Cold Solder Joints: Causes and Detection

A cold solder joint looks like a solder joint and behaves like a resistor. It forms when the alloy solidifies without having reached a proper wetting temperature, so the metal never bonds metallurgically to the pad and the lead. The joint passes a quick visual check, survives handling, and then opens intermittently in the field under thermal cycling or vibration. This article explains what makes a joint cold, how insufficient heat prevents the intermetallic layer from forming, how to detect the defect, and how to remove the causes.

What Makes a Joint Cold

Soldering is a metallurgical process, not a gluing process. For a reliable joint the alloy must melt, wet the surfaces it touches and then hold above the liquidus long enough for a thin intermetallic layer to grow at each interface. A cold solder joint is one where that sequence was cut short, usually because the joint never reached the wetting temperature or because it cooled while still moving.

The result is a fillet with a dull, grainy, often slightly convex surface instead of the smooth concave shape that indicates wetting. The alloy is mechanically attached rather than bonded, so the contact area is small and the electrical path is fragile. Because the joint can still conduct when it is new, the defect escapes functional test.

Insufficient Heat and the Intermetallic Layer

The intermetallic compound that forms between copper and tin is what makes the joint electrically and mechanically sound. It is only a few micrometres thick and it grows at a rate that depends strongly on temperature and time. If insufficient heat reaches the interface, the layer is incomplete or absent, and the joint relies on whatever adhesion the solidified alloy has to a surface it never properly wet.

Heat can be missing for several reasons that have nothing to do with the setpoint. A large copper plane conducts energy away faster than the source can supply it, a thermal relief that is too narrow limits the flow, and a component with a heavy lead acts as its own heat sink. The relevant quantity is the temperature at the interface, which is always lower than the temperature of the tool or the oven air.

Disturbed Joints During Cooling

A joint that is mechanically disturbed while the alloy is between liquidus and solidus develops the same grainy appearance as a cold joint, and the two are often confused. Movement can come from a conveyor that vibrates, a board that warps as it cools, a component that shifts because its terminations are not symmetrical, or a technician who moves a hand soldered part too early.

The distinction matters for diagnosis. A disturbed joint indicates a mechanical problem in the process, such as fixture rigidity or conveyor condition, while a genuinely cold joint points to a thermal problem. Treating one as the other wastes time, because the fixes are in different parts of the line.

Detection by Visual and X-Ray Joint Inspection

Visual joint inspection remains the first filter. A proper fillet has a bright, smooth, concave profile with a small contact angle, and any joint that is dull, rounded or shows a fracture line at the interface deserves a closer look. Optical inspection works well on through hole and perimeter joints, but it cannot see under a ball grid array or under a component body.

X-ray fills that gap by showing the shape of hidden joints. A cold or disturbed joint typically appears with irregular boundaries, voids concentrated at the interface, or a fillet that does not extend around the pad as it should. Our overview of automated optical inspection covers how AOI fits alongside X-ray and where each technique is blind.

Electrical Symptoms in the Field

Because a cold joint often conducts when it is cold and fails when it is warm, the symptoms are intermittent and temperature dependent. A unit that works on the bench and fails after a few minutes of operation, or that recovers after being cooled, is showing the classic signature. Vibration produces a similar pattern, with faults that appear only when the product is running.

The long term outcome is an open circuit or a high resistance connection that heats locally, which can then damage the surrounding laminate. Our collection of solder defect and board failure case patterns includes several of these progressions, and most of them begin with a joint that passed inspection.

Reflow Profile and Thermal Design Causes

The reflow profile determines whether every joint on the board reaches the required temperature for long enough. A profile that was developed on a bare coupon will be too aggressive or too weak on a populated board, because component mass changes the thermal response. Verifying the profile with thermocouples attached to real joints, as described in reflow oven profile verification, is the only reliable way to know.

Design contributes as much as the oven. Pads connected to planes without adequate thermal reliefs, large copper areas under small components and asymmetrical pad geometries all create joints that heat at different rates. Where a board has a wide spread of thermal mass, the profile has to satisfy the coolest joint, and the hottest one must still stay within the component limits.

Hand Soldering and Rework Causes

Hand soldering produces cold joints when the iron cannot deliver enough energy. A tip that is too small for the joint, a station with poor thermal recovery, or an operator who touches the joint briefly and withdraws will all produce the defect. Adding more solder does not help, because the problem is temperature at the interface, not the volume of alloy.

Rework has an additional risk: the surrounding area has already been heated, and the operator may be working quickly to protect nearby parts. The result is a joint that looks complete but was never properly wetted. Rework instructions should specify the tip, the setpoint and the expected dwell time so that speed does not override the thermal requirement.

Prevention Rules

Prevention comes down to controlling interface temperature and eliminating movement. Profile every board family with thermocouples on the critical joints, keep thermal reliefs on plane connected pads, match tip geometry and station capability to the joint, and let joints cool undisturbed for the full time the alloy needs. Then confirm with inspection that the result is consistent.

Statistical confirmation is worth the effort. Tracking joint related defects against profile revisions and station settings shows whether a change actually helped, and it turns a recurring field problem into a process parameter that can be managed. The gopcb engineering team applies this approach when a customer reports an intermittent failure that cannot be reproduced on the bench.

FAQ

Can a cold solder joint be repaired? Yes, if the residual flux and alloy are removed and the joint is remade with adequate heat. Simply adding solder over the top usually leaves the original unwetted interface in place and the fault returns.

Is a dull joint always cold? No. Some lead-free alloys are naturally less shiny than leaded ones, and a disturbed joint looks similar. The decisive evidence is the fillet shape and the interface, not the surface gloss alone.

Why do cold joints pass functional test? Because a cold joint still conducts while it is new and at room temperature. It only becomes an open circuit when thermal expansion, vibration or oxide growth breaks the weak contact, which may happen long after the product has shipped.