Solder Pot Level Control: 7 Rules for a Stable Wave

The alloy level in a soldering machine is the parameter that most often changes without anyone deciding to change it. Every board that crosses the wave takes solder out of the pot and leaves oxide on the surface, so the level falls steadily through a shift. Nothing is adjusted, no setting is touched, and yet the wave that was set up on Monday is not the wave that is running on Friday afternoon.

The level matters because the wave is produced by a pump that lifts alloy through a nozzle. A lower level means less alloy above the pump inlet, more surface turbulence and a wave whose height and shape drift with every board that passes. Chasing that drift with nozzle adjustments treats the symptom. Holding the pot level inside a defined band treats the cause.

Why the Level Changes the Wave

The pump moves a fixed volume of alloy per revolution, but the wave it produces depends on the head of metal above the intake. When the level is high, the intake is well submerged and the flow is smooth. When the level falls, the pump begins to draw alloy from close to the surface, where oxide films and dross are floating, and the wave becomes rougher.

A rough wave does two things at once. It changes the contact pattern, because a turbulent wave touches the board in an irregular line rather than a steady one, and it drags oxide into the joint. Both effects raise the defect rate, and both are usually blamed on flux or on preheat. Measuring the pot level before adjusting anything else is the cheapest diagnostic step available.

Setting the Working Band

Every pot has a level at which it was designed to run, and the machine manual usually states it as a distance from the rim or as a volume. That point should be converted into a working band with a high mark and a low mark, and both marks should be visible on the pot or on a dipstick.

The band is not symmetrical. Running above the high mark floods the nozzle and can force alloy over the top of the barrel, which is a common cause of solder balls on the top side. Running below the low mark starves the pump and lets the wave collapse. The useful band is usually narrow, and a workshop that has never marked it is usually running somewhere below the design point for most of the week.

Level Sensing and False Readings

Automatic level sensors are useful but they are easily fooled. A float or a probe measures the surface it sits on, and the surface of a solder pot is not alloy, it is dross. A probe that has been buried in an oxide layer will report a level that is higher than the true one, and the controller will then refuse to add alloy when the pot is actually low.

The same problem appears with an operator checking the level by eye, because a thick dross layer hides the true surface. The reliable method is to skim the area around the sensing point before the reading is taken, or to confirm the automatic reading against a dipstick at a fixed point once a shift. Where the two disagree, the manual measurement is the one to trust until the probe has been cleaned and re-verified.

Adding Alloy Without Disturbing the Wave

Any alloy addition disturbs the pot. Cold bars dropped into molten alloy chill the bath locally, raise the oxide level and set up convection currents that carry oxide into the pump intake. The disturbance is temporary, but a joint soldered during it may not be the joint the profile was developed for.

Good practice is to add alloy in small amounts, more often. Bars should be dry, clean and free of any marking ink that would burn off in the bath, and they should be placed in the quiet end of the pot away from the pump intake rather than thrown into the working zone. Where the machine supports it, preheating the bars on the pot rim reduces the thermal shock. A large addition should be followed by a settling period and a check of the wave before production resumes, and the addition should be written in the log with the time and the amount.

Dross, Skimming and the True Level

Dross removal is part of level control, not a separate housekeeping task. Oxide that is left on the surface insulates the bath, hides the level and is drawn into the pump. Skimming should be done on a schedule and at a point that does not interrupt production, and it should be done with a tool that has been warmed and dried so that it does not add moisture to the bath.

Every skim removes alloy as well as oxide, so the skim rate and the addition rate are linked. A shop that skims aggressively but adds alloy slowly will spend the day at the low end of the band. Recording both numbers is the only way to see that relationship, and it is usually the point at which a chronic wave height problem is solved.

Chip Wave, Main Wave and Hollow Wave

A machine with two nozzles has two levels to think about. The chip wave sits closer to the board entry and is shallower, so its behaviour is more sensitive to level than the main wave. When the pot runs low, the chip wave is the first to lose its shape, and the first defect to appear is usually incomplete fill on a through hole joint near the leading edge of the board.

The main wave is next, and the visible symptom is a change in the contact length rather than a change in height. A hollow wave nozzle adds a third variable, because the two crests it produces can merge at some levels and separate at others. The wave profile should be re-checked with a glass plate after any deliberate level change, not only after a nozzle change.

Sludge and Sediment in the Bottom of the Pot

Heavy intermetallic compounds and oxides settle toward the bottom of the bath. When the level is kept high, the settled material stays out of the circulation. When the level is allowed to drop, the pump intake moves closer to the sediment and begins to stir it back into the wave.

That is one reason a pot that has been neglected for months behaves badly even after the level is corrected: the sediment has been distributed through the alloy. Sampling the alloy for copper, iron and other contamination, and comparing the result with the level history, often explains a defect rate that has drifted upward without any other change. Scheduled pot cleaning and alloy analysis belong to the same control loop as the level itself.

Consumption Rate, Pallets and Loading

The rate at which the level falls depends on the product. A panel with heavy through hole content and large barrels takes far more alloy per board than a lightly loaded panel, and a pallet that carries alloy away on its underside takes some with every pass. A change of product mix therefore changes the consumption rate without any change to the machine.

Setting the level check interval from the consumption rate rather than from a fixed shift routine keeps the pot inside the band. If a product consumes twice the alloy of the previous one, the level should be checked twice as often, or the addition should be made automatically between batches. The same reasoning applies to pallets and fixtures, which should be cleaned and checked for alloy build-up as part of the same routine.

Records That Keep the Level Inside the Band

A simple log with four columns is enough: the time, the measured level, the amount of alloy added and the amount of dross removed. Read together over a week, those columns describe how the pot is actually behaving and where the band has to be tightened.

The log also protects the process when the machine is rebuilt or when a new operator takes over the shift. A pot that is refilled properly and skimmed on schedule holds its wave height with no nozzle adjustment at all. Where the alloy itself has become contaminated, the correction belongs to the alloy contamination procedure, and the evidence that the level was maintained belongs in the same record as the pot contamination checks. A pot that is running at the wrong level will also show more excess solder on heavy joints than the same product did a month earlier.

FAQ

How often should the pot level be checked? Once a shift is the minimum, and more often on a product with heavy through hole loading or a high alloy consumption rate. The check takes under a minute and it removes one variable from every other adjustment that follows it.

Can a low pot level damage the pump? Yes. Starving the intake makes the pump work with less head, increases turbulence and in some designs causes cavitation at the impeller. Repeated operation at the low end shortens pump life and is one of the reasons a machine becomes unstable long before any component visibly fails.

Is it acceptable to top up the pot during a production run? Small additions are normal and unavoidable, but a large addition during a tight-tolerance run is not. Where a full refill is needed, it should be done at a break, followed by a settling time and a wave check before the first board of the next batch is soldered.