Solder Fountain: 5 Rules for Selective Soldering
A solder fountain is the small, localised wave that a selective soldering machine raises under a single through hole component, and its height decides how much of the barrel the alloy touches. Setting that height is more delicate than it looks, because the same value that fills a connector can flood the mask beside it. The aim is a fountain that reaches the joint, holds for a controlled dwell time, and lifts without dragging a bridge to the next pin.
Selective soldering differs from mass wave soldering in scale. The bath is small, the nozzle is small, and the thermal mass of the assembly dominates the process rather than the other way round. Nozzle depth, the shape of the solder wave and the dwell time all interact, and changing one of them without the others usually moves the defect instead of removing it.
How a Solder Fountain Forms a Joint
The pump pushes alloy through the nozzle and creates a dome of moving solder above the tip. When the board is lowered, the dome touches the pin and the pad, and heat flows from the alloy into the copper. The joint forms while the board is in contact, not after it lifts, so the contact geometry is the process.
Because the alloy is moving, it also carries heat away. A fountain that stands too high loses temperature at the tip, while one that stands too low transfers heat unevenly around the barrel. Contact has to be stable for the whole cycle, with no rise and fall as the board settles onto its support.
Setting Nozzle Depth Above the Board
Nozzle depth is the distance the tip travels past the board surface, and it is usually written as a small positive or negative value. A tip that stops just above the surface relies on the dome to reach the joint; a tip that rises past the surface pushes alloy into the barrel and around the leads. The right value depends on the nozzle diameter and the clearance around the pin.
Too deep and the fountain floods the mask, wets the lead shoulder and leaves solder on the component body. Too shallow and the joint fills only part of the barrel. Set the depth with a gauge on a scrap board, then confirm it on the real assembly, because board thickness and mask height both shift the effective contact point.
Solder Wave Height and Contact Area
Wave height is the amount of alloy standing above the nozzle rim. It sets the contact area and therefore the rate at which heat reaches the joint. A tall wave wraps further around the lead and heats faster; a short wave confines the alloy to the pad and the lower barrel. Most programs use the shortest wave that still fills the joint inside the allowed cycle.
Height drifts as the pump wears and as the nozzle erodes, so it should be measured on a schedule rather than trusted. A worn nozzle can lower the wave without any change to the machine settings. The selective soldering nozzle condition is part of the height check, not a separate maintenance item.
Dwell Time and Thermal Mass
Dwell time is how long the joint stays in contact with the fountain, and it has to be long enough to bring the barrel and the lead to soldering temperature. Heavy connectors and thick ground planes need more time, while a small pin on a thin board needs very little. The value should come from a profile rather than a guess, because the gap between a cold joint and a damaged one is often only a few seconds.
Thermal mass also changes with the copper attached to the pin. A pin tied to an internal plane pulls heat away as fast as the fountain delivers it, so the dwell that works elsewhere can leave a partial fill. Measuring board temperature near the joint is the only way to know, and assemblies with ceramic parts or heat sensitive bodies need their own limits.
Flux Application Before the Fountain
Selective lines apply flux by spray or by drop jetting before the board reaches the nozzle. The flux has to be active at the moment of contact, which means the delay between application and the fountain matters as much as the volume. Boards that wait too long between the two steps lose activity and produce joints that look wet but do not fill.
Excess flux is not harmless. It runs into the nozzle and carbonises, and the residue changes the shape of the wave. Flux chemistry should also suit the assembly, since a no clean flux that is only partly activated leaves ionic residue behind. The nitrogen blanket fitted to many selective machines reduces oxidation but does not replace flux activity.
Nitrogen and Dross at the Nozzle
Dross forms where the alloy meets air, and on a small fountain the surface to volume ratio is high. Nitrogen slows that oxidation and keeps the wave clean, which matters because a skin on the wave stops the alloy from wetting the joint. Gas flow should be set to blanket the nozzle rather than to blow across it.
Even with nitrogen, some dross collects at the rim and has to be removed. A nozzle that is partly blocked raises the wave on one side and lowers it on the other, which produces an uneven fill around the same connector. Solder pot analysis together with a nozzle cleaning routine keeps that variation inside limits.
Joint Fill and Barrel Rise
The acceptance target for a through hole joint is a fillet that rises into the barrel and wets both the pad and the pin. Selective soldering can reach that when the fountain supplies enough heat, but it struggles where the hole is much larger than the lead. The gap between lead and hole wall is the path the alloy has to climb.
Fill percentage should be measured on cross sections during qualification rather than judged from the top of the joint, since a full looking fillet can hide a barrel that is only half filled. Where fill is consistently short, the usual cause is dwell or wave height rather than flux, and settings should be changed one at a time. The IPC acceptance standard defines the fill targets to work to.
Bridges, Webbing and Cold Joints
Bridging between adjacent pins is the most common selective defect. It comes from too much alloy at the tip, a dwell that is too long, or a lift that is too slow. Webbing, where a thin film of solder remains between two pads, usually means the mask surface was wetted when it should have stayed clean.
Cold joints look dull and grainy and come from insufficient heat at the moment of contact. Raising the dwell is not always the answer, because a fountain that is losing temperature at the tip will not recover however long it stays. Checking alloy temperature inside the pot and again at the nozzle separates the two causes.
Verification and Daily Checks
Daily checks should cover alloy temperature, wave height, nitrogen flow and the physical condition of the nozzle. Recording them against the machine rather than against the operator makes a drift visible before it becomes scrap, and a photograph of a reference joint is a useful boundary sample for the shift.
Verification should also include a profile on a representative assembly, since a fountain that behaves on a test coupon can fail on a board with a heavy ground plane. When settings change for a new product, the change should be recorded and the previous recipe kept, because the next build of the old product will need it. Solder skip control follows the same logic of proving the setup before the run starts.
FAQ
Why does one recipe fill a connector but not another? The difference is almost always thermal mass rather than the fountain. Pins tied to planes or to large copper areas pull heat away faster, so they need a longer dwell or a taller wave. Splitting the recipe by connector type is more reliable than hunting for one setting that suits the whole board.
How often should the nozzle be replaced? Replace it when the wave height can no longer be set inside its window at normal pump speed, or when the tip has eroded enough to change the contact shape. Nozzle life depends on the alloy, the flux and the cleaning routine, so the replacement point should follow measurement rather than a calendar interval.
Can selective soldering replace wave soldering entirely? For a board with a few through hole parts among surface mount devices it often can, and it saves the masking step that wave soldering needs. Where a board carries many through hole components, mass wave soldering is still faster, so the choice should rest on the number of joints rather than on preference.
