What on earth was causing this?

[HappyHippyDad]

Today I had a routine socket change. The present one was tripping the RCD. The customer had used one of those double conversion sockets that utilise a single back box. I noticed that the right hand switch on the socket would not turn off.

I chased out and fitted a proper double socket. All tests were fine and the switch worked as it should. I then plugged the customers plugs back in switched them back on and thought I'd switch them off again just in case they still tripped the RCD for some bizarre reason. But, the right hand switch would not switch off.

I then changed the socket again as my new socket now had a faulty switch on it. The new switch was fine, until I plugged the customers socket back in and switched it on. As soon as the socket was energised by this particular plug the switch would become faulty and not allow itself to be switched off (just on the side that had this particular plug in it), even when the plug was removed.

I then got one of the sockets I had previously used and loosened the screws holding the socket together. The switch started working again (i.e allowing itself to be switched off), I tightened the screws back up and it was fine. However, the same thing happened everytime I plugged this particular plug in. I then chopped off the plug (it was an extension lead with 8 sockets on it and 2 USB slots), fitted a new one and everything worked fine.

Why would a plug do this this???

 

[HappyHippyDad]

Out of interest have you tried it in a new outlet, one not mounted onto the backbox? How about in another outlet in the house?

I only tried it in the one double socket, but I tried it both sides and I also tried it when it wasn't screwed on to the back box.

Could be an L-N short did on what you have plugged in. Did any protective devices blow or trip?

No, no devices tripped.

Did you check the pins size and spacing with a vernier or similar, also are they twisted in any way?
Very intriguing problem!

I haven't got any verniers. I did compare it with another plug and all the pins seemed to be exactly the same size.

What are the make and model number of the socket?

Difficult to work out but the only word on it is Santai then underneath that word 'BAP2'. BS1363/A. APPD. BY BSI KM\39477

What's connected to the extension lead?

I had a socket at my parents place which supplied my computer gear. It sometimes ended up stuck in the on position which I believe was due to the inrush current causing the contacts to stick together. This happened numerous times before it eventually welded itself permanently in the on position.

That's interesting as this had 8 computer items plugged into the extension lead. I thought it may have something to do with what was plugged into it, but the new plug works fine.

 

[HappyHippyDad]

What's connected to the extension lead?

I had a socket at my parents place which supplied my computer gear. It sometimes ended up stuck in the on position which I believe was due to the inrush current causing the contacts to stick together. This happened numerous times before it eventually welded itself permanently in the on position.

Thinking about it more I think your answer is probably the one. For some reason the new plug works ok (for now!), but I bet it was the inrush current. Thanks SparkyChick

 

[marconi]

I suspect the socket make and model suffers from 'contact bounce' exacerbated by low closed contact force and weak opening force. Thus when the contacts close and touch together there is a series of openings and closings because of the springiness and mass of the switch and contact mechanical system. Every time the contacts open an arc forms which melts some of the contact area. Once the bouncing has decayed and the contacts are pressed together, the molten metal contact areas solidify holding the contacts together. Now the contacts are not just pressed together but welded together. It will take more force in this welded state to separate the contacts than the switch mechanism is perhaps able to provide. The intensity of arcing depends on the magnitude of the transient current at switch on, supply emf and circuit reactance.

Here is a reference which goes into more depth:

Troubleshooting Common Electrical Contact Problems - http://www.pepbrainin.com/technical-resources/troubleshooting-common-electrical-contact-problems/

abd this youtube clip:

 

[Lucien Nunes]

Contact welding due to inrush, as per SC. Bog standard problem, I see it all the time, not on sockets but switchgear and controls. Probably swapped out hundreds of relays with welded or weld-damaged contacts over the years.

Electronic equipment with switching PSUs inherently have a high peak inrush current as the reservoir capacitors charge in a fraction of a second. There are usually components such as VDRs to limit it so that it does not cause problems, but cheap PSUs may scrimp on these and have excessive inrush. The inrush from multiple units in parallel adds, so a cluster of electronics with poorly inrush-limited PSUs will cause quite a splash at the contacts when they close. Even well-designed PSUs have some, and it has to be allowed for when switching large units or large banks of small ones.

Switch contacts are made of high-melting point alloys that are not supposed to weld under normal conditions. But design and quality varies from one to another due to the cost of the contact material, a decent grade of electrical accessory may have more weld-resistant contacts than a cheap one. Relays and contactors are specified for different ratings at different duties according to how damaging the load is to the contacts. Special-duty contacts for interrupting aggressive loads frequently can be made of tungsten, for example, which does not weld at all due to its high melting point, although it has other drawbacks.

Welded contacts can leave circuits energised when they should not be, so there are specific strategies in safety-critical controls, both in the design of the contacts and the circuit arrangements, to minimise the risk of unexpected or dangerous effects if a contact welds. Contact behaviour, wear-out and failure is a major subject in its own right - I must have half a dozen books on it.