The IET is carrying out some important updates between 17-30 April and all of our websites will be view only. For more information, read this Announcement

This discussion is locked.
You cannot post a reply to this discussion. If you have a question start a new discussion

Recommended wiring topology for multiple lamps on one switch?

As having an array of mains downlighters or spotlights in a ceiling is now fashionable, (rather than a single central luminaire), this begs the question of how to optimally and correctly wire them all up to the light switch.

In the case where one light switch operates multiple lights (as powered by from the domestic lighting circuit), there are several possible ways to wire from the switch to each of the lamps.

[It is taken as obvious that each lamp needs to be wired in parallel, with connections to the mains neutral and the switched-live wires.]


The (twin-and-earth) cable routing from the switch to all the lights and could be one of:

1) Daisy-chain   (the cable goes from the switch to one lamp, and then on to the next lamp, until all lamps are connected)

2) Ring    (like the daisy-chain wiring, but the final connection then links back to the switch to create a "ring main")

3) Star   (separate cables go from the switch to each lamp)

4) Tree   (i.e. a balanced spanning-tree, where the cable goes from the switch to two lamps, and each of them connect to two lamps, branching until all lamps are connected)

Each topology has different implications regarding the current distribution, and its associated heating losses in each segment of cable.

For example, a Daisy-chain topology has the highest current loading in the cable segment closest to the switch, but it uses the least cable; whereas the star topology has equal current loading in all its cable segments, but it uses the most cable.


Do the IET regulations recommend a particular wiring topology for particular scenarios, and what does everyone use in practice?

Thank you
Parents

  • 1) A ring topology balances the currents in the ring such that each half caries half the current. This halves the wire thickness needed, reducing costs.



    In practice there's a bit more to consider. Since it's difficult to guarantee precise sharing of the current around a ring, the regs generally require a minimum cable rating closer to two-thirds the circuit rating, rather than half, for any ring (e.g. 20A cable for a 30A ring) - so any potential saving might not be as great as foreseen. Indeed the regs also impose a minimum conductor size for reasons of physical robustness and soundness of connections - which for lighting circuits is 1.0mm² - more than likely the size that would be used anyway - so in practice no reduction is cable size at all is likely. Indeed the extra leg needed to complete the ring will actually increase the length of cable used and the extra testing needed for a ring compared with a simple straight radial will add to the labour costs so overall increasing costs not reducing them.


       - Andy.
Reply

  • 1) A ring topology balances the currents in the ring such that each half caries half the current. This halves the wire thickness needed, reducing costs.



    In practice there's a bit more to consider. Since it's difficult to guarantee precise sharing of the current around a ring, the regs generally require a minimum cable rating closer to two-thirds the circuit rating, rather than half, for any ring (e.g. 20A cable for a 30A ring) - so any potential saving might not be as great as foreseen. Indeed the regs also impose a minimum conductor size for reasons of physical robustness and soundness of connections - which for lighting circuits is 1.0mm² - more than likely the size that would be used anyway - so in practice no reduction is cable size at all is likely. Indeed the extra leg needed to complete the ring will actually increase the length of cable used and the extra testing needed for a ring compared with a simple straight radial will add to the labour costs so overall increasing costs not reducing them.


       - Andy.
Children
No Data