Often times, people assume that the dipping process is easy. It is…once you have taken the time up-front to optimize the process. This optimization can take a bit of time and trial and error.
The two most important process parameters in the dipping process are the dip depth and the dwell time. Both of these parameters will need to be optimized for the specific component that is being used in the dipping process. If more than one component is used, the process should be optimized for each component separately. That being said, there is a connection between dip depth and dwell time. A deeper dip depth can result in a shorter dwell time, and therefore, a thinner dip depth will need a longer dwell time.
The only problem with the last statement is that there can be too deep a dip depth and there can be too shallow a dip depth. Too deep will result in wet bridging (flux bridging the component bumps),which can result in pick and place vision errors and rejected parts. Too deep a dip depth can also result in the component getting stuck in the flux reservoir. On the other side, too shallow a dip depth can result in insufficient flux volume transferred to the component bumps, and the solder joint may not form correctly. This can result in Head-in-Pillow or Snowman Opens.
The same goes for the dwell time. Too long a dwell time may result in too much flux, causing wet bridging and vision errors, or the component becoming stuck in the flux reservoir. Too short a dwell time can result in insufficient flux volume and issues with proper solder joint formation.
So, it is very important that some time is spent at the beginning of the build to ensure that the proper flux depth and the proper dwell time of the component in the flux reservoir are optimized for each individual component. There are a few ways to do this, but the easiest way to check for the proper flux deposition is to put a glass microscope slide onto the board over the area where the components will be placed.
Keep in mind that the board thickness may need to be adjusted. This is so placement pressure doesn’t break the glass slide or the pick-and-place nozzle isn’t damaged. You then would proceed with picking the component out of the tape and reel packaging with the pick-and-place equipment and then dip the component into the flux reservoir, followed by placement of the component onto the glass slide. At this point, it will be easy to remove the glass slide from the board with the component on it.
The glass slide can then be flipped over and placed under a microscope to inspect the component bumps for flux. The flux should cover 50 - 60% of the bump height on all component bumps, and you should not see any flux bridging between the bumps. Typically, you will not see flux bridging if the component was placed because the vision system wouldn’t decipher the individual component bumps for alignment; it would have scrapped the component. Flux depth and/or dwell time can then be adjusted to obtain the proper amount of flux volume (50% - 60% of the bump height). If the component becomes stuck in the reservoir, the dip depth and/or the dwell time will need to be reduced.
The best suggestion I can give for where to start with the process would be a depth of 50% - 60% of the actual bump height and a .5 – 1 second dwell time. Adjustments will need to be made from there to optimize the process.
The shelf life of the PoP Flux at room temperature within the reservoir should be roughly 8 hours. However it is important to monitor the flux reservoir to ensure that enough flux is present at all times and add more when needed. Just like any other process, I would recommend switching out the flux when the operators are changed. This way, the operator will know everything about the flux that is in the reservoir, like how much is present when it was originally put on the line and how much has been added over time. This will limit the variability in the process as much as possible.
Many times for the PoP process, nitrogen reflow is used to widen the reflow process window and reduce the potential for defects. Although it may not be necessary, I’d say 70 – 80% of customers building assemblies with PoP components and any PoP Flux will use nitrogen.