This installment of The PCB Norsemen was written by Jan Pedersen, senior technical advisor at Elmatica.
The solder-limit subject has been a "hot potato" for a quite some time, with many discussions around the new requirement from Underwriters Laboratories (UL) that UL’s Emma Hudson brought to attention in early 2018(1). The requirement demands that all UL types must be reapproved regarding solder limits. How can we best solve this situation without creating lots of new challenges for the industry?
What is a Solder Limit?
Solder limit is the time a PCB is exposed to a temperature above the UL-approved maximum operating temperature (MOT) during a soldering process. In most cases, the solder limit has been tested to 20 seconds, which means the PCB is UL certified to be exposed to temperatures above MOT for 20 seconds.
Most know this is not sufficient even for one run through the reflow process, which—according to a standard soldering profile—will be exposed for 270–400 seconds to temperatures above MOT for only one processing (Figure 1). Today, many PCBs will be exposed to six or seven soldering processes, including rework and selective processes. Using the existing model, this results in approximately 2,500 seconds above MOT.
Why Test for Only 20–30 Seconds?
If you go back in the history of electronic manufacturing, those numbers were set when wave soldering and components existed only on one side of the PCB. When we started to have components on two sides, we were already past 20 seconds.
In 2006, we were introduced to lead-free soldering and things got much worse for the solder limits. A question would be, “Why did UL not act at that time when they had a golden opportunity?” Being a rookie on the UL Technical Panel, I had to find out. It seems UL had a discussion in the technical committee back then, but strong voices used their veto to stop the process. We can only assume that a higher cost was the main driver.
Thus, we are back to where we started and the situation is much worse. Thousands of materials and PCB builds have been added with a high cost for every certificate, and since we finally understand what this is about, UL cannot continue without a big warning.
Not Just Rough Guidelines
What the industry needs to remember is that these parameters are not just made to have rough guidelines. The main reason for these parameters is safety; not just safety for the printed circuit, but for the product where the PCB plays a vital part. The PCB will surely survive all soldering and assembly processes, but if soldering processes degrade the materials, the result could be a field failure. Then the issues one will have to face are severe.
Since soldering limits have been overlooked and are generally not understood in the supply chain, the industry has made their own tests to qualify a laminate for multiple soldering processes. To advise customers to use the material most feasible for their product has been part of my job for years, and it became critical after the introduction of lead-free processes in 2006. Still, few pay attention to what state the PCB is in after surviving all the assembly processes and being delivered to the end customer. This is where UL test procedures to evaluate solder limits becomes valuable. It’s UL’s main responsibility!
Number of Standard Cycles vs. Seconds
How do we move forward and make sure we find a solution that actually works? A solution proposed a few years ago within the UL Technical Panel is to qualify solder limits by a number of standard soldering cycles instead of a number of seconds. We are very much in favor of this solution. However, how do we find a way forward for existing UL type certificates?
If UL requires all existing UL types to be retested, the cost will be huge and test capacity is not ready to meet this sudden demand. A typical factory producing multilayer PCBs with HDI can have several hundred types. The cost of retesting all these is immense! There is no doubt UL is facing pressure to start the new testing and leave the past untouched for PCB factories.
If UL decides to move away from solder limits based on time and focus on the number of processes a PCB can tolerate according to J-STD-020, then we are on the right path. In fact, this is how most global companies qualify laminates for lead-free soldering today. By implementing these tests and using them as guidelines, the industry will have parameters that are correct until UL finishes all testing. However, should UL still decide that all existing approvals must be retested by UL, we might face a situation with enormous expenses and incredibly long qualification lead times—not to mention the reactions when the industry understands the cost and workload this project applies to laminate factories and PCB suppliers.
What We’re Doing Today is Wrong
Another way forward could be to start this process by requesting laminate and solder mask suppliers to retest all their most popular materials. Then, use this as a support for existing PCB UL types with the upper limit of four cycles as an interim solution until all suppliers update their UL approvals. Again, this will release the pressure and focus on retesting the most critical items.
Still, the most important part of the whole soldering question is that safety comes first. It’s all about finding the best solution for all involved parties and hoping that no strong voices will silence the discussion because what we’re doing today is wrong.
- Solder Limits: Updates for the Age of Surface Mount by Emma Hudson, PCB007.com.