Reflow Profiling Methods and Thermocouples

Why the Profile Must Be Measured

The oven setpoints describe what the machine is trying to do; the profile describes what the board actually experiences. A board passing through the same oven at the same settings can see very different temperatures depending on its thickness, its copper loading, and where it sits on the belt. The only way to know is to measure.

Profiling is also the basis for process control. Once a profile is established for a product, it becomes the reference against which changes are judged. A shift in the profile, or a difference between the measured profile and the reference, points to a change in the oven, the board, or the fixture before it shows up as a defect.

The profile is not a single number. It is a curve with a ramp rate, a soak, a peak, and a time above liquidus, and each of those has an effect on the joint and on the components. Reading the whole curve rather than just the peak is what separates useful profiling from a formality.

Thermocouple Attachment

The thermocouple has to measure the temperature of the joint, not the temperature of the air around it. That means attaching it to the board surface, directly on a pad or a joint, so that it is in thermal contact with the material. A thermocouple held in the air or taped loosely to the surface reads the air temperature and produces a profile that is lower and later than the real one.

Attachment methods range from high-temperature tape to a small dot of adhesive. Tape is quick but can lift as the board passes through the oven, and it also insulates the couple slightly from the board. Adhesive gives better contact but takes longer to apply and cure. The method should be consistent, because changing the attachment changes the measurement.

The thermocouple wire itself is a thermal path. A heavy gauge wire conducts heat away from the junction and reads lower than the board, while a very fine wire is fragile. A fine gauge with the wire routed along the board surface, so that it comes to the temperature of the board before reaching the junction, gives the most representative reading.

Choosing Measurement Points

The points should represent the extremes of the board rather than the middle. The corner of a large thermal mass, a small component at the opposite corner, a heavy ground plane area, and a thermally isolated area together show the range the board sees. A profile measured at one point in the centre can hide a difference of tens of degrees at the edges.

The number of points depends on the board and the oven. Three or four channels are enough for a simple assembly, while a large or mixed board may need six or more to capture the variation. The channels should be recorded with their locations so that the data can be interpreted later.

Where a component is thermally sensitive, it should have its own thermocouple, because the requirement for that part may be different from the requirement for the joints. Measuring only the solder joints and assuming the component is within its limit is a common mistake.

Reading the Profile

The ramp rate is the slope from ambient to the soak. A ramp that is too fast can cause thermal shock, spatter from the paste, and uneven heating between components of different mass. A ramp that is too slow wastes time and allows more oxidation.

The soak, or the plateau before the peak, brings the whole assembly to a uniform temperature and activates the flux. Its length and temperature determine whether the flux has done its work before the alloy melts. A soak that is too short leaves unreacted flux and produces residue; one that is too long consumes the flux and causes poor wetting.

Time above liquidus is the period when the alloy is molten and the joint is forming. It has to be long enough for the solder to wet and the joint to form, and short enough to limit intermetallic growth and thermal stress. The peak temperature has to be above the alloy liquidus with enough margin for the coldest point on the board, which is why the margin is set by the coldest channel rather than the hottest.

Profile Verification in Production

A profile is developed on a representative assembly and then used as the reference for production. The verification in production can be simpler than the original development: a periodic check with a profiled board confirms that the oven has not drifted, and the oven’s own sensors can be monitored continuously between checks.

The frequency of verification depends on the product and the oven. A new oven, a recently maintained oven, or a change in the product mix all justify a check, and a stable line with a well-maintained oven can go longer between checks. The point is that the profile is a controlled parameter and should be verified rather than assumed.

The records should include the profile, the date, the oven identification, and the thermocouple locations. When a defect appears later, comparing the current profile with the reference shows whether the oven is responsible in a few minutes rather than after a long investigation.

Common Profiling Mistakes

The most common mistake is measuring air temperature rather than board temperature. The profile looks clean and the joints still fail, because the board was never as hot as the graph suggested. Good attachment and a fine thermocouple gauge are the corrections.

The second is measuring only one point. A single channel in the middle of the board can miss the coldest and hottest areas, and the resulting profile is a compromise that suits neither. Multiple channels at the extremes are needed to know the real range.

The third is using a profile from a similar product. A board with a different thickness, layer count, or copper distribution will heat differently, and the profile has to be developed for the actual assembly. Reusing a profile because the boards look similar is a common source of defects that appear only on one product.

FAQ

Where should thermocouples be attached? Directly on the board surface at the joints that matter, at the coldest and hottest locations. A thermocouple measuring air temperature does not describe the board.

How many thermocouples do I need? Enough to capture the extremes of the board, typically three or more, with additional channels for thermally sensitive components.

What is the most important profile parameter? Time above liquidus and the minimum peak temperature across all channels, because the coldest point determines whether the joint forms.

How often should the profile be verified? After any oven maintenance or change, when the product changes, and periodically on a stable line. The oven’s own sensors should be monitored between checks.

Why does my profile look good but the joints fail? The measurement was probably taken from the air rather than the board, or from a single point that does not represent the coldest area on the assembly.

Conclusion

Reflow profiling turns the oven’s settings into a measurement of what the board actually experiences, and its value depends on attaching the thermocouples to the board, measuring at the extremes, and reading the whole curve. Used as a controlled reference, it makes oven drift visible before it becomes a defect and gives the process a documented basis. For related topics, see our notes on SMT assembly, PCB assembly, quality management, and PCB capabilities for how thermal processes are controlled in 2026.