How safety switches work
Back to Modding & Electronics
Ever wondered how that safety switch of yours actually worked? I replace faulty ones as part of my work, and so took the opportunity to pull one apart and reveal the inner workings for all to see.
Here is the complete safety switch (Residual Current Device, Earth leakage detector, whatever) I'll refer to it as an RCD as it's easier to type. This one in particular is a combination RCD/Circuit Breaker, so it provides circuit protection as well as protecting people.
Here I have removed the sides to expose the entrails within.
Just a close up of the left side, the parts labelled are as follows:
B: Trip Mechanism
C: Spark Arrestor
D: Bi-metallic strip
The contacts can be seen as a small piece of silver metal extending down from the piece marked 'B'. The square contact surface on the copper piece to the left, as well as the moving arm are made of a silver compound to resist corrosion etc. (Don't bother trying to salvage it, you'll need thousands to get rich!)
Now a close up of the right side, labels are as follows
E: Test mechanism
F: Trip plate
G: Trip solenoid
H: Control circuit
I: Balance Coil
We will go into further detail of these parts in a moment.
First how does it operate?
The RCD is connected in such a way as to enable it to monitor how much current is entering a circuit and how much is leaving that same circuit. If there is an imbalance of 30mA or over (the minimum current required to kill a human), ie. current is flowing from active, through a person and to ground. There will be less current leaving the circuit through the RCD than entering. The RCD will detect this and isolate the power, usually within 15 to 20 ms. The RCD also has over-current detection, that operates in exactly the same way as a conventional circuit-breaker. The RCD shown has a current rating of 20 amps, making it suitiable for one power circuit or two light circuits etc.
Now for the detail.
I will start with the circuit-breaker side of things (just because that is the order of letters...)
This circuit-breaker has two modes of operation, electromagnetic and thermal. The picture shown above is of the electromagnet. This enables the breaker to trip with spikes of current. Ie, a motor drawing too much current at startup. You can see a small rod extending from the coil on the left-hand side. When there is too much current flowing through the coil which is connected in series, this rod will extend, pressing onto a striker-plate. This plate then trips the breaker (much like springing a mouse trap).
Here you can see the trip mechanism. The small white lever below the blue lever is the part the rod from the electromagnet presses against. The blue lever is what is used to manually activate/deactivate the RCD. The bit of braid you can see at the bottom continues the circuit to the next part in the chain...
The bi-metallic strip. This is for the slow current increases. The problem with the electromagnet is if you were to slowly increase the current to overload, it will not trip untill it is dangerously overloaded. This strip takes care of that. As the current increases, it heats up. This heating causes it to bend, where it eventually gets to a point where it presses a lever and trips the RCD. The problem with the strip though is it will not respond to current spikes as it doesn't have enough time to heat up. That's where the coil covers, so they compliment each other to cover both types of fault.
This little device (also known as an arc chute) is placed near the contacts that open to break the circuit, in the path any arcing will take. What it does is effectively extinguish any arcing that may occur. The fins act as a heatsink and to break up the spark to help dissipate it so it will not continue to let current flow and create a fire inside the RCD.
Contact detail (picture oriented as in normal operation)
This is how it all works. When the contacts open due to an overload, an arc is set up. Convection currents (heat rises) carries the arc upwards, past the arc guides into the arrestor, which extinguishes the flame. This all happens in milli-secconds.
Now for the other side...
This is the mechanism that allows you to test for the correct operation of the RCD. It completes a circuit inside the RCD that simulates a fault condition. This test should be carried out every month or so. You can also see the yellow indicater that tells you if the RCD is on or off. (see the first pic, it is the little yellow window)
This is what the trip solenoid (shown next) presses against to deactivate the RCD in case of earth leakage. The pin to the right of the white piece (trip plate) extends to the circuit breaker side of the RCD so when the trip solenoid activates, the RCD will turn off. (It can be seen in the trip mechanism pic as a metal loop coming through the blue lever).
This solenoid is activated by the control circuit and presses against the trip plate. The actuator can be seen as a white rod extending from the left of the solenoid. The wires can just be seen as a blur extending from the black bit.
This is the 'brains' of the RCD. It measures if there is an imbalance, and if there is, activates the solenoid to deactivate the RCD. All in under 20ms. The PCB measures about 15mm square.
This is what sends the signal to the control circuit. The wire for the active and the wire for the neutral are wound in opposite directions around a ferrite ring. During normal operation, the induced currents cancel each other out. As soon as there is an imbalance, there is an induced current produced. This is picked up by a third coil of thin wire that is connected to the control circuit. The control circuit then detects this induced current and trips the RCD.
A few notes:
RCD's ONLY operate if there is an imbalance in the active and neutral, or overcurrent. If you hook yourself up across active and neutral and do not overload the circuit, the RCD *WILL NOT* trip, as there is no imbalance (leakage to earth). However, if you touch an earth and active, it will trip. Ie. For an RCD to trip, there must be leakage of current to earth.