Cold Joint: 5 Rules for Complete Solder Melting

A cold joint is a solder connection that solidified before the alloy and the surfaces reached the temperature needed to form a proper bond. It is recognised by a dull, grainy surface that contrasts with the bright, smooth fillet of a good joint. The shape is often wrong as well, with a convex blob instead of a concave fillet.

The cause is always a shortage of heat at the joint, but the heat can be missing for several reasons. The reflow profile may peak too low, the thermal mass of the copper may pull heat away faster than the oven delivers it, the flux activity may have been exhausted before the alloy melted, or the solder temperature at a hand iron may have been lost on contact.

What a Cold Joint Looks Like

A cold joint has a rough, matte surface because the alloy solidified as a mass of small grains rather than as a continuous cast structure. The grain boundaries scatter light, which is why the joint looks dull. A good joint, cooled from a fully molten state, has a smooth surface that reflects light in a single direction.

The profile of the fillet carries more information than the colour. A proper fillet curves smoothly from the pad to the pin, while a cold joint often forms a blob that sits on the pad without rising. Where the alloy never fully melted, the paste residue and the partly fused powder are still visible inside the joint under magnification.

Peak Temperature and Time Above Liquidus

The reflow profile has to take the joint above the liquidus of the alloy and hold it there long enough for the surfaces to wet. For a lead free alloy, that means a peak roughly twenty to thirty degrees above the liquidus, and a time above liquidus measured in tens of seconds rather than a few.

Measuring the peak at the board edge says nothing about the joint. The thermocouple has to be attached to a representative joint on the product, and the profile has to be taken with the panel in its production configuration. The thermocouple attachment method decides whether the number means anything.

Thermal Mass and Heat Sinking

Copper conducts heat away from the joint as fast as any oven can deliver it. A pad connected to a plane, a large connector body and a thick ground layer all act as heat sinks, and the joint on the far side of them is always the last to reach temperature. That is why the coldest joint on a board is usually at the same reference on every panel.

Thermal relief spokes are the design answer, and their width and count decide how much heat the plane pulls away. Where the design cannot change, the profile has to be adjusted for the heaviest joint on the board, which means the light joints run hotter than they would otherwise need to.

Soak Zone and Flux Activation

The soak zone brings the whole assembly to a uniform temperature before the peak, and it is where the flux becomes active and begins to remove oxide. A soak that is too short leaves the heavy areas behind, so the peak arrives while parts of the board are still climbing. The soak zone is therefore a heat equalising step rather than a time filler.

A soak that is too long exhausts the flux activity before the alloy melts, and the joint then forms on an oxidised surface. That failure looks like a cold joint but has a different cause, and the fix is to shorten the soak or to choose a paste with a longer activation window rather than to raise the peak.

Ramp Rate and Preheat Control

The ramp into the soak determines how evenly the board heats. A fast ramp raises the surface temperature quickly while the interior lags, and the difference is greatest on thick boards and on assemblies with a large ground plane. Slower ramps cost cycle time but reduce the temperature spread across the panel.

Ramp rate also affects the flux and the paste. A ramp that is too fast can spatter the flux and leave voids, while one that is too slow allows the paste to slump. The preheat ramp rate should be set from a profile measured on the product, not from a generic oven recipe.

Hand Soldering and Iron Temperature

A hand iron cold joint comes from contact time rather than from the set temperature. The tip loses heat the moment it touches a pad or a plane, and a small tip on a heavy joint cannot recover fast enough. The result is a joint that never reaches soldering temperature even though the iron display reads several hundred degrees.

The practical controls are the largest tip that fits the joint, the shortest contact that still melts the alloy, and a tip that is properly tinned and free of oxide. Where a joint needs a long contact time, the correct answer is more thermal capacity rather than a higher setting, because raising the temperature damages the board without solving the heat shortage.

Wave and Selective Soldering Cold Joints

In wave soldering, a cold joint usually means insufficient preheat or too short a contact with the wave. The board arrives at the wave already cool, the alloy gives up its heat to bring the joint up to temperature, and the joint leaves the wave before it has fully melted. Conveyor speed and preheat both control that balance.

Selective soldering has the same problem on a smaller scale. The alloy temperature in the pot may be correct while the temperature at the nozzle tip is lower, and a fountain that is set too low loses heat before the joint is reached. The solder fountain height and dwell are the two settings to check first.

Telling Cold Joints From Other Defects

A disturbed joint looks similar but happens after melting. The alloy was fully liquid, then the joint moved before it solidified, and the surface shows ripples or a torn edge rather than the uniform grain of a cold joint. The distinction matters because the causes are different, and a reflow profile change will not fix a handling problem.

Poor wetting is the third case. The alloy melted completely, but the pad or the pin was contaminated, so the solder beaded instead of spreading. The surface can look dull, yet the fillet angle tells the story. A wetting failure has a high contact angle, while a cold joint has the right shape and the wrong surface.

Verification, Cross Sections and Records

Verifying the process means profiling the product rather than the oven. A periodic profile on the heaviest assembly on the line, recorded against the oven settings, shows whether the process has drifted. Between profiles, the joint surface on the first article is the fastest indicator that something has changed.

Where the cause is not obvious, a cross section shows whether the alloy melted fully and how much intermetallic formed at the interface. The section should be taken from the joint that runs coldest, not from a convenient one, and it should be kept with the profile data so the two can be compared.

FAQ

Is a cold joint always caused by the reflow oven? No. The oven is one source of heat among several, and a cold joint can come from a hand iron that lost temperature, a wave that the board crossed too quickly, or a selective nozzle that stands too low. The first question is which process made the joint, not which setting to change.

Can a cold joint be reflowed again to fix it? Reflowing the joint can complete the bond where the alloy and the surfaces are still clean, and it is a common rework for a single joint. Where the joint also failed to wet because of contamination, a second pass will not help, and the joint has to be cleaned and rebuilt.

Why is a cold joint a reliability risk if it still conducts? The grainy structure has less strength and lower fatigue resistance than a properly formed joint, and the bond to the pad is incomplete. Such a joint can pass test and still fail after thermal cycling, which is why the defect is treated as a reliability problem rather than a cosmetic one.