Application Note #20001 Rev: 2008-02

Background

The implementation of lead-free electronic assembly has raised several critical issues for electronics assemblers. One of the most critical is that the lead-free reflow process has a much tighter process window than the traditional tin-lead reflow process. The tin-lead reflow process window is a wide one, with the lower limit generally set around 195-200C, (for standard solders with a liquidus temperature of 183C), and the high limit generally set around 240C. The typical tin-lead high and low process limits, which yield a process window of ~40C, are wide enough that a carefully monitored process can be expected to produce a high yield with little fear of defects caused by process drift. This wide process window means that the potential temperature differential across the board (Delta Temperature or DT), which typically is 5-20C, is not a critical profiling factor.

The process window has shrunk dramatically for lead-free assemblies. With the widely used Sn/Ag/Cu alloys (217C liquidus), most paste specs call for process peak temperatures from 232C to 260C. Bismuth alloys require slightly lower peaks. The process window of 28C, about 70% of the traditional tin-lead process window, appears to be tight but achievable. The issue is that many commonly used electronics components have maximum temperature tolerances between 240-260C, which drastically reduces the process window. This is the source of the current concern about reducing the “Delta T” (DT) of a product profile.

Delta Temperature is the peak temperature differential a product experiences during the reflow process. The product is exposed to fixed temperature convection in the various zones of the reflow oven, but smaller components (0210’s, etc) will heat up faster and experience higher peak temperatures than larger or more dense components (BGA’s). The DT is the difference between the temperatures experienced by the highest and lowest mass components on the board, and in traditional processes, a DT of 20C on large assemblies is not uncommon. In a lead-free process, this high DT is a recipe for disaster, as a process that will get high mass components hot enough to reflow properly will inevitably damage low mass components. The problem of narrow process windows will be further exacerbated by the trend to more complicated assemblies with increased component density.

The long-term solution to this problem is to raise the temperature limits of common electronics components to 250-260C, but this is several years off. In the short-term, the search for a solution has focused on reducing the Delta Temperature across the board. The obvious solution is to reduce belt speed in the reflow oven, as this will allow temperatures to equalize across the board. The obvious problem with this approach is that it will increase cycle time at the reflow oven and possibly make it the point of constraint on the assembly line. It may also cause problems with the profile, especially in the soak zone, where an extended profile may cause the volatiles to burn off and the flux to activate before the reflow portion of the profile.

Recommendations

KIC has performed extensive research on this issue. We have found that focusing exclusively on the DT of a given assembly can result in a degradation of overall profile quality. The SlimKIC 2000/KIC Explorer with the Navigator option will find the best profile for a given board that a process is capable of, and will consider all critical process statistics in finding the optimal profile. For most combinations of product and process, KIC’s research indicates that the KIC Navigator will find an acceptable profile, and KIC recommends using the Navigator normally as a first step in developing a lead-free process.

The KIC 2000 software displays the DT (range) of the data across all the thermocouples attached to product for every statistic that is part of the defined process window. This will be the last row in the statistics table for both the original and predicted data. This is a calculation and display of the process DT only. KIC does not include DT as a process window setup limit for the Navigator optimization. There are critical reasons in terms of overall process quality and the operating efficiency of the KIC software that have caused KIC to chose not to include DT as a process statistic used in calculating the optimized profile.

Typically, someone concerned with the DT is trying to get it under a certain value. If DT was added as a statistic the Navigator would have to apply the PWI to the data and then optimize on that. This would force the Navigator to calculate the PWI for the DT. For example, if you specified a DT of 10C and the actual DT is 11.5C. This is outside your defined process limit, and the question becomes; how does this affect the overall process PWI? The problem is worsened by the fact that DT is not a critical process statistic. It is an indication of process capability, but what matters is that the board is reflowed within the solder paste spec and the components thermal capabilities. In other words: if the DT is too large, the process will be out of spec anyway, and the KIC Navigator will optimize the process based on the existing specs limits. If the DT is within the process spec, the PWI will tell you how robust the process is and whether there is cause for concern. Although DT is not included in the KIC software setup screen, the Navigator cannot optimize the process without minimizing the DT.

Procedure

KIC recommends always using the Navigator normally, without focusing on DT. However, there may be cases, especially with very large assemblies, where the Navigator may not be able to find a profile with in the process window. The procedure below may allow you to develop a profile that will successfully reflow unusual products.

As noted above, the KIC 2000 software displays the process DT. This simplifies the task of analyzing the profile data and adjusting the peak temperature process window to reduce the DT. For example: the paste spec allows for a 20C peak window (205-225) and you want a DT of 12C. You run the first profiling pass, and the KIC 2000 will show the DT. After the first pass, the process is in spec, and the DT is 15C. You can then analyze the data from the individual thermocouples and determine which side of the peak limits is the most logical to tighten. By reentering the process spec screen, you can change the peak temperature process limit and force the Navigator to search for a recipe that will minimize DT. In our example above, you might change the high peak temperature limit from 225C to 220C.

The procedure is:

  1. Run your first pass with normal process limits so you know if you are in spec to start with.
  2. Review profile data and calculate the mean of the peak temperatures. This will let you know which side of the process spec (high or low) you can adjust with the least effect on the overall profile quality. For example, if most of the peak temperatures are near the upper limit, and one is significantly lower, you can adjust the lower limit to tighten the peak window. Tightening the process window will force the Navigator to reposition the profile target before searching for a solution that will reduce DT.
  3. Once the peak spec has been altered, returning to the graph screen will prompt the software to automatically recalculate the predicted profile and the overall process PWI. The Navigator will attempt to reduce DT based on the reduced peak temperature process window. Note that this tweak will alter other profile statistics, and will probably increase the overall process PWI.
  4. Run a second profile and achieve a tighter DT, then change the process limits back to the original process limits, adjust the process window, save as part of that profile, and then go to production.

It must be emphasized that this operation requires engineering knowledge and should not be delegated to an operator or technician. You will need to be able to look at the data and make a judgment call as to which way the profile needs to be tweaked. The software has password protection, so you can choose to only allow the process “owner” to perform this operation.