Selective Soldering: Key Checks Before Release
Selective soldering applies molten solder to individual through hole joints on a board that also carries surface mount parts. It replaces hand soldering where the volume justifies a machine and where the process has to be repeatable, and it is used wherever a printed circuit board has to be assembled on both sides without the through hole parts being subjected to a wave. The process is a small wave delivered from a nozzle, and its quality rests on three parameters: the flux, the preheat and the contact time.
This article describes how the process works, how the nozzle and the profile are chosen, and how the common defects are traced.
What The Process Does
A nozzle pumps a small standing wave of solder from a pot, and the board is moved over it so that each joint is wetted in turn. A flux is applied first, either by spraying or by a drop jet that places a measured amount on each position, and the board is preheated from below or from above so that the solder does not have to supply all the heat.
The board is held by a fixture, usually with a clamp that supports it around the area being soldered. The nozzle approaches from below and may also apply a small amount of solder from above on some machines, and the sequence for each joint is a programmed set of positions and times rather than a single pass. That programmability is what makes the process repeatable but also what makes it dependent on the programming being right.
Nozzle Types And Sizes
A nozzle may be a single small round tip, a rectangular slot or a multi hole arrangement that covers several pins at once. The choice follows the joint: a single pin with a small thermal mass takes a small round nozzle, a connector row takes a slot or a multiple nozzle, and a joint with a large surrounding copper area needs a larger nozzle that can deliver more heat.
The nozzle size also affects the flow. A nozzle that is too large delivers a wave that floods the area and can push solder onto the mask and onto the surface mount parts, while one that is too small cannot transfer heat fast enough and the joint does not fill. The distance between the nozzle and the board is set so that the wave touches the joint without touching the laminate, and it is one of the settings that is checked most often because it changes with the flatness of the board.
Flux, Preheat And Contact Time
The flux has to be applied to the exact position, and the amount has to be enough to clean the joint and not enough to leave residue that will not be cleaned. A drop jet dispenses a measured drop on each position and is the most controllable method, while a spray covers the whole area and is faster but less precise. The flux chemistry has to match the cleaning requirement, and a no clean flux that leaves a residue has to be compatible with the coating that follows.
Preheat raises the board towards the soldering temperature so that the nozzle only has to supply the difference. A typical preheat brings the board to 100 to 130 degrees Celsius, measured on the top surface near the joint, and the figure is set from the thermal mass of the board. Contact time then completes the joint, and it is the parameter that has to be long enough to fill the barrel and short enough not to damage the parts or the mask.
Thermal Mass And Board Support
The single largest influence on the contact time is the copper attached to the joint. A pad connected to a plane takes heat away as fast as the nozzle supplies it, and the time needed to fill the barrel may be two or three times that needed on an isolated pad. The layout therefore determines the profile, and a board that mixes heavy and light joints has to be soldered with different parameters for each, which is a normal use of a selective machine.
Board support matters for the same reason. A board that is not held flat has a varying distance between the nozzle and the joint, so the wave touches some joints more deeply than others. The support pins or the fixture are positioned under the areas being soldered and away from the components on the underside, and the flatness of the board is checked before the program is set rather than after a lot has been soldered.
Programming And Verification
A selective program is a list of positions, approach angles, contact times and nozzle assignments. It is built from the board data and then tuned on a first article, and every position is verified by inspecting the joint rather than by assuming that the program is correct. The verification is done on the first board of each lot when the product is new, and on a sample thereafter.
The measurements that support the program are the board temperature at the joint before contact, the contact time, and the appearance and fill of the joint afterwards. A thermal profile measured with a thermocouple attached to a test board at the start of a product is the most useful single record, because it shows what the solder actually had to do. The surrounding assembly sequence is described under PCBA development process.
Common Defects And Their Causes
An unfilled barrel usually means that the contact time is too short, the preheat is too low or the joint is attached to more copper than the profile allows. A bridged joint means that the nozzle is too large, the flux is excessive or the mask between the joints has been disturbed. Solder on the surface mount side usually means that the wave was too high or that the board was not held flat.
Flux residue that darkens and hardens indicates too much flux or a preheat that has driven the activator off before the joint was made. A cold joint with a dull surface indicates a contact time that is too short, while a joint with a rough and grainy surface indicates a disturbed joint, which in a selective machine usually means that the board moved while the solder was solidifying. The alloy itself contributes to the appearance, and the differences are described under lead free versus leaded solder.
Layout Rules That Help The Process
The layout can make the process easy or difficult. Joints that are separated from a plane by a thermal relief need less heat, while joints solidly connected to a plane need more, and mixing the two in one connector makes the program a compromise. Keeping the through hole parts on one side of the board and the nozzles able to reach them without crossing a tall component reduces the programming work considerably.
The board outline, the position of the tooling holes and the clearance for the fixture are all part of the same design, and they are described under board outline and mounting design. Where the design permits it, grouping the through hole pins into a compact area rather than scattering them across the board is the single change that improves the process most, because it allows one nozzle and one set of parameters to serve them all.
FAQ
Is selective soldering faster than hand soldering? It is faster and far more repeatable for any volume that justifies programming the machine, and it removes the variability that comes from different operators and different irons.
Can selective soldering damage nearby surface mount parts? It can if the preheat is too high or the nozzle is too large. The parts near the joint see the flux and the heat, and the layout should keep sensitive parts away from the through hole area.
Does the process work with a lead free alloy? It does, with a higher preheat and a slightly longer contact time, and the nozzle materials and the pot have to be compatible with the higher temperature and with the more aggressive alloy.
