Soak Zone: 6 Rules in a Reflow Profile

The soak zone is the flat band a reflow profile holds between the initial climb and the spike, usually somewhere between 150 C and 180 C for a lead-free paste. It is the part of the curve where the board stops gaining temperature quickly and simply sits, letting heat spread through the assembly before the alloy melts.

It looks like wasted time, and on a simple board with even copper it often is. On a mixed assembly with connectors, shields and thermal tabs, the soak zone is what keeps a small chip and a large module inside the same process window. Setting it well is a matter of matching the band to the board, not of copying a number.

What the Soak Zone Does Between Ramp and Spike

Three things happen during the thermal soak. Heat conducts from the surface into the inner layers, so the temperature difference across the board narrows. The flux activates and begins to reduce oxide on pads and powder. Volatiles that were driven off during the climb continue to leave the deposit while the paste is still fluid enough to release them.

None of these need the alloy to be molten, which is why the soak zone sits below the melting point of the paste. The zone ends when the board enters the spike and the paste passes liquidus, at which point the window for flux action closes and the joint has to form.

How Thermal Soak Equalises a Mixed Assembly

A board with a thin area and a heavy area heats unevenly, and the difference is largest at the end of the ramp. During the soak zone, the heavy region catches up while the light region holds, because the oven air is close to the board temperature and heat transfer slows as that gap narrows.

The result is that both regions cross liquidus closer together than they would without a soak. This is the reason a reflow profile with a soak zone is more forgiving on mixed technology boards, and why the zone tuning that sets the band matters more than the exact peak.

Setting the Soak Band and Its Duration

The band is set by the paste chemistry stated on the datasheet, and the duration is set by the board. A typical window is 60 to 120 seconds, with the longer times used on thick boards and on assemblies with a large copper imbalance.

Too short and the flux never fully activates, leaving solder balls and poor wetting. Too long and the flux oxidises, the paste dries, and joints form dull and grainy. The band width also matters: a wide band is easier to hold across a panel than a narrow one, and a narrow band demands better oven control.

Flux Activation Inside the Soak Window

Flux activation is a chemical reaction with a rate that rises with temperature, so the soak zone is where most of the cleaning action happens. The activator has to reduce the oxide on the copper pad before the alloy wets it, and it has to do so while the paste is still able to flow into the joint.

Where flux activation is incomplete, the symptom is not always obvious. It can appear as a slightly dull fillet, a small solder ball beside the joint or a marginally higher contact resistance. The peak temperature cannot repair that, because the damage was done before liquidus.

Soak and the Time Above Liquidus Budget

Time above liquidus is the interval when the alloy is molten, and every paste has a limit for it. A soak that runs long does not eat into that budget directly, but it does raise the board temperature at the start of the spike, which shortens the time needed to reach peak.

The two are therefore linked even though they are measured separately. When a profile is adjusted to fix a fill or wetting problem, the reflow profile record should show the soak and the time above liquidus together, because moving one changes how the other behaves.

Component Limits and Moisture Sensitive Parts

Components set a ceiling on the soak as well as on the peak. Plastic bodies, electrolytic capacitors and parts rated to a maximum body temperature all experience the soak in full, and a long soak at the top of the band can push a body past its rating even when the joint is comfortable.

Moisture sensitive devices add a second constraint: absorbed water expands during the climb and the soak, so parts that have exceeded floor life should be baked before assembly. The storage and baking rules that apply are set out in moisture control in SMT storage.

Measuring the Soak on Real Hardware

The soak that matters is the one measured at the joint, not the one printed on the oven display. Thermocouples attached as described for thermocouple attachment give the board reading, and at least one channel should sit on the heaviest joint on the assembly.

Plot the rate of change rather than only the temperature. A curve that is still climbing at 2 C per second when the soak nominally starts is not soaking; it is ramping through the band. The flat part of the measured curve is the true soak, and it is often shorter than the oven setpoint suggests.

Common Soak Mistakes and Their Symptoms

The most common mistake is a soak zone that exists on the setpoints but not on the board, which produces all the symptoms of no soak at all: uneven joints between heavy and light areas, solder balls and incomplete wetting on the trailing edge of large pads.

The second is a soak zone that is too aggressive because the oven was tuned for a different product. The symptom there is flux that has burned off, seen as graping, poor spread and a residue that is difficult to clean. The graping pattern usually follows the pads with the most paste.

Recording the Profile and Reviewing Changes

Keep the measured curve, the soak band and duration, the peak temperature and the time above liquidus in one record tied to the product revision. Add the thermocouple map, because a curve without a map cannot be compared with the next measurement.

Recheck whenever the paste, the board, the oven or the component mix changes. Where a board has been through thermal excursions during fabrication, the notes in thermal deformation help explain a shift that the profile alone does not account for, and the IPC profile standards give the limits to check against.

FAQ

Is a soak zone always necessary? No. Simple boards with even copper and small components often run a straight ramp to peak, which is faster and gives a brighter joint on some pastes. The soak earns its place when the assembly has a wide spread of thermal mass.

What soak time should be used for a lead-free paste? Most datasheets quote a band and a range rather than a single value, commonly 60 to 120 seconds. Start near the middle of the range, measure at the heaviest joint and adjust within the range rather than outside it.

Does a longer soak reduce voiding? It usually does, because more of the volatile material leaves before the alloy melts. The gain flattens once the flux has given up its solvent, and adding time past that point only oxidises the paste, so measure the void result rather than assuming more is better.