Selective Soldering: Design Rules and Process Limits

Selective soldering fills through holes one joint at a time with a small nozzle, which makes it the natural choice for mixed technology boards where wave soldering would damage the surface mount side or where only a few through-hole parts remain. The process is far more controllable than a wave in theory, and far more sensitive in practice, because a small nozzle delivers heat at a limited rate into a joint whose thermal mass may be much larger than the nozzle appears to handle.

What the Process Has to Achieve

The requirement is the same as any soldering operation: the barrel and the pad must reach a temperature above liquidus long enough for the solder to flow through the hole and form a fillet on both sides, without exceeding the limits of the components or the laminate. Selective soldering achieves this with a local heat source, so the rest of the board stays cool. That is its principal advantage and also the reason it needs careful process development, since the joint itself receives very little help from its surroundings.

Boards that come to selective soldering are usually a mixture of technologies, which means the thermal environment is uneven by design. A connector standing next to a large ground plane is a different problem from the same connector on a board with sparse copper, and both may appear on the same panel. Through hole practice covers the hand soldering equivalent, and the two processes share most of their thermal logic even though the equipment looks entirely different.

Flux and Preheat

Flux is applied either by spray or by a small dispenser that follows the same path as the nozzle. The deposit needs to be sufficient to clean the joint and small enough not to leave residue that interferes with test or with a subsequent coating operation. Because the nozzle arrives seconds after the flux in the same local area, the chemistry is expected to act quickly, and a flux chosen for wave soldering may not be the best match. Flux application practice provides the framework even where the equipment differs.

Preheat in selective soldering is usually applied to the whole board from below, raising the assembly to a temperature that shortens the time the nozzle must spend on each joint. The target is high enough to reduce the thermal shock and low enough to leave the flux active. Boards that carry temperature sensitive components often have to run at a lower preheat, and the compensation is then made in dwell time and in nozzle temperature, which is where the process window begins to narrow.

Nozzle Selection and Geometry

Nozzle diameter is chosen to suit the joint, not the machine. A nozzle that is too small cannot deliver enough heat and forces long dwell times; one that is too large contacts neighbouring joints and can deposit solder where it is not wanted. The shape matters as well, since a nozzle with a wide mouth produces a different wave profile from a narrow one, and the available contact area determines how quickly the joint reaches temperature.

Nozzle condition is a maintenance item that is easy to neglect. Solder attacks the nozzle material over time, the opening erodes and changes shape, and dross builds up around the mouth. A nozzle that has drifted produces joints that no longer match the developed profile, and the operator will usually compensate by increasing dwell time, which transfers the problem to the components and the laminate rather than solving it.

Dwell Time and Thermal Mass

Dwell time is the variable that most often decides success. The joint must be in contact with the wave long enough for the solder to wet and fill the barrel, and the time required grows with the thermal mass that surrounds the hole. A joint connected to an internal ground plane draws heat away continuously, and a nozzle that fills a simple joint in two seconds may take considerably longer on the same board when a plane is attached.

Because the relationship is not linear, the process should be developed on the worst joint rather than the average one. Measuring the barrel temperature, or verifying the fill by microsection, gives the evidence that the dwell time is sufficient. Extending dwell time globally to fix one difficult joint overheats every other joint on the board, so the usual approach is to program per-joint parameters and accept the longer cycle on the few parts that need it. Plating thickness in the barrel is part of the same calculation.

Atmosphere and Solder Condition

Nitrogen is common in selective soldering because the small wave exposes a large surface area to the atmosphere and oxidises quickly. An inert atmosphere improves wetting, reduces dross and widens the usable window, particularly with lead free alloys. Where nitrogen is used, the oxygen level should be measured rather than assumed, since a partially effective blanket gives the cost without the benefit.

The solder in the pot follows the same chemistry as any other pot, and its composition drifts as copper and other elements dissolve into it. Periodic analysis is required to keep the liquidus temperature within the range the process was developed for. Because selective pots are small and are often topped up frequently, the drift can be faster than in a wave machine, and the analysis interval should reflect the actual throughput. Solder pot analysis provides the method.

Process Development and Documentation

A selective soldering program is a sequence of parameters per joint: nozzle, flux quantity, preheat, dwell time and travel speed. Developing it means starting from a conservative setting, verifying the joint by inspection and, where the product justifies it, by microsection, and then tightening the parameters until the cycle time is acceptable with margin remaining. The result should be recorded per joint rather than as a global average, because that is how it will be run.

Documentation has to travel with the program. A product transferred to another machine, or a program recovered after a software change, should behave the same way, which requires the nozzle type, the flux, the profile and the verification record to be stored together. Where the process is validated by a first article, the acceptance of that article should be part of the same record so that the basis for the settings is never in doubt.

FAQ

Why does one joint need so much more dwell time? Usually because it is connected to a large copper area that conducts heat away. Verify by measuring the barrel temperature or by microsection rather than by lengthening the dwell time for the whole board.

Can selective soldering replace wave soldering entirely? For boards with few through holes it often does. Where a large number of holes have to be filled, the wave remains faster and the comparison becomes economic.

How often should nozzles be replaced? When the joint results change or when the opening is visibly eroded. Keep a spare of each type in use so a worn nozzle does not stop the line.

Is nitrogen necessary? Not always, but it widens the window and improves wetting with lead free alloys. Measure the oxygen level if nitrogen is used.