Reflow Soldering: What Each Zone Must Hit
Home
Yiwu, Zhejiang, China
Mon – Fri 9:00 AM – 6:00 PM · Sat 9:00 AM – 4:00 PM
[email protected]

What Is Reflow Soldering: Four Zones And The Numbers In Them

Alan Xiao, Founder of Sourcing Pioneer
Alan Xiao Founder of Sourcing Pioneer
· Updated November 10, 2026 · 10 min read

Reflow soldering melts solder paste in an oven, not under an iron.

Paste is printed through a stencil onto the pads. Components are placed on top of the sticky paste. The whole board then travels through a heated tunnel and comes out with permanent joints.

The process is controlled by a temperature curve called a profile. That curve has four sections, and each one has numbers attached.

Most explanations name the four sections and stop. This page gives the numbers, then shows which defect each wrong number leaves behind.

Studio still life of a stencil, a solder paste syringe and a small circuit board on beige
Paste, stencil, oven, curve.

The Short Answer

Reflow soldering is the standard way to attach surface mount parts. Solder paste holds the part in place, and heat turns the paste into a joint.

The paste is a mix of tiny solder powder and flux. The flux cleans the metal and lowers surface tension so the molten alloy wets the pad and the part.

Heating usually happens in a long convection oven with heated zones. Infrared lamps and a hot air pencil do the same job for repair and prototypes.

What makes or breaks the process is the temperature curve, not the oven brand. A correct curve on a modest oven beats a random curve on a good one.

Paste Is Solder And Flux, Not Wire

Wire soldering and reflow use the same alloy in different forms. The difference is how much of the job the flux has to do.

In paste, the flux is measured into the mixture. It must survive printing, sit on the board while parts are placed, then activate at the right moment in the oven.

That timing is why the profile matters. Flux that activates too early is used up before the alloy melts, and the joint does not wet.

It also explains the smell of a reflow line and the residue on a finished board. Both come from the flux doing its job and then being left behind.

The Print Decides Half The Outcome

Before any heat arrives, a stencil decides how much paste lands on each pad. That volume then decides how the joint forms.

A stencil is a thin metal sheet with apertures cut where the pads are. A squeegee drags paste across it, and the paste stays in the openings when the sheet lifts.

Too little paste gives a weak joint with a poor fillet. Too much gives a part that floats and shifts, or balls of solder squeezed out at the side.

The awkward part is that both failures look like oven problems afterwards. A joint that did not wet may simply have had half the paste it needed.

So when a defect appears, look at the print before the profile. Aperture size, stencil thickness and squeegee pressure sit upstream of every zone.

Four Zones, One Purpose

The four zones are not four separate steps for their own sake. Each one prepares the board for the next, and skipping the preparation shows up as a defect.

ZoneWhat it doesThe number attached
PreheatRaises the whole board steadily, drives off solventsSlope, up to about 3.0 °C/s
SoakActivates the flux and evens out the boardUsually 60 to 120 seconds
ReflowMelts the alloy and forms the jointPeak and time above liquidus
CoolingFreezes the joint with a fine grainAbout 4 °C/s suggested

Preheat has to be linear, not a jump. Too fast a rise boils the solvents, and the escaping gas throws solder off the pads.

That throwing is called solder balling, and it is the classic sign that the first zone was rushed. Small balls end up on the mask, sometimes bridging two pads later.

Soak is where the board catches up with itself. Big parts and small parts heat at different rates, and the soak is the chance to close that gap before the alloy melts.

The Numbers That Actually Matter

Here are the figures a profile is judged against. They come from paste data and from the process literature, not from an oven brochure.

ParameterUsual valueWhat goes wrong outside it
Ramp rateUp to about 3.0 °C/sFaster: solvent boils, solder balls
Soak time60 to 120 secondsShort: flux not active, uneven board
Peak temperature20 to 40 °C above liquidusLow: no wetting. High: damage
Time above liquidusAt least 30 s, wetting under 60 sShort: cold joint. Long: brittle joint
Cooling rateAbout 4 °C/s suggestedToo slow: coarse grain, weaker joint

Two of those rows deserve a note. The peak is set by the part that tolerates heat least. The usual practice is to take that part’s limit and subtract 5 °C.

The 30 second minimum for time above liquidus is worth treating carefully. The literature notes that most pastes specify it, and also that there is no clear reason for that exact figure.

One likely explanation is measurement. A profiler reads a handful of points, so a generous minimum keeps an unmeasured area from being the one that never melted.

The Profile Follows The Board

A profile is not a setting copied from a paste datasheet. It is measured on your board, with thermocouples attached at points that matter.

What decides its shape is the copper: how much ground plane there is, how thick it is, and how many layers it hides in.

A board with a large ground plane acts as a heat sink. The surface heats quickly while the plane lags, and that lag is where uneven joints come from.

The mix of parts matters for the same reason. A row of small resistors next to a big connector gives two very different thermal paths on one board.

So the soak exists to let the slow parts catch up. A bigger gap between the fastest and slowest point means a longer soak, not a hotter peak.

This is why a factory that quotes a profile from a similar job is guessing. Ask for the profile measured on your panel, with the points marked.

What The Oven Actually Controls

The oven controls two things: the temperature in each heated zone and how long the board spends in it. Belt speed sets the second one.

Most production ovens move heat by forced convection, blowing hot air onto the board. Older infrared types heat by radiation, which is less even across a board with mixed colours and metals.

The number of zones matters less than the control. Six zones with a measured curve beat ten zones set by guesswork.

The reason is drift. Heaters age, filters clog, and a curve that was right in March can be four degrees off in September without anyone touching a setting.

That is why the profile is treated as a measurement rather than a recipe. It gets repeated on a schedule, logged, and compared, which is what the process literature means by statistical control.

Why 260 °C Is A Line

Above 260 °C things inside a package start to suffer. The die and its connections are not built for it, and the metal layers keep growing compounds they should not.

Lead-free alloys raise the pressure. A tin-silver-copper alloy melts at 217 to 220 °C, which is about 34 °C above the old tin-lead eutectic.

That pushes the working peak to roughly 235 to 245 °C for wetting and capillary flow. The gap between that and 260 °C is the room a profile has.

Some parts eat that room. Electrolytic capacitors, connectors, optoelectronics and older plastic bodies are the usual offenders, which is why 260 °C rated versions exist.

The common alloy is SAC305, at 3.0 percent silver and 0.5 percent copper. Cheaper mixes with less silver melt a little higher, which narrows the room again.

Diagram of a four zone reflow profile with soak, peak and cooling marked
Four zones, and the three temperatures that decide them.

What A Wrong Profile Leaves Behind

The useful thing about reflow defects is that most of them point back at one zone. Read the failure, then read the curve.

What you seeWhich zone caused itWhat went wrong
Small solder balls on the boardPreheatRamp too fast, solvents boiled and spat
Dull or cold looking jointsReflowToo little time above liquidus
Voids, trapped fluxReflowToo little time, flux not fully out
Brittle joints, weak in testReflow and coolingToo long above liquidus, thick intermetallics
One part not wetted at allSoakBoard never reached balance; a cold spot
Part damaged insideReflowPeak over the part’s limit, above 260 °C

Two rows are easy to confuse. A cold joint and a void look alike. One is too little heat; the other is heat applied for too little time.

The fix differs. Raising the peak does nothing for a board that needed another twenty seconds in the soak.

Brittleness is the opposite failure. A joint kept liquid too long grows a thick layer of intermetallic compound, and that layer cracks under thermal cycling.

Mixed Boards And Through-Hole Parts

Many boards carry both surface mount and through-hole parts. There are two ways to solder the second group.

The classic way is a wave solder pass after reflow. The alternative fills the through-holes with paste and lets the reflow oven do both, which the trade calls intrusive reflow.

Saving a process step sounds like free money. On lead-free alloys it is harder. The paste has to push into a hole and wet a lead while the rest of the board sits at peak.

Some through-hole parts also carry plastics that cannot take the profile, and those still need a hand iron afterwards. That is the step the saving was supposed to remove.

So ask which of the two routes the factory will use, and which parts still need hand work. The answer changes both the cost and the places where a defect can hide.

Flow diagram comparing reflow plus wave soldering with intrusive reflow
Two routes for the through-hole half.

What To Ask The Factory

Four questions separate a real process from a copied one.

First. Show the profile measured on this panel, with the thermocouple points marked on a drawing.

Second. Name the peak you will run, and the part on the board that set that limit.

Third. Give the soak time and the time above liquidus, with the band you keep them inside.

Fourth. Show how the oven is checked between runs. A profile is a measurement, and a measurement drifts.

A factory that answers all four has a process. One that sends a paste datasheet instead has a hope.

Two more checks cost nothing. Ask for a sample board cut open, so the joints under a package are looked at rather than assumed.

And ask what happens when a profile check fails. If the answer is that production carries on until the next shift, the schedule exists on paper only.

Keep the records with the batch. A profile log with a date and an operator name is what turns a claim into a conversation.

Rework Is A Different Process

Repairing one joint does not use the tunnel. A hot air pencil or a small bench oven heats the part locally, and the rest of the board stays near room temperature.

That local heat is harder to control. The part sees the peak while its neighbours do not, and a clumsy pass can lift a pad or cook the plastic around it.

So rework has its own profile, usually shorter and cooler, and its own risk. It also tends to void the warranty of the assembly step. That matters when a supplier offers to fix a batch by hand.

Six Lines for the Order

Put these in the assembly brief and the questions stop being arguments after the fact.

One. Alloy and paste, with the liquidus temperature written down.

Two. Peak temperature and the part that sets it.

Three. Soak window and time above liquidus, as a range rather than one number.

Four. The cooling rate, because it is the row everyone forgets.

Five. Whether through-hole parts are reflowed or waved, and which ones get a hand iron.

Six. The inspection standard, so a joint is judged against a written rule. The standard a solder joint is judged against is where that rule comes from.

Two more reads help. Where a small board run gets built covers the step before volume. And checking boards one by one is what catches a drift the profile no longer shows.

The line list that comes back comparable is the format that gets all six lines back in one document.

Checklist panel of six lines to put on a reflow assembly order
Six lines that turn a profile into a commitment.
PROFILE CHECK

Send me the assembly quote or the profile page from the factory. I will check whether the peak names the part that limits it, whether the soak and time above liquidus are given as windows, and whether the cooling rate appears at all.

Check The Profile

Questions Buyers Ask

What temperature does reflow soldering reach?

Peak is usually 20 to 40 °C above the alloy’s liquidus point. For a lead-free SAC alloy that means roughly 235 to 245 °C, with 260 °C treated as a limit for the parts.

Why four zones instead of one ramp?

Each zone does a different job. Preheat drives off solvents slowly, soak activates the flux and evens the board, reflow melts the alloy, and cooling sets the grain structure.

What causes solder balls on the board?

Usually a preheat ramp that is too steep. The solvents boil and the escaping gas throws solder off the pads before the alloy melts.

How long should the board stay above liquidus?

Most pastes ask for at least 30 seconds, with wetting ideally finished inside 60. Too short leaves cold joints and voids, too long makes the joint brittle.

Can reflow also solder through-hole parts?

Yes, by filling the holes with paste. It is harder with lead-free alloys, and parts with plastic bodies may still need a hand iron afterwards.

Does the oven brand matter?

Less than the curve. A measured profile on a modest convection oven beats a copied profile on an expensive one, because the board decides the curve.

Questions & Comments

We read every one and reply within 24 hours

If your boards come back with tiny solder balls on the mask, tell me what the first zone was doing. That is usually the whole story.

Subscribe
Notify of
guest
0 Comments
Most Voted
Newest Oldest
0
Would love your thoughts, please comment.x
()
x