A cold solder joint is a defective solder connection in which proper wetting and bonding between the solder and the surfaces being joined have not been achieved. It may look rough, irregular, grainy or poorly bonded to the pad or component lead. In repair work, especially in British terminology, similar defects are also commonly described as dry joints. However, not every dull, cracked or unreliable solder connection is technically the same type of defect.
Appearance alone is not enough for a reliable diagnosis. Lead-free solder joints can naturally look duller and less smooth than older tin-lead joints while still being completely acceptable. A suspicious joint becomes much more significant when its appearance matches the behaviour of the circuit: intermittent resets, signal loss when a connector is moved, faults that appear after warm-up, sensitivity to vibration or operation that changes when the PCB is gently flexed.
What does a cold solder joint look like?
Cold solder joints are easiest to inspect on through-hole components, connectors, larger surface-mount parts and high-current connections. A properly formed solder joint should show good wetting of the surfaces being joined and a geometry appropriate to the component and soldering process. The solder should not simply sit on the pad like a separate blob. It should form a genuine metallurgical connection with the land, lead or termination.
A defective joint may appear unusually rough, grainy, irregular or porous. The solder may form a rounded blob with a clear boundary where it failed to wet the pad or component lead. In other cases the problem is visible as an incomplete fillet, exposed surface that should have been wetted or an obvious separation at the interface.
Cracks can also be visible, particularly around through-hole leads that have been exposed to mechanical loading or repeated temperature changes. A circumferential or ring-shaped crack around a component lead is a common clue that the electrical connection may have become intermittent. A moving terminal, lifted land or damaged plated-through hole can produce very similar symptoms.
The important point is that a dull solder joint is not automatically a cold solder joint. Lead-free solder frequently has a more matte, grey or grainy appearance than traditional tin-lead solder. The more useful indicators are poor wetting, abnormal fillet geometry, visible separation, cracking and a clear relationship between the suspect connection and the actual fault in the equipment.
Cold joint, dry joint, disturbed joint or cracked solder?
Several different solder defects are often grouped together under the informal label “cold solder joint”. For practical troubleshooting this can be misleading because they may have different causes.
A cold solder joint is primarily associated with inadequate heating, poor solderability or contamination that prevents proper wetting and formation of the solder connection.
A dry joint is a widely used repair and servicing term, especially in the UK. Depending on context, it may describe poor wetting, a mechanically weak solder connection or a joint that was disturbed before the solder had fully solidified. It is useful practical terminology, but it is not always applied as precisely as the defect classifications used in electronics manufacturing standards.
A disturbed solder joint forms when the component lead, termination or other part of the connection moves while the solder is solidifying. The surface may look wrinkled, uneven or fractured because the developing structure of the solder was mechanically disrupted during cooling.
A cracked or fractured solder joint can develop after the original soldering process. A joint that was initially acceptable may crack after repeated thermal cycling, vibration or mechanical stress. Such a service-life failure should not automatically be described as a cold joint.
It is also useful to distinguish nonwetting and dewetting. With nonwetting, molten solder contacts a surface but fails to form the required bond over it. Dewetting occurs when solder initially wets a surface and then pulls back, leaving irregular areas with insufficient solder coverage. Both can create unreliable connections, but the underlying process is not necessarily identical to a classic cold joint.
What causes a cold solder joint?
The most common cause is insufficient heat transfer to the complete connection. It is not enough for the solder itself to melt. The pad, component lead or terminal must reach the conditions required for the solder to wet the surfaces properly and form a stable intermetallic connection.
A soldering iron set to an apparently correct temperature can still produce a poor joint if the tip is too small, the contact area is poor or the connection has a high thermal mass. Large copper areas, ground planes, connectors, power terminals and heavy component leads can remove heat from the joint much faster than a small signal pad.
Other causes include oxidised component leads or PCB finishes, contamination, poor solderability, insufficient flux activity, exhausted flux and an unsuitable heating profile. In automated assembly, an incorrect reflow profile can produce defects even when the solder paste and components themselves are correct.
Movement during cooling is a separate but related problem. If a wire, lead or component shifts while the solder is passing through solidification, the resulting connection may look rough or wrinkled and may have reduced mechanical integrity. In repair work this is often called a dry or cold joint, although technically it is better described as a disturbed solder joint.
What symptoms can a bad solder joint cause?
Intermittent solder faults are often much harder to diagnose than a permanent open circuit. A damaged connection may conduct normally while the equipment is cold and stationary, then fail when the board heats up, vibrates or experiences mechanical stress.
| Symptom | Possible solder or interconnect problem | What to inspect |
|---|---|---|
| Random controller or module resets | Intermittent supply connection at a connector, regulator, relay or high-current component | Power connectors, regulator pins, relay terminals, power resistors and visibly heat-stressed joints |
| Signal drops when a cable or PCB is moved | Cracked solder joint, damaged connector, lifted pad or broken trace | Connector pins, ring cracks around through-hole leads, pad movement and nearby copper traces |
| Fault appears only after warm-up | Thermal expansion opens a crack or changes contact resistance | Joints around hot components, power stages, drivers, regulators and mechanically constrained parts |
| Equipment works when the board is pressed or flexed | Cracked solder, broken PCB trace, damaged via, lifted pad or package-related interconnect fault | Solder joints, laminate, vias, traces and mechanically stressed areas |
| Continuity test passes but the fault remains | Intermittent crack is making contact at the moment of measurement | Joint geometry under magnification and behaviour when the original operating condition is reproduced safely |
Connections carrying substantial current, heat or mechanical load deserve particular attention. Board-mounted connectors, transformers, large capacitors, relays, power resistors, power semiconductors and soldered wires can all place more stress on the joint than small signal components.
Why temperature and vibration reveal solder faults
Temperature changes are particularly effective at exposing marginal interconnections because the PCB, component package, leads and solder do not all expand by exactly the same amount. Repeated heating and cooling can therefore create cyclic mechanical strain in solder joints.
A solder connection that was originally manufactured correctly can eventually develop a fatigue crack after a large number of thermal cycles. The crack may remain electrically closed at one temperature and open at another. This explains why a board can pass a simple resistance or continuity test on the bench and still fail after the equipment warms up.
Vibration can produce similar behaviour, especially where a heavy component or connector transfers mechanical forces into its solder joints. A marginal connection may open only for a fraction of a second, creating resets, communication errors or signal dropouts rather than an obvious permanent failure.
This distinction matters diagnostically. A fault discovered after many years of service is not necessarily evidence that the original joint was “cold”. It may instead be a fatigue failure of a joint that was initially acceptable.
Why a continuity test does not prove the solder joint is good
A continuity tester answers a very narrow question: is there an electrical path at the moment of the measurement? It does not determine whether the solder joint is mechanically intact or whether it will remain conductive under temperature, vibration, current load or PCB movement.
A cracked joint can temporarily close as materials contract or when the board is lying in a particular position. The meter may beep normally even though the same connection opens when the board warms or a connector is touched.
For this reason, electrical measurement should be combined with visual inspection and the actual symptom. Where it can be done safely, diagnostic testing is more useful when the operating condition that triggers the failure can be reproduced in a controlled way.
Cold solder joint or damaged PCB?
Not every intermittent electrical connection is a solder defect. A cracked copper trace, damaged plated-through hole, lifted land, worn connector, fractured SMD component or an internal component fault can produce almost identical symptoms.
With a through-hole component, the problem is sometimes relatively obvious. A circular crack may be visible around the lead, the solder may not wet the land correctly or the pin may move relative to the PCB. If the copper track has cracked next to the pad, however, resoldering the visible joint may make no difference because the actual break is further along the circuit.
A damaged via can be even more deceptive. The solder visible on the surface may look acceptable while the electrical discontinuity exists inside the plated hole or at an internal layer connection. Multilayer boards therefore require more caution than a simple single-sided PCB.
Always relate the suspicious connection to its electrical function. If the equipment loses power and the suspect joint is on an input connector, regulator or power relay, the evidence is consistent. If resoldering that point changes nothing, the diagnosis must be reconsidered rather than repeatedly adding more solder.
Through-hole, SMD and BGA solder faults are diagnosed differently
Through-hole connections
Through-hole joints are usually the easiest to inspect visually. Useful features include wetting of both the component lead and the PCB land, solder fillet shape, visible cracks and movement of the lead. For plated-through holes, proper hole fill and the integrity of the barrel are also important.
A joint can look acceptable from one side while a problem exists deeper inside the plated hole. This becomes particularly relevant for thermally demanding connections connected to large copper planes.
Surface-mount components
With surface-mount devices, part of the connection may be hidden under the component termination. Visible fillet shape can still provide useful information, but it does not always reveal the full condition of the interconnect.
Fine-pitch assembly can also produce defects that resemble a cold solder joint. One example is graping, where solder-paste particles fail to fully coalesce during reflow and leave a grainy appearance. This is a process defect, but it should not automatically be described as a classic cold joint simply because the surface looks rough.
BGA and other hidden joints
Ball grid array packages create an entirely different inspection problem because the solder joints are located underneath the component. A magnifying glass or service microscope can inspect the surrounding PCB, but it cannot confirm the condition of most BGA solder balls.
Professional inspection of hidden interconnections may require X-ray imaging, and more advanced failure analysis can use computed tomography, cross-sectioning or other methods. Symptoms such as sensitivity to temperature or board flex may suggest an interconnect problem under the package, but they do not prove that the fault is a cold solder joint.
A short diagnostic checklist
- Disconnect power and inspect the suspect area under good lighting and suitable magnification before attempting rework.
- Compare the joint with neighbouring connections produced by the same soldering process instead of judging surface shine in isolation.
- Look for poor wetting, abnormal fillet shape, visible cracks, movement of pins, lifted pads and damage to nearby traces or vias.
- Determine whether the fault depends on temperature, vibration, connector movement or mechanical stress.
- Do not treat a simple continuity beep as proof that the connection is reliable.
- Relate the suspicious joint to the actual circuit function: power, signal, connector, driver, relay or high-current path.
- Remember that hidden joints under BGA and similar packages cannot be fully evaluated by external visual inspection.
- After controlled rework, test the equipment again under conditions similar to those that originally triggered the fault.
Blindly reflowing every joint on a board is poor diagnostic practice. It may temporarily change the symptom without identifying the real failure mechanism, and unnecessary heating can introduce additional damage. Rework should follow a diagnosis rather than replace it.
When is resoldering not enough?
Local rework is reasonable when the defect is clearly associated with the solder connection: poor wetting, an obvious crack around a through-hole lead, a mechanically disturbed joint or a connection that has separated from the lead or pad while the PCB itself remains intact.
If the fault returns, repeatedly adding solder to the same location is unlikely to solve the underlying problem. The real cause may be a lifted pad, cracked copper track, damaged via, excessive mechanical stress, overheating or failure inside the component itself.
It is also important to address the reason the joint failed. A large connector that continues to move relative to the PCB may crack a newly repaired joint again. A power component operating above its intended temperature can continue to impose excessive thermal cycling. Correcting the solder without correcting the mechanical or thermal cause may only delay the next failure.
Hidden interconnects require even more caution. A device that reacts to temperature or PCB flex may have a cracked BGA connection, package damage or a fault in the board itself. External reflow without confirming the failure mechanism is not the same as a controlled repair.
The most useful troubleshooting rule is simple: suspect a solder fault when the appearance, electrical function and behaviour of the equipment all point to the same location. A dull surface alone is not enough. An intermittent reset alone is not enough. The strongest diagnosis comes from combining visual evidence with circuit context and reproducible symptoms.
