by devil » Mon Sep 10, 2007 2:37 pm
As an electrical engineer having written books on various aspects of assembly, I can say that soldering is a much used assembly method and is extremely reliable. I would have no qualms about any correctly designed solder joint for carrying currents far higher than 13 A. It is therefore erroneous to vilify soldering. You probably have thousands of electrically conductive solder joints in your house and car, all operating reliably.
So, let's analyse what could have gone wrong here. The first thing that could cause overheating of a solder joint is a current far in excess of that for which it is designed. As a guess, for it to reach liquidus (183°C for a eutectic tin/lead solder), it would require a minimum of 100 A, probably a lot more. This is impossible because, even if the fuse were replaced by a bar of metal, the ring main circuit breaker would trip. This is therefore extremely unlikely.
A second possibility, and quite probable, is that the solder joint was imperfect. For a solder joint to be good, intermetallic compounds have to be formed between the solder mass and the copper of the wire and between the solder mass and the bus. If something inhibited this formation of intermetallics, then there may be a poor electrical contact which could heat up sufficiently to melt the solder. Possible causes for this to happen are heavy oxidation of the metals, contamination of the metals, inadequate plating of the metals or a lack of soldering flux. A post-soldering inspection by a trained operator should detect this.
A third possibility is that soldering flux residues are too corrosive, thinning down the wires close to the joint. This can have a "snowball" effect. Corrosion thins the wires which heat slightly, the higher temperature causes the flux to corrode the wires more, causing a further rise in temperature and so on. High humidities, such as we have had this summer, could exacerbate the problem.
A fourth possibility is that the solder joint was badly designed. If I were to design a solder joint from a wire to a bus, I would have the denuded wire loop sharply through a hole in the bus and then soldered. Alternatively, the bus could be crimped onto the wire and then soldered. If the wire is simply placed along the bus, without any mechanical support, then this would be a poor choice.
A fifth possibility is a manufacturing defect termed a "dry joint". This occurs when a non-eutectic solder is used and the wire is disturbed during the cooling phase when the temperature is between liquidus and solidus. This deranges the crystallisation of the different intermetallic phases and can result in a poor electrical connection.
A sixth possibility is a simple mechanical fault, such as the cable grip being inadequate and too much mechanical stress is placed on the joint.
Now, I wish to ask a couple of specific questions. When did you buy this lead? Does it bear a label or marking ROHS or similar? I ask these because the EU ROHS Directives came into force last year and these specify that solder must not contain lead. These lead-free solders have a higher melting point and totally different mechanical characteristics, including a greater fragility. I work in conjunction with a group determining reliability issues with lead-free solder joints, so this example may be of interest to us. Is there any way I could examine the faulty item?
FYI, the BS or CE mark is not concerned with HOW an item is manufactured. It is concerned with whether it is safe in normal use (high-voltage testing, maximum current testing, shuttering of the sockets etc.). FYI, most moulded BS 1363 plugs have the wires soldered to the pins and the reliability of these is excellent.
I consider it wrong to imply that a soldered appliance is inferior to other methods of connection, rather the opposite.