity is far greater on these tensioning devices (FIGURE 6). These patterns still reflect the foil corners undergoing higher stress than their center points, and the dishing effect is greatly exaggerated.

[FIGURE 6 OMITTED]

* Vendor E-a--Range across the foil = 40 N/cm, foil center = 47 N/cm.

* Vendor E-b--Range across the foil = 16 N/cm, foil center = 28 N/cm.

* Vendor E-c--Range across the foil = 24 N/cm, foil center = 51 N/cm.

Test 2: Same foil, two different master frames by same supplier. See FIGURE 7.

[FIGURE 7 OMITTED]

* Vendor F-a--Range across the foil = 21 N/cm, Foil Center = 47 N/cm

* Vendor F-b--Range across the foil = 24 N/cm, Foil Center = 44 N/cm

* Vendor F-c--Range across the foil = 2.5 N/cm, Foil Center = 45 N/cm

Summary of Elasticity Results

Although the patterns of both types of mounting technology have similar traits--higher in the corners, lower in the center--the spring/pressure tensioned systems display higher center elastic force and greater range across the whole foil area compared with the mesh-mounted stencils.

The typical stencil central point value is around 40 N/cm of a newer mesh-mounted stencil. The deviation from this value across the foil area is +10 N/cm (25% deviation).

For the mesh-less systems, the typical central point value is 46 N/cm, with a typical range deviation from this value of +20 N/cm (43.5% deviation).

As both systems (mesh mount and tensioning frames) have been in full production worldwide for years, there has been no specific evidence that one works better than the other with regard to print deposition performance. For discussion: if the perceived tension of a stencil is that the higher the better, then would that imply that the corners of each foil system print better than the center? Maybe edge-justified stencils print better than center-justified stencils? Either way, it would make better sense if these values were tightly controlled frame to frame, foil to foil where this consistency would remove variation and doubt from this argument.

The concentrated effort toward higher performance stencils for improving the transfer efficiency on smaller features is attempting to use the higher tension model to achieve better print performance. Given the effects of surface energies of stencil apertures, variations of solder paste chemistries and PCB surfaces/topography, it would also stand to reason that each individual image in the stencil will bring about infinite variables, due to positioning and sizes of the printing features within this image. On a side note, if the foil is to be held in a planar state during board separation for increased paste release performance, the mass of solder paste resting on the top side will clearly deform the stencil during the separation sequence.

In FIGURE 8, a represents the idealistic yet unrealistic stencil to board during the print separation process; b represents a slack foil and how it will sag based on little to no applied tension, and c represents a taut foil with a paste mass on the squeegee side. As the board separates, the unsupported paste mass will deflect the foil, causing a sag and therefore potentially uncontrolled separation. As the paste mass/bead deteriorates in size, so does its mass; therefore, the effect is lesser on the separation process.

[FIGURE 8 OMITTED]

As an exercise using the Catenary curve equation and a paste mass of 500g, it would take a tension force of over 300,000 N/cm to overcome the sag caused by the effect of the paste mass.

Area mapping of resonance across the foil. The Equi-Tone handheld device has been developed as a quick and objective measurement method/ device. First phase development results are very encouraging.

Harmonic oscillation. Given Hooke's law and our perspective that a stencil be viewed as a spring, then pulling the foil and releasing would set it into oscillation about its equilibrium position. However, due to the various mounting system techniques, the foil as a spring will be affected at a location across its area, bringing about different oscillations within these locations.

To test this theory, all the above tested stencils were compared with measuring the frequency responses in each of the nine locations previously selected and tested.

Frequency response across the foil. Moving a frequency recording device around each of the previously selected nine points highlighted not only patterns, but uncovered unwanted overtones. Split frequencies, beat frequencies and other mixed frequencies are noise to our signal. To isolate the primary frequency at each location, the use of Fast Fourier Transform (FFT) Spectrum Analyzer (FIGURE 9) and sophisticated filtering was required. These techniques would eliminate any false readings and overtones, and a stable primary frequency response could be isolated.

[FIGURE 9 OMITTED]

It is clear that the frequencies generated were complex and required various filtering to identify the primary frequency. Once this was achieved, the novel device was able to accurately and repeatedly identify the specific tone.

The testing and comparison study could then take place, with the results plotted below.

Novel device v. displacement. Using the novel device, each of the previously measured stencils was tested for comparison. The frequency data collected were paired and plotted with their respective stencils. As seen below, the patterns generated resemble the N/cm plots fairly closely.

Summary

It is clear from these findings that the steel used in stencils will deflect based on its own elastic properties. The mesh-mounted stencil and newer direct-tensioned frames are shown to have similar patterns. The mesh-mounted stencils have two modes of elasticity: one being the mesh, the other the foil. The lower elastic properties of the mesh permit the foil to be more copianar under stress than are direct-tensioned foils. Direct-tensioned systems clearly distort the foil and create larger deviations across the foil area. This phenomenon is counterintuitive when it comes to the idea of uniform separation between foil and board post-print. Further increasing the tension applied to the foil in these direct-tensioned systems may create further distortion and is therefore likely to create unpredictable print depositions. This research also points out that there are clear differences not only among vendor-to-vendor master frame technologies, but also different master frames from the same supplier.

Compared to other analyzers, the novel device clearly shows it can plot similar characteristics when area mapping the foil. Therefore, it is a viable alternative tool for fast and convenient foil characterization and verification.

Subsequent research will look at how these technologies and their elastic properties affect print deposition.

REFERENCES

(1.) RST Gauge, company website.

(2.) Murakami Screen, company website.

by RICKY BENNETT, PH.D., AND RICHARD LIESKE

RICKY BENNETT, PH.D., is founder of Lu-ConTechnologies; rbennett@lu-con.com. Richard Lieske is director of applied product development at Lu-Con Technologies.
COPYRIGHT 2013 UP Media Group, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

 
Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:STENCILS
Author:Bennett, Ricky; Lieske, Richard
Publication:Printed Circuit Design & Fab
Date:Jul 1, 2013
Words:2190
Previous Article:Connecting flex layers using conductive paste: a novel process eliminates copper plating, potentially increasing board flexibility and reliability.
Next Article:The role of failure-oriented-accelerated-testing for field functional IC packages: devices are known to pass qualification testing, then fail in the...

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters