Thermal Profiling of Large Assemblies

A large assembly does not heat evenly, and no single thermocouple can describe it. The profile that governs the result is the one at the coldest joint and at the hottest joint, and the process has to satisfy both at the same time.

Profiling a large board is therefore an exercise in finding the extremes and then deciding which one sets the process. On most assemblies the answer is the centre of a large plane for the cold extreme and a small part near the edge for the hot one.

Thermal Mass and Its Distribution

Thermal mass is the product of the material and its heat capacity, and copper has far more of both per unit volume than laminate. A ground plane that covers a third of the board changes the heating behaviour of everything attached to it.

The distribution matters more than the total, because a small isolated area heats quickly while the same copper connected to a plane heats slowly. The connection, not the area, is what the profile has to account for.

Probe Placement Strategy

The probes should be placed at the joints that are hardest to heat and at those that are easiest, with one probe on the board surface as a reference. The difference between the two extremes is the process window that remains.

Attachment is a practical problem on a large board, since tape fails and wire pulls. Adhesive cured under the probe and a routed wire that does not press on the surface both help, and a dedicated coupon can remove the problem entirely.

Soak and Differential Heating

The soak exists to let the assembly reach a more even temperature before the alloy melts, which is exactly what a large board needs. A longer soak reduces the difference between the cold and hot areas and widens the usable peak window.

The limit is the flux, which has a finite life at soak temperature. Extending the soak beyond the flux capability produces poor wetting regardless of how even the board has become.

Conveyor Speed and Dwell

Speed is the simplest way to change the dwell, and it changes the whole curve at once. For a large board the speed is usually reduced so that the mass can absorb enough energy in each zone.

Reducing speed also increases the time above liquidus, which affects the intermetallic growth. The two effects have to be balanced, which is why the curve is assessed as a whole rather than by adjusting one number.

Support and Warpage

A large board can sag under its own weight at reflow temperature, and the sag changes both the placement and the thermal contact. Centre support, a pallet or a mesh conveyor all reduce it, and the choice affects the profile as well.

Warpage also changes the contact between the board and any tooling, so a board that was supported at the edges behaves differently from one that sits on a flat belt.

Differential Between Components

A large package and a small chip part on the same board have very different heating rates, and the difference can be large enough that one is at peak while the other is still in soak. The placement of heavy components relative to the plane is therefore a thermal design decision.

Where the difference cannot be resolved, a second profile with a different speed may be needed for a second pass, and the two passes documented separately.

Oven Capability

A large thermal load requires an oven with enough heating capacity to hold its zones at temperature, and an oven that is adequate for small boards may not be. The measurement is the zone recovery under load rather than the setpoint.

Where the oven struggles, the symptom is a profile that changes with the number of boards in the tunnel. Running a single board for profiling therefore flatters the process in a way that production does not.

Verification and Records

The profile should be run with a representative panel and repeated when the product mix on the line changes significantly. A profile taken on an empty tunnel is a starting point and not a qualification.

The records should include the extremes, the reference probe and the panel loading, so that a later comparison is meaningful. These records support the same process evidence described for manufacturing processes.

Practical Sequence

The practical order is to profile with a full panel, find the cold and hot extremes, adjust the soak and the speed, and then confirm with a cross section on the hardest joint. Only then is the profile released.

A large assembly that has been profiled in that order produces a process that is robust to a normal variation in loading, which is the property that matters in production.

Process Control and Verification

Running a first article through the same checks as the production panel confirms that the two agree, and that comparison is the cheapest form of process control available at prototype stage. Keeping a sample from the panel turns a dispute into a measurement, because the same coupon can be re-examined by both parties without rebuilding the batch.

Reviewing the design before the data is released is cheaper than correcting it after the panel is in the tank, because every step downstream inherits the decision made at the front end.

Related reading: our fabrication notes, board quality and design release notes cover the same ground.

FAQ

How many probes does a large board need? Enough to find the extremes, which in practice means three to six including a reference on the surface.

Should the profile be run with a full tunnel? It should be run with a representative loading, since an empty tunnel gives a different profile on a large board.

Is a longer soak always better? Only up to the point where the flux is exhausted, after which wetting deteriorates.

Can a large board be run at the same speed as a small one? It can if the oven has the capacity, and the profile has to be demonstrated rather than assumed.