Semaphore Signalling
When I first set out, the idea of signalling did seem rather ridiculous and I have admit I did take some time to come around to the idea... I studied some of the books and had chats with professional signalling system designers. We spoke different languages, so I ended up designing one that is based not on any one system -but is broadly based on GWR practice. The system is electromechanical and does not follow 'accepted' signalling practice -as pulling one lever causes a sequence of events not one event.
It more closely resembles a pre WW2 binary logic comparator in short -a mechanical computer...
The diagram shows the electrical connections and diodes that 'program' the activators (I call the system MEDUSA).
To me this is an incredibly simple 8 switch truth table with 3 presets. The mechanics of getting the levers to push the microswitches is pure Henry Greenly and is made of pieces of K+N affixed to the underarms of the levers.
This allows: using buckets and vanes
A open to allow B to (open or close) -but when B is open A cannot close.
E open to allow D to (open or close) -but when D is open E cannot close.
H to open or close.
J open to allow F to (open or close) -but when F is open J cannot close.
F open to allow G to (open or close) -but when G is open neither J or F cannot close.
The instructions given in the book are for 1924 'O' gauge equipment and I have cribbed the system -but I used brass section rather than french polished plywood!!!
The points and signals are operated by 'Memory Wire' from C+L Finescale, and the single semaphore signals are the Cambrian Models NS2 kit. The gantries are home made from GENUINE Bassett-Lowke GWR white metal castings from the 1930's -which I stole off e-Bay.