Weird result determining order of sockets on ring circuit

[JohnW2]

As to currents, in the absence of faults, I would have thought a figure of 8 would reduce the ability to overload one leg of the ring.

Indeed so, as I've often written. Except in the ususual situation in which loads on the circuit are already 'completely balanced', adding any 'cross-connections' is bound to move the situation towards one in which vthe total load is equally shared between the two legs. That would be even more true with 'multiple cross-connections' (rather than the one which exists in a 'figure of 8' arrangement). Cross-connections also increase redundancy, hence reduce the effects of any faults (breaks) in the circuit.

As I've also often said, and you repeat, as far as I can make out the only real downside of a 'figure-of-8' is the difficulty/problems it creates in relation to testing.

Kind Regards, John

 

[EFLImpudence]

Agreed, however, I could, of course, be wrong but reading Eric's comment:

"From what I remember it's down to where on the ring the figure of 8 joins, at the centre of the ring it is not really a problem, but near to the consumer unit it can cause too much current to be drawn on one leg."

Because of the "joining at the centre of the ring" I get the impression that he is talking about a ring on ring situation rather than just a bridge which is the usual configuration..

E.g.

 

[SimonH2]

Ah yes, the ring-on-ring topology is effectively putting a point load (everything on the "distant" ring) onto the main ring - with all the issues that can cause. (IIRC) as long as the junction is more than something like 1/4 of the main ring from the CU then there's no problem with typical installation methods and circuit lengths for a domestic install.
But, if it's the choice between the diagram on the left, and opening the joint to make it one ring, then again - the current in the shorter leg may be less with the figure of 8 than with the single ring. For the figure of 8 to be worse in terms of current in the shorter leg of the main ring, you'd need a lot of load at the top of the smaller ring - and for it to be a problem, the junction would need to be fairly close to the CU.

 

[JohnW2]

Agreed, however, I could, of course, be wrong but reading Eric's comment: "From what I remember it's down to where on the ring the figure of 8 joins, at the centre of the ring it is not really a problem, but near to the consumer unit it can cause too much current to be drawn on one leg." ... Because of the "joining at the centre of the ring" I get the impression that he is talking about a ring on ring situation rather than just a bridge which is the usual configuration.. E.g. ...

Possibly, but I'm not sure that it makes much difference to what I/we were saying. You could re-draw the 'ring on ring' to look like a 'conventional figure-of-8", the only difference being that the 'cross-connection' would be of 'zero length'.

Kind Regards, John

 

[bernardgreen]

In a cottage near here I found a "figure 8" where a ring had been extended by removing a single socket and fitting 6 terminals behind a blanking plate. No connection between the "sides" of the ring. This was in a wall about 4 foot above floor level. One cable vertically up and three vertically down.

 

[JohnW2]

Ah yes, the ring-on-ring topology is effectively putting a point load (everything on the "distant" ring) onto the main ring - with all the issues that can cause.

That's true, but (although I haven't yet had the time to think very deeply) I think it's probably still 'better' (at least in some respects) than the situation which would exist if you broke both rings at their point of contact, thereby turning the whole thing into a single ring.

Kind Regards, John

 

[John D v2.0]

That's true, but (although I haven't yet had the time to think very deeply) I think it's probably still 'better' (at least in some respects) than the situation which would exist if you broke both rings at their point of contact, thereby turning the whole thing into a single ring.

Kind Regards, John

Could be better if the total length of the combined ring would be too long to meet disconnecting times or have too high voltage drop.

For a ring on a ring, it would have to be connected symmetrically to be balanced properly in terms of current. The easiest way would be to attach at the mid point or at the cu. However you could have a ring with the cu at 12o clock and the extra ring leaving at 4o clock and rejoining at 8o clock.

Basically if you can put a mirror across it and it reflects perfectly it's going to be OK for current balance but subject to resistances.

 

[JohnW2]

Could be better if the total length of the combined ring would be too long to meet disconnecting times or have too high voltage drop.

That's fairly unlikely with the sort of ring lengths one is likley to see in a 'standard' domestic installation but, in any event, I don't think it is necessarily the only sense it which it could be 'better'

For a ring on a ring, it would have to be connected symmetrically to be balanced properly in terms of current.

Sure, for any ring arrangement, no matter how complex, it has to be perfectly symmetrical (and have loads perfectly symmetrical) in order to be 'perfectly balanced'. However, we are not talking about 'perfect balance' but, rather, about things which make a circuit 'more balanced' - particularly in the presence of unbalanced loads.

The easiest way would be to attach at the mid point or at the cu.

If the 'attachment' was at the CU, one would effectively have two totally separate rings, which is a bit different.

However you could have a ring with the cu at 12o clock and the extra ring leaving at 4o clock and rejoining at 8o clock.

That's obviously more like the sort of figure-of-8 one comes across. However, even if it is not symmetrical (e.g. joined a 'string of sockets', hence cables, at 4 o'clock and 10 o'clock) that cross-connection would usually still improve the balance of currents in some of the cables (in the presence of imbalanced loads) to some extent.

Kind Regards, John

 

[John D v2.0]

I need a lie down (somewhere cool)
This is interesting though, but I feel the main problem is the next person to test the circuit will be confused!
Does anyone provide diagrams of circuits as installed with new installations? I think here it would be required.

 

[JohnW2]

I need a lie down (somewhere cool) ... This is interesting though, but I feel the main problem is the next person to test the circuit will be confused!

Indeed. The big problem of these circuits is that they present testing problems - particularly when the person faced with the task does not initially even realise that they are dealing with an 'unconventional circuit'.

That (fairly big) downside aside, I don't think that there is anything about a 'convention figure-of-8' circuit (I'd have to think more about the 'ring on ring') which makes it in any sense 'worse' than a single ring - with at least some senses in which it is arguably 'better'. However, given the testing problem, I doubt that it will ever 'catch on'!

Kind Regards, John

 

[John D v2.0]

I just wonder what shape your house would have to be to benefit from one! I'm imagining the star ship enterprise!

 

[JohnW2]

I just wonder what shape your house would have to be to benefit from one! I'm imagining the star ship enterprise!

I don't think that's usually the point. I think that figure-of-8's most commonly arise 'unintentionally' ('accidentally'), when an original ring is 'extended'.

Kind Regards, John