Swinging overlaps. Complex or simple?

Yes; the principle is quite straightforward, the application can be slightly involved as you "lose" a set of points out of your route calling/route priming circuits [1] and time has to be designed in to allow the setting and detecting of the blades in the new lie. Can't remember if that's 'time of operation locking' or not.

[1]perhaps not quite that low in the circuit hierarchy.

You wouldn't be able to swing it if there was even a remote possibility of another move grabbing the other overlap.

You would probably find that there is some form of split in the approach locking circuitry that maintains the possible overlaps as all applying in the same direction, obviating any racing. TORR and sectional route releasing play a small but significant part.

The overlap would be set but not locked - I'm away from my library, so this is all off the top of my head, and possibly using incorrect terminology. When I'm back in Cantuar I'll try to hunt out a circuit. I can only find an ALSR/JR circuit for route normalisation and approach locking disengageing, which only covers the 'rigjt, can I swing you?' part of the process

Having now found the 'textbook' circuit for swinging overlaps there are two main relays:

• OSR - this is the 'swing relay' it is fed either through the NWKR/RCUR up *or* the RWKR/NCUR up for the points to be swung, then the WKLJR for the same set of points. The OSR is usually slow release and will disconnect the detection of the facing points for the duration of the swing. Of course, because of the design of this particular circuit there are either NC or RC URs activated by the point keys (ie. this is not a LU style circuit, but the principle holds I suppose).
  
• WKLJR. WKLJRs prevent the points motoring for longer than necessary, in this example it is slow release (7-9 secs) to allow the blades for go across. If the blades are not detected then the controlling signal over the swung points will be restored to red via the OSR being disconnected and therefore detection being restored to the GR of the controlling signal and dropping it.

I might draw up some extract circuits and post them - but the complexity is not in the swinging part of the circuit, it is more in the point control side of things (NKLIPR/RKLIPRs, plus a lot of TIPRs in the example, which reflect relay-driven circuitry - again, this isn't directly LU style-circuits) Ignore everything apart from the OSR/WKLJR circuits - the rest is there as necessary back story.

Watch this space....

Indeed: look at the front contact of 104WZR in the feed to 104WKLJR and that gives the key to it.

104WZR is the same as 102WZR in fig 4.29.

Do Camden Town s/b Barnet platform and Tooting Broadway s/b count as this?

Yep, I see it, my problem was in reading the text under the 'swinging overlap' paragraph heading and then constantly looking elsewhere to see the whys and wherefores that were being referenced. I find the layout of the book somewhat clumsy and being unfamiliar with mainline signalling some of the relay acronyms are not immediately explicit to me. I am intrigued to see slow to release relays in signal circuits, which I expect are used to reduce circuit complexity.

Having spent the early years of my working life as a Strowger exchange telephone engineer I am quite au fait with relay timing using slow to operate and slow to release relays using armature and heel end slugs respectively or by use of standard single or double coil relays using contacts, diodes, capacitors and resistors etc to change pickup and release timing according to conditions.
Indeed both these principles (slugging or timing circuitry) were used to effect every time somebody picked up a phone handset and dialled a number, while the first selector 'A' relay contact operated and released to dial pulses at 10 pps the 'CD' relay held to it, an extreme case of 'bobbing' if ever there was one! Early designs used a slugged slow to release relay whereas later designs used a standard double coil relay and components to achieve the slow to release feature.


As for the IRSE books, apart from the red and green 'tomes' I have only a couple of the thin green booklets which IIRC are mostly points circuitry though I do still have somewhere the tracings of some pushbutton circuitry taken from the booklets once held in the staff library at 55, Broadway where I often spent hours of spare time educating myself as a young wireman in the late 1970s in order to sit a promotion examination.

From memory it's further back - The detection is suppressed IIRC as far back as 35 or 340.

Following on from my own thought I came across a reference online which suggests that swinging overlaps are inherently dangerous (pretty much my thinking too) and are actually banned by some railway authorities around the world!

Please do Tom - it would be quite interesting. Could it possibly be BB31 before a train has taken BB34/331; or BB340 following a move onto NR, perhaps?

Hopefully the very rough-and-ready sketch from memory will help:



There are also a couple of route levers 25 and 30 I think.