Research and recommendations on using lead-free PCB surface treatment processes during the circuit testing phase

Introduction: The emergence of lead-free PCBs has raised new issues for in-circuit test (ICT). This paper describes the existing PCB surface treatment processes and analyzes the impact of these processes on ICT. It points out that the key factor affecting ICT is the contact reliability between the probe and the test point, and introduces the specific changes that need to be made in the PCB construction process to meet the requirements of ICT.

Figure 1: The user adopted a recommended OSP rule set but still saw a 12% impact on first pass yield.

Historically, the main concern of the test engineer has been to ensure that he has an effective test program that will perform well in production. "In Circuit Test (ICT)" remains a very effective method for detecting manufacturing defects. More advanced ICT systems can also add real value in the configuration of test functions by providing a method to program Flash memories, PLDs, FPGAs and EEPROMs during test. Agilent's 3070 system is the market leader in ICT.

ICT still plays an important role in the manufacturing and testing process of printed circuit board assembly (PCA), but how will people's pursuit of lead-free PCBs affect the ICT stage?

The push for lead-free soldering has resulted in a great deal of research into PCB surface finishes. These studies are primarily based on the performance of the technology during the PCB build process. The impact of different PCB surface finishes on the testing phase has been largely ignored, or focused solely on contact resistance. This report will present details of the impacts observed in ICT and the need to respond to and understand these changes.

The purpose of this article is to share PCB surface treatment experience and train engineers on the changes required to achieve ICT PCB production process. This article will discuss the issues of lead-free PCB surface treatment, especially in the ICT stage of the manufacturing process, and reveal that the successful testing of lead-free surface treatment also depends on the beneficial contribution of the PCB construction process.

Successful ICT testing is always about the physical properties of the contact point between the test probe of the bed of nails fixture and the test pad on the PCB. When a very sharp probe contacts a soldered test point, the solder will dent because the contact pressure of the probe is much higher than the yield strength of the solder. As the solder dents, the probe penetrates any impurities on the surface of the test pad. The uncontaminated solder below now contacts the probe to achieve good contact with the test point. The depth of probe insertion is a direct function of the yield strength of the target material. The deeper the probe penetrates, the better the contact.

An 8-ounce (oz) probe can apply a contact pressure of 26,000 to 160,000 psi (pounds per square inch), depending on the surface diameter. Because the yield strength of solder is about 5,000 psi, the probe makes better contact with this relatively soft solder.

PCB surface treatment process selection

Before we get into the whys and wherefores, it is important to describe the types of PCB surface finishes that exist and what they offer. All printed circuit boards (PCBs) have a copper layer on the board that will oxidize and deteriorate if left unprotected. There are a number of different protective layers that can be used, the most common being hot air solder leveling (HASL), organic solder preservatives (OSP), electroless nickel gold immersion (ENIG), silver immersion, and tin immersion.

Hot Air Solder Leveling (HASL)

HASL is the main lead surface treatment process used in the industry. The process consists of immersing the circuit board in a lead-tin alloy, and the excess solder is removed by a "wind knife", which is hot air blown on the surface of the board. For PCA processes, HASL has many advantages: it is the cheapest PCB, and the surface layer can be soldered after multiple reflows, cleaning and storage. For ICT, HASL also provides a process for automatic solder coverage of test pads and vias. However, compared with existing alternatives, the flatness or coplanarity of the HASL surface is poor. Now there are some lead-free HASL alternative processes, which are becoming more and more popular due to the natural substitution characteristics of HASL. HASL has been used for many years with good results, but with the emergence of "environmentally friendly" green process requirements, the days of this process are numbered. In addition to the lead-free problem, the increasing complexity of boards and finer spacing have exposed many limitations of the HASL process.

Advantages: Lowest cost PCB surface process, maintains solderability throughout the manufacturing process, no negative impact on ICT.

Disadvantages: Typically uses leaded processes, which are now limited and will eventually be eliminated by 2007. Can cause solder bridging and thickness issues for fine pin pitches (<0.64mm). Surface unevenness can cause coplanarity issues during assembly.

Organic solder preservative

Organic solder preservatives (OSP) are used to create a thin, uniform protective layer on the copper surface of a PCB. This coating protects the circuit from oxidation during storage and assembly operations. This process has been around for a long time, but has only recently gained popularity with the quest for lead-free technology and fine pitch solutions.

OSP has better performance on PCA assembly than HASL in terms of coplanarity and solderability, but requires significant process changes in the type of flux and the number of thermal cycles. Careful handling is required because its acidic characteristics degrade OSP performance and make copper susceptible to oxidation. Assemblers prefer to handle metal surfaces that are more flexible and can withstand more thermal cycling cycles.

With OSP surface finish, if the test points are not soldered, this will cause contact problems with the bed-of-nails fixture at ICT. Simply changing to a sharper probe type to penetrate the OSP layer will only result in damage and puncture of the PCA test vias or test pads. Studies have shown that changing to higher probing forces or changing probe types will have little impact on yield. Untreated copper has a yield strength an order of magnitude higher than lead solder, and the only result will be damage to the exposed copper test pads. All testability guidelines strongly recommend not probing directly to exposed copper. When using OSP, a set of OSP rules need to be defined for the ICT stage. The most important rule requires opening the stencil at the beginning of the PCB process to allow solder paste to be applied to the test pads and vias that need to be contacted by ICT.

Advantages: Comparable to HASL in unit cost, good coplanarity, lead-free process, improved solderability.

Disadvantages: The assembly process needs to be significantly changed. Probing unprocessed copper surfaces is not conducive to ICT. Overly sharp ICT probes may damage the PCB. Manual precautions are required, which limits ICT testing and reduces test repeatability.

Electroless Nickel Gold Immersion

Electroless nickel immersion gold (ENIG) is a coating that has been successfully used on many circuit boards. Although it has a higher unit cost, it has a flat surface and excellent solderability. The main disadvantage is that the electroless nickel layer is very fragile and has been found to crack under mechanical stress. This is called "black block" or "mud crack" in the industry, which has led to some negative publicity for ENIG.

Advantages: good solderability, smooth surface, long storage life, can withstand multiple reflow soldering.

Disadvantages: High cost (about 5 times that of HASL), "black block" problem, manufacturing process uses cyanide and some other harmful chemicals.

Silver Immersion

Silver immersion is a new method for PCB surface treatment. It is mainly used in Asia and is gaining popularity in North America and Europe.

During the soldering process, the silver layer melts into the solder joint, leaving a tin/lead/silver alloy on the copper layer, which provides a very reliable solder joint for BGA packages. Its contrasting color makes it easy to inspect and it is also a natural alternative to HASL in soldering processes.

Silver immersion is a very promising surface treatment process, but like all new surface treatment technologies, end users are very conservative. Many manufacturers regard this process as a "under investigation" process, but it may become the best lead-free surface treatment process option.

Advantages: good solderability, smooth surface, natural alternative to HASL immersion.

Disadvantages: Conservative attitudes among end users mean that there is a lack of relevant information within the industry.

Tin immersion

This is a newer surface treatment process that has many similar characteristics to the silver immersion process. However, there are significant health and safety issues to consider due to the need to prevent the use of thiourea (a possible carcinogen) in the tin immersion process during PCB manufacturing. In addition, there is also concern about tin migration (the "tin burr" effect), although anti-migration chemicals can achieve some success in controlling this problem.

Advantages: good weldability, smooth surface, relatively low cost.

Disadvantages: Health and safety issues, limited number of thermal cycling cycles.

Summary of PCB surface treatment

Table 1: Condition parameters for OSP experiments.

Table 2: The conclusion is that ICT performance is greatly improved when solder test points are used. Given some issues with the fixture and process, the user believes that once these issues are addressed, they can achieve a first pass yield of between 80% and 90%.

Above are the main methods for lead-free PCB processing. HASL will continue to be the most widely used PCB processing process, and nothing will change for test engineers in this case. In some countries, HASL has been banned by law and alternatives have been adopted. As PCA manufacturing expands into more different global regions, more and more lead-free processing processes will be seen in ICT testing. Although OSP is not a natural replacement for HASL, it has become the preferred alternative processing solution that PCA manufacturers are investigating. This will cause real ICT test reliability issues when the process is not changed to allow solder paste on test pads and vias.

The conclusion is that there is no perfect PCB surface treatment process, and each method has its own problems that need to be considered. Some of these problems are more serious than others, and all of these lead-free PCB surface treatment processes need to be modified in the process steps to prevent fixture contact reliability issues in ICT.

Comparative considerations of HASL, OSP and silver immersion at the ICT stage

Now I want to focus on these surface finishing techniques and how they affect ICT performance. Surface finishing leaves soft solder “caps” and exposed vias on the test points, which are ideal for ICT test targets. One property that HASL has that OSP does not is the ability to absorb forces, HASL is eutectic SnPB and is exceptionally soft. This soft target has two benefits: accommodating the probe and absorbing energy.

With OSP PCBs there is no such soft target. The copper surface is very hard in comparison and does not absorb much energy, so the area of ​​direct contact that the probe can "bite" into is reduced. The copper plating on the outer layers is generally between 10 and 50 microns. Combine the copper plating with the OSP coating, and you can see that a probe designed to probe a HASL board will not work on a board with an OSP finish.

Studies have shown that OSP creates a hard "crust" on the test target with longer transfer times between reflow and ICT. The optimal transfer time to ICT is less than 24 hours. There are many other process factors that affect the degree to which OSP causes trouble to the test engineer, some of which are: OSP provider type, number of passes through the reflow oven, whether the wave process is removed, nitrogen reflow or air reflow, and the type of simulation test at ICT.

Direct probing of the copper surface, coupled with the higher probe force required to penetrate the OSP layer, creates a real potential threat of damaging the thin copper layer and causing internal shorts. Therefore, our recommendation is to never probe exposed copper surfaces.

Recent cases have shown that board vias or test points may be poked through after 5 to 10 fixture excitations.

For some PCA manufacturers, the impact of OSP on ICT has become so problematic that they have stopped using OSP altogether. Others are learning how to follow the “OSP Rules” listed below.

"OSP Rules" for ICT test fixtures and procedures:

Pay attention to the latest industry testability recommendations, such as www.smta.org. - Always apply solder paste to the test connection points (test pads or vias), do not probe the exposed copper layer covered by the OSP. If you cannot change the stencil, prepare:

* Great impact on first pass yield (FPY)

* It may be necessary to change the fixture probe to obtain a higher force, for example from 2 Newton to 3 Newton

* May need to change fixture probe type to a more pointed type*

* May require a "double click" gripper actuation method, or utilize gas dynamics, manipulators*

* Simulation test program constraints may need to be compromised, opened up, or even ignored*

*Research has shown that these asterisked rules may have a relatively small impact on yield and the only way to ensure reliable test contact is to ensure the test pads are soldered.

Some manufacturers see the immediate cost savings of OSP and consider it as the first choice for lead-free replacement processes. However, some companies have recently undergone a complete about-face and are re-examining their strategies when considering the real costs associated with production interruptions and delays.

Silver Immersion

Immersion silver is a 0.4 to 0.8 micron layer of metal over the copper that provides the "meat" for the test probes to bite into. Immersion silver is not as widely used as HASL or OSP, but initial studies suggest it is a natural replacement for HASL as a manufacturing process. There have been some preliminary studies of ICT reliability that indicate etch time (surface roughness/finish) and surface thickness are important considerations for repeatability. No issues have been reported with fixture contact reliability with silver finishes at the ICT stage, so no adjustments to the test fixture are required, but adjustments to the probes or test software may be needed.

The etch rate is important for ICT testing because it determines whether the silver finish will be shiny or dull. During the silver deposition step, the silver is deposited onto the contours of the copper surface, so if the surface roughness increases, and thus the area increases, it will appear as a dull surface, while the surface with the lowest roughness will appear as a shiny surface.

This surface treatment has been studied very limitedly in the industry, but it looks the most promising technically and commercially. Recent experience shows that this surface treatment poses no problem for ICT. PCB manufacturers now offer boards with silver surface treatment at the same price as HASL products.

User Research - Transition from HASL to OSP

Study 1—One European OEM Manufacturing Site

A set of data from an OSP trial is shown below. This trial was initiated by a user who found that there was a significant impact on yield when introducing an OSP PCB surface finish as a lead-free alternative to one of their product lines. These results were obtained independently by the user and were not influenced by Agilent in any way. The test equipment used was the market leader Agilent 3070 in-circuit tester, which is widely used in the industry as the most stable test platform.

Study 2—European Contract Manufacturers

This contract manufacturer study was driven by similar experiences as Study 1, where poor yield performance was obtained in ICT. The goal of this study was to discover the root cause and provide feedback to the OEM.

This experiment was divided into two experimental parts. The first experimental part determined the effect of changing the probe type and the probe force applied to the bare copper OSP test points. The second experimental part focused on the performance when the test points were soldered.

During the first experiment, the parts were stored in nitrogen to prevent oxidation before being transferred to ICT. Three types of probes were used, standard 7oz (2N), 7oz & 10oz (3N) e-type probes and spiral probes.

The experimental results found that different probe types did not improve yield when probing exposed copper OSP plating. It took an average of 5 fixture excitations to "break through" the OSP plating. Once the device's pin contact test passed, the rest of the test program will proceed normally and no changes to the analog test constraints are required. However, the experiment found that the degree of damage to the PCB test points/vias will also change significantly if the test force and probe type are changed. The conclusion of this experiment is that unsoldered test points are a big problem in ICT. Using higher forces or sharper test probes will only cause PCB damage.

The second experiment was a comparison of OSP boards with soldered test points, and a small number of pure OSP samples were also included in this experiment. A total of 86 boards were tested, of which 77 were soldered and 9 were not.

During the trial, the contract manufacturer utilized their global ICT equipment experts to test the test fixture. This trial presented several issues where fixture pressure and recording were affected and did not meet the user's specifications. This highlights the fact that the quality of future and current ICT test fixtures must always be monitored to ensure that acceptable standards are met.

Another issue that users have to face is the application of solder paste on the board. They have very small test objects, 30,000 per inch, and because of the need to consider the placement of heat sinks, the maximum solder height limit in some areas is greater than 0.11 mm.

Conclusion

There seems to be a trend among some companies towards OSP being considered a natural replacement for HASL. This choice is likely driven by a perceived unit cost savings. ICT engineers should be aware of this trend: OSP plated PCBs will not perform as well as other alternative lead-free finishes unless the test pads are covered with solder. If the process is not changed, the potential savings in initial cost may be offset by the cost of changing fixture probes, fixture maintenance, test software modifications and scrap of damaged boards. We have seen a lot of reverse in the OSP choice recently. The advice to customers who have not yet given up on leaded HASL is to consider the pros and cons of all possible lead-free PCB alternatives and ensure that all manufacturing stages are included in the trial, including testing! We do not have any definitive results on the effect of silver PCB finishes on ICT. We have discussed with customers using silver finishes and they have not seen any fixture contact issues using this finish.