Issue link: https://iconnect007.uberflip.com/i/937683
70 PCB007 MAGAZINE I FEBRUARY 2018 appliances. The methods were designed for process control, not for the qualification of a production lot, as many believe and expect them to do. In reality, we still don't have a suf- ficient and fast test method to do a proper lot qualification. Another issue that has recently been brought up in discussions with test laboratories and in- ternally in the Automotive Addendum task group is the ability to test local areas. Today's methods measure the cleanliness of the entire printed circuit and are not able to find and lo- cate local contaminated areas. To control qual- ity and reliability you need the ability to mea- sure cleanliness in those critical areas. New tools will make this possible, and it is a prior- ity for the task group to evaluate and qualify improved methods to check the cleanliness of a specific area. In addition to new test methods, we are also discussing at what time in the production pro- cess we should do the testing. Performing a cleanliness test of your printed circuit before soldering may provide a result contaminated by natural substances from solder mask. We believe this will vary by solder mask curing and will be hard to control. A solution on how to avoid this could be to send the bare PCB through a reflow process and perform a test on this one, since reflow will give the PCB an ad- ditional cure. This theory has been supported by a test report provided by Bosch. A Standard for Thickness The second challenge we spend much time on these days is the thickness of solder mask. The industry currently operates with two stan- dards: standard a and b, referring to applica- tion methods. The group has reached the con- clusion that this is not optimal. We consider adding a demand for a specific thickness, suit- able for all application methods—a demand that will meet the requirements and still be manufacturable. Expect Better Inspection Our third challenge and area of focus is the level of inspections and inspection meth- ods on the finished board. IPC requires that the inspections are after certain AQL (accept- able quality limit) demands. Automotive pro- duction requires 100% visual inspection. With the volume in this industry, and with increased miniaturization, a 100% visual inspection by an operator is not practical. Implementation of automated visual inspection is under discus- sion in the task group and will be added to the standard as a recommendation. How do we measure wicking? The fourth challenge that has been addressed by the task group lately is wicking, and how this is measured. According to our experience, the parameters provided by IPC are at best con- fusing. IPC requires that wicking is measured from the drilled hole edge to the point where the wicking ends. This might result in a false result where you don't measure the wicking it- self. In the new and revised edition of the Au- tomotive Addendum, we will be precise that the measurement should start and stop where the wicking starts and stops. Only then will the results be correct. Standards for PCB production specifica- tions need to be alive and kicking, facilitating a modern production that meets the demands of today, not the ones in the past. As chair and member in several task groups, I know how much work lays behind a standard. The whole idea with task groups is to work diligently to- wards an improvement, assuring that all needs and consequences are considered and moni- toring the challenges from all stages of produc- tion. We need to extract knowledge from pro- duction files and convert it into useful infor- mation to improve the standards so they match the demands not only from IPC, but also the industry. With that in mind, I believe we also need a digital article specification that picks up re- quirements from all bits in the supply chain, but that is another topic. PCB007 Jan Pedersen is Senior Technical Advisor at Elmatica. This column was originally published as a blog on the Elmatica website and reprinted here with permission.