Lo Volts

[Bob Eager]

In addition (a separate issue) don't certain kinds of motors overheat if run on low voltage?

[John Rumm]

Yup, induction motors can draw too much current, which overheats the windings. (they are also far more likely to stall at startup on low volts as well - leaving them in a high current draw state)

[dennis]

You would have thought they would have put mesh in the walls to act as a faraday cage.

[Dave Liquorice]

I knew sustained low volts was "not to be recommended" but that excellent post really shows how nasty the consequences can be. One for the FAQ?

[The Natural Philosopher]

Its still bollocks.

If the cable can blow before the fuse trips on *any* installation on ANY voltage, its ipso facto dangerous.

Playing around with the numbers to try and prove its different in some carefully chosen case doesn't alter that.

Fuses blow on over current. Wires blow on over current. If the voltage isn't enough to blow a trip its not enough to blow a cable, or the cable is ill matched to the trip. Period.

A little knowledge...

[John Rumm]

Ah that would be the technical rebuttal then.

Well you may be disappointed to learn that this applies to all the standard circuits where RCD protection is not present. They are designed to operate safely within a certain set of parameters. If you change one of those dramatically then you can't guarantee that the design will remain appropriate.

In mitigation remember you are dealing with the situation where there are two faults simultaneously - a substantial and sustained under volt - something that is rare anyway, and an earth fault at the same time - another rare event. As an engineering exercise, it is not worth significantly increasing the cost all installations to cope with these eventualities which have a very low probability of happening together.

The figures were realistic for a long ish circuit using standard cable sizes and no other special de-rating factors, and with common circuit protection as found in many homes.

The one "carefully chosen" aspect was to look at a spur since this is traditionally more vulnerable. The second example I gave *is* slightly pessimistic - since the extended time to clear the fault is also long enough that the overall cable sheath temperature will get a chance to start rising, and hence the cable heating won't be purely adiabatic - this may be enough to save the main ring conductors (depending on the specifics of the actual installation), but still holds out chances somewhere between slim and fat for the spur's 1.5mm^2 earth.

Earth fault circuit protection without a RCD relies on getting a substantial fault current to operate the protective device quickly. Too much resistance for the available voltage stops that happening.

With regard to the current carrying capacity of the main conductors, what you say is correct. They acting alone (in the case of a radial) or as a pair (in a ring) need to have a higher current carrying capacity than the circuit is protected at.

However we are dealing with an earth fault here an not an over current. In many cases the maximum cable length for a given rating of MCB is dictated by the maximum allowed earth loop impedance (in others it may be limited by voltage drop). This is set to ensure the situation I described above is not encountered in normal operation.

Earth loop impedances are related to voltage by ohms law, if you make a significant design parameter change such as the operating voltage, you can't expect the carefully calculated and tested limits for the circuit to still be correct.

The earth connection is not designed to carry the full operating current of the circuit for any extended period of time. If it were, it would need to be at least the same size as the main conductors, and generally its not. All it is expected to do, is open a protective device quickly enough (and if you are lucky limit the touch voltage somewhat while it is doing it).

Agreed.