Preconditioning Chamber for SMT Components MSL 1 to 6

A preconditioning chamber drives moisture into a plastic package until it holds what a factory floor would, then a reflow heats it. The whole exercise hunts, in a few days of soak and a pass through the solder peak, for the packages that crack when the trapped water flashes to steam.

A plastic chip package is not waterproof. Its epoxy body drinks moisture from the air; carried into the heat of soldering, that water flashes to steam and can split the package open. Preconditioning is the test that provokes this on purpose. A chamber soaks a package to a measured amount of moisture, the amount it would hold after a set time in open factory air, then a reflow runs it through the solder heat. A package that comes through sound is safe to assemble at that moisture; one that cracks or delaminates is caught here, in the lab, before it would fail on a customer’s line. The level a part earns, MSL 1 through 6, is the subject of the classification standard’s own article; this one is about the chamber that drives the soak and the failure it hunts.

Why a wet package cracks

The trouble begins with the plastic itself. The epoxy mould compound over a chip is hygroscopic, pulling water vapour from humid air into its bulk, where the moisture gathers along the interfaces inside, the die-attach beneath the chip, the lead frame, the face of the die. None of it shows at room temperature. A soaked package looks and works the same as a dry one, right up to the moment it meets the soldering heat.

That heat is what turns the water into a weapon. The moisture sitting at the interfaces reaches the reflow temperature along with everything else, far past the boiling point of water, so it flashes to steam inside the sealed package. The steam has nowhere to go fast enough, so its pressure builds against the bonds around it, bonds whose adhesion was never made to hold back a head of steam. The wetter the package and the hotter the peak, the harder that steam pushes.

So preconditioning recreates the worst moment a package meets. It loads the package with the water a real exposure would leave in it, then runs it through the solder heat that flashes that water to steam, all under watch in the lab. The test exists because the failure stays invisible until that heat arrives, and a part that would crack in a customer’s reflow oven is caught instead in a chamber, where the cost of finding it is a sample.

The pop

The failure preconditioning hunts has a name that fits the sound it makes: popcorning. A package soaked with moisture and driven to the reflow peak holds water at its interfaces that flashes to steam; saturated steam at the peak temperature of a lead-free reflow, somewhere near 260 degrees, carries a vapour pressure of several megapascals, tens of atmospheres, all of it pressing outward on bonds built to hold nothing of the kind. What makes the pop so destructive is the suddenness of it. The package crosses the boiling point of its trapped water in a few seconds as it ramps to the peak, so the steam does not form gently; it forms in a rush, the pressure spiking faster than the mould compound can relieve it. A package that might bear the same pressure applied slowly is split by the speed of its arrival. The faster the body reaches the peak, the less time the vapour has to seep out through the plastic, and the higher the pressure climbs before anything gives way. When something does give, the weakest bond goes first. A void left in the die-attach gives the steam a ready pocket to expand into; a thin skin of mould over a die corner concentrates the stress until it tears; the lead frame, smooth metal against epoxy, holds least of all where water has already crept along the join. In the loudest version the whole body swells, then splits with an audible report, the pop that named the effect. The quiet version does the deeper harm. A delamination that never breaks the surface still lifts the die-attach or tears a wire bond at its heel, fractures the die as the swelling package drags on it, leaving a part that looks whole and even tests good while carrying a latent crack the field will finish months later. One reflow can do all of this; a board sees reflow two or three times across a build, between its two sides and any rework, so the soaked package has to survive the steam on every pass, a marginal interface that held on the first heating letting go on the third. The pop is the whole reason a package carries a moisture rating, and reproducing it honestly is the whole reason the preconditioning chamber exists.

The quiet failure that hides

The crack that pops the package open is the easy one to catch. Worse is the delamination that never breaks the surface, the failure that leaves a part looking whole. The steam can lift an interface inside without splitting the body, peeling the mould compound off the lead frame or the die without a sound, a separation an eye cannot see and a quick test can miss.

A hidden delamination is a wound waiting to open. It can tear the heel of a wire bond, where the wire leaves the package and the lifted mould drags on it, breaking a connection that worked when the part was new. It can crack the die, fractured by the swelling package pulling on the silicon, or open a channel along an interface that lets moisture creep in to corrode a metal trace months later in service, long after assembly passed the part as good.

This is why the test does not stop at whether the package survived. A part that came through the reflow without a visible crack can still carry a delamination that dooms it, so the inspection looks inside, mapping the hidden separations the steam opened. The quiet failure is the one preconditioning works hardest to catch, since it is the one a customer’s own assembly would pass.

The reflow profile that triggers it

The heat that finds a wet package is no single jump to the peak; it is a shaped profile whose shape decides how hard the moisture hits. The board climbs first through a preheat, a degree or two a second, to drive off solvents and wake the flux. Then it holds in a soak zone, somewhere from 150 to 200 degrees for a minute or two, where the flux does its work and the whole assembly evens out in temperature before the peak.

Only then does the profile ramp to the peak. It crosses the solder’s melting point, the liquidus, near 217 degrees for the common lead-free alloys, then stays above it for a controlled stretch, the time above liquidus, held to something like 60 to 90 seconds so the joints form without cooking the parts. The peak itself is brief, a few seconds near 245 to 260 degrees, before the cooling ramp brings the board back down.

The peak is when the trapped water is hottest and its steam fiercest, so it is the instant the pop lives or dies. Preconditioning runs the soaked part through this whole profile, to the peak its package class allows, more than once, since a real board meets reflow two or three times. Each pass is another chance for a marginal interface to let go, so the test repeats the profile to match the heat a package will take across a build.

The peak a package sees is set by its size. A thin, small package is classified to a higher peak, up near 260 degrees, since its little mass sheds heat fast; a thick or large one is capped lower, nearer 245, unable to lose heat quickly enough to stay safe higher. The preconditioning runs each part to the peak its own class allows, so the steam inside meets the heat the package will take in real assembly.

The moment it all turns on

At the peak, the trapped water flashes to steam and pushes.

The soak that loads the water

Everything the test reveals rests on the soak, the part of the sequence the chamber owns. The chamber drives a measured amount of water into the package and no more, the amount the part would hold after its rated time in open factory air. Load too little and the reflow bites softer than the field will; load too much and the part fails a heat it would have survived. The soak has to land the package at exactly the moisture its rating stands for.

How much water the package holds is set by the humidity. The epoxy holds more as the humidity rises, climbing a sorption curve that steepens at the damp end, so a few percent of drift near the top changes the final moisture far more than the same drift lower down. This is why the chamber’s humidity has to be held so tightly: a soak that wanders a few percent over days mis-loads the package, so the grade that comes out is wrong.

How deep the water reaches is set by diffusion and by time. Water creeps inward at a pace that rises sharply with temperature; the time to soak a body all the way through grows with the square of its thickness, so a thick package takes far longer to wet to its core than a thin one. The soak pairs a warm push with a long dwell precisely so the water reaches the centre, never the skin alone, since a package wet only on the outside reads back as a different part than it is.

The classification standard tabulates, level by level, the moisture each rating carries and the soak that lands it, the subject of its own article. Here the point is simpler. Whatever level a part is being qualified for, the chamber holds the one condition, a temperature and a humidity over a length of time, that brings the package to that level’s moisture, the whole result resting on holding it true.

The soak starts from a known empty state. Before any water goes in, the chamber bakes the part dry, clearing whatever moisture it arrived carrying, so the soak that follows adds a measured amount onto a clean baseline, never onto an unknown. Without that dry start, a part that came in a little damp would finish the soak wetter than its level intends, so the grade would drift. The bake is what makes one lab’s soak comparable to another’s.

What the soak chamber must hold

The chamber that does this is judged on how exactly it holds its condition. The grade a part earns is only as honest as the soak that loaded it, and the soak is set by the chamber, by the temperature and humidity it holds across the soak hours. A chamber off its setpoint does not produce a slightly wrong grade; it produces a confident grade that is wrong, a floor life the part cannot keep.

Humidity is the sharp edge, because the uptake is so steep at the damp end. A chamber running a couple of points high over-soaks the load, pushing more water into the parts than their level should carry, so a part that would pass is failed and rated needlessly fragile. Running low leaves the parts under-soaked, passing a heat they should fail. The tighter the chamber holds humidity, the more the grade can be trusted.

Uniformity carries the same weight across a full load. A chamber preconditions a batch together, then treats them as one grade. Let the humidity run high in one corner of the working space and the parts there take up more than the ones in a drier pocket, so the grade records where a part sat, never the part itself. Even humidity across every position is what lets one soak speak for the whole load.

Reading the damage with sound

The damage the soak reveals is read without opening the package, by sound. An acoustic microscope sends a focused ultrasonic pulse in through a water couplant; the pulse passes cleanly through solid epoxy and bonded interfaces. Where it meets a delamination the steam has opened, it meets air, which throws the pulse straight back with its phase flipped, a sharp echo from a gap too thin to see.

Scanned across the package, those echoes build a map. The reflections gather into a plan-view image where a delamination shows as a bright, sharp patch over the die or along the lead frame, read against the standard’s limits on how much separation each interface may carry. The critical faces, the surface of the die and the die pad beneath it, are held tightest of all, since a separation there is the one most likely to tear a bond or crack the die.

The acoustic map does not work alone. An electrical test catches a bond lifted or a die cracked clean through; a cross-section, the destructive last word, settles a borderline call by putting the interface under a microscope. A delamination past the limit fails the part even when it still tests good, since the opened interface is a latent failure the field will finish, exactly the outcome preconditioning exists to catch.

Reading the limits

Not every delamination fails a part, so the standard draws lines on where and how much is allowed. A small separation in an unimportant corner may carry no risk; the same separation over the die or under it is a fault, so the acceptance criteria name the interfaces that matter and the extent each may carry. The map from the acoustic scan is read against those limits, interface by interface.

The faces around the die are held tightest. A delamination on the surface of the die, or on the pad it sits on, is the one most likely to crack the silicon or tear a bond, so the standard allows the least there, often none across the active area. Out at the package edge, where less rides on the bond, a little more separation is tolerated. The limits map onto where a failure would do real harm.

So the grade rests on two things at once. The package has to survive the soak and the reflow without a crack. The scan has to read its interfaces within the limits the standard draws. A part that clears both is cleared to assemble at its rating; one that exceeds the limits is failed, even when it still works on the bench, because the opened interface is a crack the field will finish.

The packages that drink hardest

Not every package carries the same risk; its geometry tells the bulk of the story. A large, thin package is the hard case. It offers a wide area of interface for the steam to work on, a thin cap of mould that flexes and tears easily, with a long path for any bake to drive the moisture back out. Big plastic ball-grid arrays on laminate substrates, thin quad-flat packs, leadless packages with a large exposed pad, all tend to sit at the thirstier levels for this reason.

The ratio of die to package counts too. A big die under a thin skin of mould concentrates stress at its corners and hands a delamination a wide interface to spread along; a small die buried in plenty of epoxy is better held. The substrate adds its own term: a laminate base drinks moisture of its own and brings an interface a plain metal lead frame does not.

None of this changes the test, only the result. The same preconditioning sequence runs whatever the package, the same soak and the same reflow; the geometry decides what grade comes out the far end. So the chamber treats a thin ball-grid array and a small chip-scale part alike, holding the soak true for both, letting the package’s own build settle where it lands on the scale.

A maker fights to certify the lowest level a package can honestly carry, since a deeper rating burdens every line that buys the part with tighter handling, the detail of which the floor-life article covers. The chamber’s grade, then, is a cost the buyer pays in care, the reason a maker pushes for the gentlest rating its package can pass and a designer weighs a part’s moisture sensitivity alongside its size and its price.

The lead-free penalty

One change made moisture sensitivity worse across the whole industry, with nothing to do with the packages themselves. The move off tin-lead solder, driven by the restriction of hazardous substances, raised the reflow peak. The old eutectic alloy melted near 220 degrees; the lead-free alloys that replaced it need a peak some 25 to 40 degrees hotter to do the same job. That hotter peak makes the trapped steam fiercer and the mould compound weaker at the same instant.

So many parts dropped a level when they were requalified for lead-free assembly. A package comfortable at one rating under the cooler old reflow could no longer pass the same soak under the hotter new one, its floor life shortening to match. The preconditioning that grades a part today runs to the lead-free peak, which is why the ratings read tighter than they once did, why a part carried over from the leaded era cannot keep its old grade without passing the soak again under the hotter profile.

Why a part carries a number

What comes out of all this is a number, the moisture sensitivity level the part carries, from 1 for a package indifferent to humidity to 6 for one baked before every reflow. The number is shorthand for how long the part may sit in open air before it has taken up enough to risk the pop, the floor life the rating sets. How the levels are defined and what floor life each one buys is the work of the classification standard, set out in its own article.

Here the number is the chamber’s output, the grade the soak and the reflow together assign. A package that survives the soak-and-reflow at a given level earns it; one that cracks or delaminates is dropped to a thirstier rating it can meet. The preconditioning chamber is where that grade is made, by loading the water and applying the heat, so the number on the part is a measured fact, never an estimate. The preconditioning that comes before a reliability test serves the same end by a different route, a subject of its own.

The grading runs in order

The grading is a fixed sequence, each step setting up the next. It opens with a reading: each part is measured and scanned before anything else, so a flaw found later is known to be the soak’s doing, never one the part arrived with. Then the bake dries it to a baseline. The soak loads the rated moisture. The reflow applies the heat. The final scan reads the result. The order is the test.

Run out of order, the sequence means nothing. A part soaked before it is baked carries the wrong moisture; a part scanned before it is reflowed shows no damage; a part reflowed too few times misses the failure a third pass would find. So the chamber and the oven and the scanner work as one chain, the soak the step the chamber owns, the step the whole grade turns on.

What an honest preconditioning chamber provides

Everything a preconditioning chamber needs follows from the one fact that the grade it produces is only as good as the soak it holds. The chamber has to load the package with exactly the water its level stands for and drive that water to the core, holding the same condition across a whole batch, repeating it the same way every time, so the grade means the same in any lab. A box that misses any of these prints a number that is wrong.

It has to hold humidity to a fine tolerance. Because the uptake steepens at the damp end, a drift of a couple of percent over a soak of days mis-loads the package, so the chamber holds its humidity tighter than a casual climate box ever would, proven against a reference and logged across the run.

It has to soak evenly across the load. A batch graded together has to see one condition at every position, so the chamber moves its air to keep the humidity even from corner to corner, so a part at the edge takes up the same water as one at the centre. An uneven soak grades the parts by where they sat, never by what they are.

And it has to dry the package before it wets it. The part arrives carrying whatever moisture it picked up since it was made, an unknown the chamber clears with a bake to a dry baseline before the controlled soak begins, so the soak adds back exactly the water the level calls for and no more. From that clean start, the soak is the measured thing the whole grade rests on, the chamber that holds it true making the number mean anything at all.

Common questions

What does a preconditioning chamber do?

It drives a measured amount of moisture into a plastic surface-mount package, holding it at a set temperature and humidity until the package carries the water it would after a rated time in open factory air. A reflow then heats the soaked part. The exercise hunts for the packages that crack or delaminate when the trapped water flashes to steam, the failure called popcorning.

What is popcorning?

Popcorning is a package cracking at reflow. Moisture soaked into the plastic flashes to steam at the solder peak, near 245 to 260 degrees for lead-free, where saturated steam carries a pressure of several megapascals. The pressure builds faster than the mould can relieve it and splits the package, sometimes with an audible pop, or quietly delaminates an interface that fails later in the field.

Why does the soak have to be so exact?

Because the package’s moisture is set by the soak; the grade is set by the moisture. Water enters the epoxy faster than the humidity rises, so a drift of a few percent over a soak of days changes how much the package finally holds. Soak too wet and a good part fails; too dry and a marginal one passes. Only a tightly held, even soak gives a grade that means the same anywhere.

Why did lead-free solder make this worse?

Lead-free alloys melt hotter than the old tin-lead, needing a reflow peak some 25 to 40 degrees higher. That hotter peak makes the trapped steam fiercer and the mould compound weaker at once, so many packages dropped a moisture sensitivity level when they were requalified for lead-free assembly, their floor life shortening to match the harsher heat.

How is the damage found?

By an acoustic microscope, which sends ultrasound into the sealed package and reads the echoes. A delamination opened by the steam reflects the pulse sharply, since the sound meets air, so the scan maps the separation without opening the part. An electrical test and a cross-section confirm a borderline result; a delamination past the standard’s limit fails the part even if it still works.

Part of the Envsin guide to semiconductor reliability testing. A preconditioning chamber loads a plastic package with the moisture a floor life would, so the reflow that follows finds the part that pops, the failure a moisture rating exists to prevent.