lollypop circuit

[studentspark]

Kitchen ring final feeding one double socket, washing machine, dishwasher and boiler.
Protected by a 32a MCB

I was testing continuity of the LNE conductors and was getting a really low a reading .
Turns out the two 2.5 t&e go into a 45 amp isolator, then onto the sockets.
The continuity I was measuring was the small loop from the board to the isolator.
I tested the end to end continuity of the conductors of the load side of the isolator and got readings I would expect.

I think this is known as a lollypop circuit.
So two 2.5 cables in parallel supplying a 2.5 ring final circuit ,
Protected by a 32 amp MCB.




Would like to know your thoughts on this.

Can throw you a bit when testing[/url]

 

[PrenticeBoyofDerry]

Normally a lollipop circuit would be defined as a radial serving a ring.
If the the two 2.5mm2 cables in parallel are not loading any appliances on route, I suppose it could be considered a radial.

 

[JohnW2]

Normally a lollipop circuit would be defined as a radial serving a ring.

Indeed, and that's what it is. Although, as the OP has observed, it can result in confusion/hassle as regards testing, it is electrically really no different from a standard ring final - technically speaking, there is always at least a bit of a 'common path' (from MCB contacts to MCB terminal) prior to the origination of the two arms of the ring. Indeed, if the 45A isolator were replaced by a 32A MCB, it would be a conventional ring, originating from that MCB - having a 'distribution circuit' upstream of the OPD from which a ring originates does not alter anything.

If the the two 2.5mm2 cables in parallel are not loading any appliances on route, I suppose it could be considered a radial.

As far as the isolator, it obviously would be a radial. However, I don't really see how one could describe what happens beyond the isolator as being 'a radial', could one?

Kind Regards, John

 

[plugwash]

Would like to know your thoughts on this.

Things to watch for

1: rings where the loads are close to the supply are more likely to have balance issues than those that have long "tails" at both ends. Rings with a long tail at one end but not the other are even worse.
2: the two paralell cables feeding the switch need to be the same length, otherwise they will take unbalanced shares of the load. Similar concerns apply if there are any loads upstream of the switch.

 

[JohnW2]

Things to watch for ... 1: rings where the loads are close to the supply are more likely to have balance issues than those that have long "tails" at both ends.

Indeed, but that's true of any ring. From what the OP has said, I suspect that the isolator (fed by parallel cables) is probably very close to the CU, so probably has minimal bearing on how close loads are to the end of the ring.

Rings with a long tail at one end but not the other are even worse.

I'm not sure that is really true. Having a load close to one end of the ring would obviously tend to 'balance' a load close to the other end, in the sense of making the currents at the two ends of the ring more similar (and higher). However, such a 'balancing load' would do nothing to reduce the current in the short arm of the ring due to a load close to the other end of the ring; on the contrary, adding such a 'balancing load' would increase current at both ends of the ring. It's not really 'balance' that matters, but the actual current in the initial part of each arm of the ring.

2: the two paralell cables feeding the switch need to be the same length, otherwise they will take unbalanced shares of the load.

True - but, particularly if the isolator were (as I suspect) close to the CU, it's very unlikley that such would not be the case. If one were deliberately designing a lollipop circuit, I suppose it would be best to have just a single, larger CSA, cable from CU to the origin of the ring (4mm² would probably suffice in this case), so as to avoid any risk of unequal cable lengths. Perhaps the greatest concern about parallel cables is that, even if of identical lengths, one can theoretically becoming disconnected (or poorly connected), leaving the other potentially overloaded, without that being apparent to users.

Similar concerns apply if there are any loads upstream of the switch.

I don't really understand that comment.

Kind Regards, John

 

[deadshort]

Normally a lollipop circuit would be defined as a radial serving a ring.
If the the two 2.5mm2 cables in parallel are not loading any appliances on route, I suppose it could be considered a radial.

I've only tested a lollipop circuit once , I soon had it licked ....

It's Friday !

DS

 

[AdrianUK]

The lollipop isn't really new.....

I first saw one in a broadcast radio studio & then in recording studios.... a 32A MCB feeding a length of 6.0mm2 T&E to a 45A DP isolator by the door, the output side of which fed a 'conventional ring final' around the studio.

The idea is that it makes it easy to isolate all the gear in the studio by operating a single switch as you leave.

 

[deadshort]

Yes, and gives one hell of a thump when you come back in ! do the Speaker cones pass you on their way out in door !

DS

 

[JohnW2]

The lollipop isn't really new..... I first saw one in a broadcast radio studio & then in recording studios.... a 32A MCB feeding a length of 6.0mm2 T&E to a 45A DP isolator by the door, the output side of which fed a 'conventional ring final' around the studio. ... The idea is that it makes it easy to isolate all the gear in the studio by operating a single switch as you leave.

That makes sense - it's cheaper than the non-lollipop alternative, which requires a 4-pole isolator.

Apart from the potential confusion/difficulties associated with testing, I don't really seen any problem with lollipop circuits.

Kind Regards, John

 

[ban-all-sheds]

do the Speaker cones pass you on their way out in door !

 

[AdrianUK]

Yes, and gives one hell of a thump when you come back in ! do the Speaker cones pass you on their way out in door !

DS

Most professional amps have a 'soft-start' that only connects the speakers to the outputs after a short delay, by which time the 'thump' has been harmlessly dumped into loading resistors.

I must admit that I have occasionally had to solve issues with the power-on surges taking out the B32 MCB when so much gear with SMPS is powered up at once but that's usually solved by changing to a C32 or even a D32 if the earth-loop impedance permits.

 

[AdrianUK]

Apart from the potential confusion/difficulties associated with testing, I don't really seen any problem with lollipop circuits.....

Me neither. I guess the first time you meet one in testing it can throw you, but once you realise what's happening it's fine. I guess it doesn't fall into the category of 'conventional circuits' as defined in the good book but it isn't unsafe either.

I've implemented a variation in the office at work - a kind of ring with radials off it....... its a ring of 4.0mm2 3core SWA fed from a 32A BS88 fuse which feeds several 32A BS4343 'ceeforms' above the suspended ceiling. Each ceeform feeds a powerpole via a length of 6.0mm2 TRS which supplies 4 off single socket outlets via an RCD mounted on the pole. Each powerpole supports 4 desks each with a PC/monitor. I know it doesn't fit any of the conventional circuit models but I don't think it fundamentally unsafe.

Adrian

 

[Iggifer]

Not really any different to a ceiling mounted busbar

Other than the fact a busbar is a radial in itself

 

[ericmark]

You have two things a lollipop and conductors in parallel. If either of the two 2.5mm² feeding the lollipop circuit have even a single socket then it's a figure of 8 and not permitted although I can't recall the rule it breaks only remember it was something we should test for.

Using two 2.5mm² in parallel is common for feed to grid switch in a kitchen which in turn as multi spurs since the overload protection device is in the plugs the 3 meter limit comes in. Join two of those spurs to form a ring final and the 3 meter limit goes however since grid switches are normally 20A there is still a problem.

From your description your system complies with the regulations.

I had a heated argument with my son also an electrical engineer about standard methods. I said it's down to the electrician to read the paperwork and to adapt his testing procedure to comply with the way wired however odd that system may be. He said to wire in a way any other electrician would not expect to find is wrong even if it complies with the regulations as future electricians may as a result make mistakes.

He does have a point, we were talking about two rings left and right side of a house rather then up and down in a house as side to side gives a better ELI. But it really does not matter what one does if within regulations but not what you expect to find then errors can be made.

In your case it's something I would not personally do but can hardly fail it when some one else does it.

Of course the regulations state records should be made and these should clearly be passed to anyone testing. But in the real world we rarely see paperwork before testing and inspecting. Should we list lack of installation certificate as a deviation?

To my mind the EICR should not have anything in it which does not breach a regulation or that is of a non permanent nature. Attaching an amendment which highlights areas of concern is OK but it to my mind should not appear in the main document. But the ESC takes a different view.

I also still consider anything permitted before BS7671 was added to the regulations name as not anything to do with any condition report today. We were at one time allowed knife switches but that was well before BS7671 started. So anything permitted before 1992 forget it. BS7671 did not exist then. Again the ESC takes a different view.

To my mind we have had three editions of BS7671, 1992, 2001 and 2008 and even to go back two editions is questionable with a PIR we looked at the previous edition so BS7671:2001 would still be considered as complying. It's stretching it to include BS7671:1992 but to include the 15th Edition and lighting without an earth to my mind is going too far.

Lets see is I have disturbed a hornets nest?

 

[JohnW2]

Apart from the potential confusion/difficulties associated with testing, I don't really seen any problem with lollipop circuits.....

Me neither. I guess the first time you meet one in testing it can throw you, but once you realise what's happening it's fine. I guess it doesn't fall into the category of 'conventional circuits' as defined in the good book but it isn't unsafe either.

Indeed. I also reckon that it's as near as makes no difference to a ring final per the 'good book', particularly if the 'stick' of the lollipop is a single (larger CSA) cable, rather than a pair of cables in parallel.

I've implemented a variation in the office at work - a kind of ring with radials off it....... its a ring of 4.0mm2 3core SWA fed from a 32A BS88 fuse which feeds several 32A BS4343 'ceeforms' above the suspended ceiling. Each ceeform feeds a powerpole via a length of 6.0mm2 TRS which supplies 4 off single socket outlets via an RCD mounted on the pole. Each powerpole supports 4 desks each with a PC/monitor. I know it doesn't fit any of the conventional circuit models but I don't think it fundamentally unsafe.

It's not really far from being a conventional circuit. If you disconnected one end of the 'ring' from the fuse, it would be a (compliant) radial with branches, and I can't see that connecting an end of the radial back to the source actually changes that. The issues really only arise (as with the standard BS7671 ring final) when the cable used would not be adequate in the absence of the ring (i.e. would be inadequate for a radial fed by the same OPD).

Kind Regards, John