“Breaking down” the key process variables.

The original CC-Tango test vehicle¹ was extremely useful in generating a database of qualification and reliability results for conformal coating application designs, and in understanding the interactions between their various process variables. However, the initial paper on the CC-Tango TV contained no upstream (masking/cleaning) or downstream (inspection/rework) process evaluations for their impact on the coating process. That is remedied in this article.

Given conformal coating equipment, processes and chemistries are in continuous flux, a process qualification tool was required to understand these changes in both coating materials and application equipment.

New conformal coatings, such as low-volatile organics (LOC) and UV-based materials have been marketed. Depending on the intended end-use, common reliability tests may include fungus resistance², thermal shock³, hydrolytic stability⁴, dielectric withstanding voltage (DWV)⁵, flammability⁶, moisture and insulation resistance⁷, adhesion⁸ and salt-fog chamber testing, as well as mechanical product-level testing.

The CC-Tango TV is intended to generate baseline reliability data on conformal coating and to assess the impact of TV design features on the reliability of various DfM impacts. Although space restrictions limited the number of design feature variations included and tested, the TV included two variations of chip components to examine the effect of component locations on bubble (void) creation for reliability – a common process error.

A standard process for evaluating any new conformal coating materials or process alterations is required.

To further improve results achieved from the initial CC-Tango test vehicle evaluations, the various upstream/downstream processes were required to be tested, evaluated and understood for more complete overall process awareness.

The upstream/downstream processes assessed were cleaning operations, masking materials/processes and inspection/rework operations (Figure 1).


With the upstream/downstream processes tested, a closer look at the actual conformal coating applications in both material and application types could be evaluated now in isolation, leading into a process that could be used for site enablement and improvements.

Upstream/Downstream Processes

Cleaning. The first upstream/downstream aspect evaluated was the cleanliness levels. The single most common deterrent to adequate conformal coating coverage and adhesion is surface contamination. In that, within the presence of ionic residues, oils, residual water and FOD (foreign object debris) on the board or component surfaces can result in corrosion, poor adhesion and future failure of the conformal coating. This can occur even when using low-residue fluxes/no-clean processes, especially if traces of flux residue are on the assembly because of manufacturing processing or improper cleaning. Cleanliness standards for evaluating these variations include IPC-SC-60, IPC-SA-61 and IPC-AC-62.^9,10,11 This stated, the best method to minimize these potential issues is a thorough cleaning and subsequent drying process.

With time, testing and research, four specific variable branches, consisting of 11 variables, have been found to affect the output of a cleaning process. These four branches can be labeled: Assembly Materials, Chemistries, Components and Cleaning Process. The variables within this section have many subsets to explore, including various assembly materials, chemistries, components and potential cleaning processes. However, this work’s main intent is to explore conformal coating operations; thus, a simple subset of the various global manufacturing processes employed were based against either a standard cleaning process or not being cleaned.

The interesting finding was that, regardless of cleaning process used, for all tests completed, the washed samples produced less conformal coating coverage and adhesion issues.

Masking. Masking is another process that if not performed correctly can leave residues on the assembly that can affect the conformal coating coverage or adhesion. Masking materials include the following forms: silicone boots, covers, peelable masking, and masking tapes.

It is important to validate compatibility between masking and coating materials. In that, within the various masking materials, some contain a substance that is incompatible with coating materials and may even inhibit coating curing.

With time, testing and research, three specific material branches, consisting of eight material types, have been found to affect the masking process. These three branches can be identified as protective devices, encapsulants and tapes (Figure 2).


The variables within this section have many subsets that can be explored, including the vendors and materials for each branch variable. Again, based on this work’s main intent, a simple subset of the masking tapes was evaluated (Figure 3).


Requirements used to evaluate the masking materials were:

• Ease of placement/removal.
• Filtration of conformal coating.
• Delaminating/adhesion loss of coating when removed.
• Paste residue after removal.
• Dewetting around the tape material.

A key point that should be noted from this evaluation was all masking materials were non-ESD compliant and required the use of an ESD ionizer during both removal and application. Overall, with ionizers, no samples exceeded ESD limits.

Inspection/rework. The third upstream/downstream aspect evaluated was the inspection/rework processes used in the modifications and evaluations for conformal coating.

Compared to cleaning and masking upstream processes, inspection/rework is considered an inline and downstream process to the conformal coating process. This said, a common element to both the inspection and rework operations is lighting, playing a fundamental role in both detection and evaluation as to where the conformal coating is, where it is not and the material quality.

With time, testing and research, various light sources were found with wavelengths between 254-365 nm, with an optimal wavelength of 365 nm that produced both a quality source and safe inspection for operations, in both regular and darkened lighting.

Test Vehicle Design

For a typical conformal coating reliability test, the test vehicle design requirements are governed by IPC-CC-830.12 The specification was not, however, written with end-user inspection requirements in mind. Regardless, many of the primary considerations do not change, and the specification remains a useful starting point in designing TVs and test plans for evaluating conformal coating materials in a manufacturing process.

The CC-Tango TV layout was based heavily on the concepts to produce a one-shot process to establish a consistent global application/inspection method for existing conformal coating customer products, and new accounts and sites.

The CC-Tango TV was sourced in a standard 0.062" thickness with six metal layers. A total of 137 components are included within the TV to assess the various topologies that a conformal coating process would encounter and process defects that could arise in a process.

With the listed components, the CC-Tango TV was divided into five sections covering the following test requirements:

• Connector wicking and daisy-chain components.
• Various pitch, lead formats and surface orientations.
• Discrete spacing and configuration for bubble (void).
• PCB tooling hole wicking for material penetration.
• Nozzle placement accuracy and definition.

Figure 4 shows these test strategies. CC-Tango design overview and descriptions are in the following sections of this article. For reference to the conformal coating development tools within each section, see Keeping.¹


Section 1: The key aspects of this section for cleanliness testing are the connectors and BGA components. In that, the connectors chosen actually are supported off the assembly by 2.28 mm within this model and provide an area for post-soldering residues to build after wave soldering.

However, for the BGA components, post-soldering residues are a prime concern for cleaning, with inspection under these types of devices difficult or impossible. These are a couple reasons why many of these devices are now being underfilled prior to conformal coating.

Section 2: The key aspects of this section for cleanliness testing are the same physical variations in surfaces and component types that were selected. However, not for the material application requirement, but for flux entrapment during either SMT reflow or wave soldering (Figure 5).


Cleanliness for these components is critical, prior to being coated. In that, these components can vary in cost and removal due to contamination, depending on the coating material, and could damage the assembly in the process.

Section 3: The key aspect of this section was split between cleanliness and masking material testing. For cleanliness, with the various discretes and configurations, a lower standoff was provided for flux entrapment to be evaluated, with a FR-4 tab exposed for masking material and coating testing (Figure 6).


Section 4: With regard to the upstream/downstream processes, only a small impact on this section was included for cleanliness testing: the entrapment for FOD within the various tooling/via holes or flux entrapment runoff during SMT reflow or wave soldering (Figure 7).


Section 5: Covers nozzle placement accuracy and edge definition for control awareness testing for material and rotation (Figure 8). The purpose of this section was to investigate the placement accuracy and edge definition control of the process being evaluated. Poor results in this section should be avoided, since inadequate control of a film pattern could lead to coating migration into an undesired location. This can be caused from several variables, such as start/stop, overlap, and even spray atomization. If a start/stop is not corrected, a process defect described as a dog-bone can occur. The defect of this issue is a controlled pattern has a process limitation on closeness to a keep-out; however, if a dog-bone occurs, material pooling can occur at the end and lead to coating migration into the keep-out location. Sample defects are shown in Appendix 6 and 7. With regard to the upstream/downstream processes, this section was designed for application testing with minimum impact on either cleanliness or masking testing.


Conformal Coating Process

With the knowledge of the upstream/downstream processes, we can look into the process requirements that must be completed to maximize the coating application results and provide the highest return for the manufacturer’s and end-customer’s requirements.

Material. The first and primary process requirement that must be understood is the coating material selection that is completed. In that, each class of material has both advantages and disadvantages. Several common types of coatings are generically described as acrylic resin (AR), urethane resin (UR), epoxy resin (ER) and silicone resin (SR). (For briefings on their advantages and disadvantages, see the online version of this article).


Table 1 summarizes the various material properties, based on vendor technical data information sheets. It should be noted these values are based on averages of the data collected and variations exist.


Coating equipment. The second process requirement that must be understood is the coating process/equipment. Two coating equipment sets were evaluated for functional and complete coverage processes. The main difference between the two coverage types is the requirement for side coverage on nonconductive components. In that, for both coverage types, all conductive, metallic or lead surfaces are required to provide full conformal coating coverage. However, functional coverage will permit nonconductive, hermetically sealed areas to have material dewetting or creep from these locations. This initial requirement will then dictate the application process type required of either select film coating or an atomized spray pattern, with the latter required for complete coverage. Sample images for both process types are included within Appendix Sections 1 and 2.


Requirements to evaluate the coating process/equipment were:

• Initial equipment costs.
• Yearly equipment expenses.
• Processing cycle time.
• Service and maintenance.
• System capability.

Initial equipment costs. For this ROIC evaluation, initial equipment costs were not based on the system costs only, but included all tooling and accessories required to perform the required coating processing requirements. An interesting finding was that Vendor B’s base machine costs were lower than Vendor A’s, yet after all tooling and accessories were added, Vendor A’s initial equipment costs were lower. This is a critical aspect that must be understood when initial equipment analyses are completed. If not included upon initial capital assessments, either more tooling would be required or inaccurate tooling may be in place at higher costs than planned.

Yearly equipment expenses. We can focus on a different angle for higher resolution. This factor would be the floor space cost as based on equipment size and factory floor space cost (which varies per geography). For this evaluation, floor costs were considered comparable, with the key difference that Vendor B’s system footprint was slightly larger, providing a higher yearly equipment cost versus Vendor A’s.

Processing cycle time. With two distinctly different application types – functional and complete coverage – we need to select a single process for comparison. This noted, the following labor cost values will be assessed on a functional select coating operation: if both systems would require full assembly masking for a spray process with comparable material and labor costs incurred. An interesting finding in this section relates to the functional select coating process. This factor is the pattern definition of the film pattern, as compared to the spray pattern. The difference in the film pattern definition resulted in a higher level of pattern control for the application, leading to various cost increases (Table 2).


Service and maintenance. Both vendors provide excellent service upon request, along with required regular maintenance. Furthermore, continuous service depends on the process being completed, with higher service required for complete coverage, as compared to functional coverage. Tooling upgrades for maintenance can be acquired within reasonable timeframes to request. The sole support issue noted for Vendor B could be related to communication between site and vendor.

System capability. Various capability requirements were identified and evaluated independently of each other to provide an unbiased rating on the process. Functional and complete coverage, along with dispensing requirements, were evaluated (Figure 9). Overall, the two systems provided distinctly different strengths and weaknesses. Vendor A had four specific items considered strengths, including systems controls in place, pattern control and transfer effectiveness. This would suggest this system was more capable for select coating. However, Vendor B had one aspect that was slightly stronger: The system nozzle design could apply high viscosity material flow with tilt and rotation techniques. That, however, negatively affected placement and control accuracy.


Procedure/Results

A clear description should be provided as to which materials and processes will be evaluated (Table 3). The green squares are the two sections that were evaluated in this report.


A three-variable, two-level matrix is run to establish a process window based on customer requirements for coverage and thickness.

Five CC-Tango assemblies are run with their process requirements compared from process-to-process / material-to-material, as shown in the results for an acrylic process using both film and spray processes.

Two evaluations were completed, one for select coat process and one for atomized spray process.

As noted in Figure 10 and Table 4, there were distinct line definitions within the accuracy, along with the overall glossy appearance, that are common properties for film coat applications.


Figure 11 and Table 5 show the inconsistency within the accuracy section, along with the overall matte appearance, common properties for atomized spray coat applications. A key point is that the inconsistent/fuzzy edges in the accuracy section are created by the atomized portion of the coating that is also an advantage in being able to provide more uniform coverage to odd-shaped surfaces, materials and edges.


With the comparison between the two processes film and spray (Figure 12), there are direct advantages and obstacles for each process:


1) Cycle time. The overall cycle time for the running of the CC-Tango TV was 7.36 times longer for the spray process, as compared to the film process. Provided volumes are low, a spray process could match a film process in capital requirements. However, if volumes were of larger formats for either an inline or batch process, additional capital could be required, based on the assembly sizes or volume requirements, and would lead to a per cost increase due to capital and additional space requirements. Furthermore, as noted, a spray process requires additional masking that also increases the overall cycle time for the process.

2) Missing/dewets. There was a higher quantity of missing/dewetted locations: approximately 11 times greater for the film compared to the spray process. Most defects could be modified within an inline touchup process prior to final assembly inspection and shipment. The cost of this increase could result in hours of additional rework, depending on coverage requirements. A sample of these defects is shown in Appendix 3.


3) Bubble/voids. No bubbles or voids were found within the spray process application; however, four were found within the film. With finer process optimization for the film process, this value of zero defects could also be obtained. Sample bubble defects are shown in Appendix 5.


4) Wicking defects. The overall defects for this section were 297 to 3 for the spray process versus the film process, respectively. The main concept obtained from this data collection was that manual application of masking is required for a spray process to prevent coating of keep-out locations, and can be optimized and not included for a film process saving both in material and labor costs, along with cycle time. The labor for this masking process increase as a result of component times could lead to ranges between 5-30 min. per assembly more, compared to a film process. Defects are shown in Appendix 4.


5) Drainage results. There was no significant difference between the film and spray process penetration standings. However, the film process was applied slightly thicker than the spray and would require a via/hole of diameter 0.041", as compared to just over 0.035", to permit such flow and potential defect creation.

6) Thickness. The thickness for both processes was within operating limits with no signs of deviation from this requirement.

Overall, the three main process variables as described are the coverage requirements (missing/dewets), thickness and adhesion of coating to required areas. With these three process variables met, a stable process can be defined. These six process defects and requirements are summarized in Table 6.


Conclusion

Many of the most complex electronics assemblies have conformal coating processes in place that may not be optimized. Conformal coating results from this TV permit the effect of coating variations on reliability to be assessed, and should be useful in developing site specifications similar to IPC-CC-830 for use in standardizing reliability testing for conformal coating within a company.

The CC-Tango TV procedure and process has demonstrated that, by using this procedure, a comparison between various cleaning agents, masking materials, coating materials and application equipment can be completed. Comparisons between film and spray processes were completed with pros and cons provided for each.

In summary, if a functional conformal coating process was required, a film process could save 7.36 times in application cycle time, and 5-30 min. in masking labor time and materials, compared to spray.

If a complete conformal coating process was required for a high-complexity topology assembly, a spray process could save 11 times the rework/inspection loops compared to film.

Future Work

A version of the CC-Tango TV will be created with Pb-free solder to work with new fluxes being implemented to accommodate density increases and transition to Pb-free solders, along with different component packages to deal with the higher required interface properties for the increased density components. These changes have produced various interactions with the conformal coating materials and processes used.

A version of the CC-Tango TV with the inclusion of taller transformer packages for testing. Components of this nature would provide an extra obstacle for the coverage requirement to overcome, such as the shadow effect of smaller components, and evaluate the specific process.

A version of the CC-Tango TV with the inclusion of 0403 and 0201 components for cleanliness testing. Components of this nature would provide an extra obstacle for the cleanliness requirement to overcome and evaluate the specific process.

Acknowledgments
The author gratefully acknowledges the contribution of Hector Barrera of Celestica, Mexico and Ti Loon Ang of Celestica, Malaysia, in the data collection for the report, along with the assistance of material and equipment suppliers in the qualification installation and testing of the CC-Tango test vehicles. Finally, the author wishes to thank Jeffrey Kennedy of Celestica, Arden Hills, for assistance in the overall review and support in the designing and data collection for the CC-Tango process.

References

1. Jason Keeping, “Process Development and Optimization Using a Newly Designed Conformal Coating Test Vehicle” SMTAI, October 2007.
   
   
2. IPC, IPC-TM-650, “Test Methods Manual,” method 2.6.1.1, Fungus Resistance – Conformal Coating, July 2000.
   
   
3. IPC, IPC-TM-650, “Test Methods Manual,” method 2.6.7.1, Thermal Shock – Conformal Coating, July 2000.
   
   
4. IPC, IPC-TM-650, “Test Methods Manual,” method 2.6.11.1, Hydrolytic Stability – Conformal Coating, July 2000.
   
   
5. IPC, IPC-TM-650, “Test Methods Manual,” method 2.5.7.1, Dielectric Withstanding Voltage – Polymeric Conformal Coating, July 2000.
   
   
6. UL, “Standard for Tests for Flammability of Plastic Materials for Parts in Devices and Appliances,” October 1996.
   
   
7. IPC, IPC-TM-650, “Test Methods Manual,” method 2.6.3.4, Moisture and Insulation Resistance – Conformal Coating, July 2003.
   
   
8. IPC, IPC-TM-650, “Test Methods Manual,” method 2.4.1.6, Adhesion, Polymer Coating, July 1995.
   
   
9. IPC, IPC-SC-60, “Post Solder Solvent Cleaning Handbook,” August 1999.
   
   
10. IPC, IPC-SA-61, “Post Solder Semi-Aqueous Cleaning Handbook,” June 2002.
    
    
11. IPC, IPC-AC-62, “Post Solder Aqueous Cleaning Handbook,” January 1996.
    
    
12. IPC, IPC-CC-830B, “Qualification and Performance of Electrical Insulating Compound for Printed Wiring Assemblies,” August 2002.

Ed.: This paper was first published at the SMTA Pan Pac Symposium in January 2008 and is used here with permission.

Jason Keeping is project manager, conformal coating/Potting Sector, at Celestica Inc. (celestica.com); jkeeping@celestica.com.

SIDE BAR:
Conformal Coating Characteristics

Acrylic resin (AR): Acrylics are fairly easy to apply and dry rapidly, reaching optimum physical properties in minutes, while providing a long pot life. Furthermore, acrylics have good humidity resistance, are low exothermic and give off little or no heat during cure, eliminating damage to heat-sensitive components, and do not shrink due to the solvent evaporation process during cure. Their main disadvantage is solvent sensitivity, but this is also an advantage, as it makes them the easiest coatings to rework.

Urethane resin (UR): Polyurethane coatings are available in both single- and two-part formulations. Both formulations provide good humidity and high chemical resistance. Due to these variations, some polyurethane coatings are easy to apply and cure rapidly, with other materials more difficult to apply because of their shorter pot life. With their wide range in good-to-excellent material properties, these characteristics also become their drawbacks in that chemical and mechanical removal techniques may become difficult and costly.

Epoxy resin (ER): Epoxies are more complicated to apply than other materials because of their shorter pot life since they are usually available only as two-part compounds. Furthermore, as a result of their cross-linking design, they provide average humidity resistance and high chemical resistance compared to other coating materials. However, their strength versus other coating materials is their abrasion resistance, but this also adds to their rework complexity, as chemical removal may attach to epoxy-coated components and the board itself and cannot easily be removed via mechanical methods due to abrasion resistance.

Silicone resin (SR): Silicones are a different form of coating, compared to other materials, with their main advantage being resistance to higher continuous temperatures and thermal expansion properties. Furthermore, silicones have high humidity resistance, and thanks to their 100%-solids design, have extended pot life, providing a fairly easy application and quick drying. Their main disadvantage is abrasion resistance and rework complexity. With the abrasion resistance a disadvantage, this should lead to lower rework complexity. However, the rework disadvantage is not in the coating removal, but rather the removal of residues that may be left from the coating.