Linked Section Control
on the Kyle of Auchendammit Branch
The ideas presented here are NOT original
The concept was developed by L. E. Carrol in articles in early (1950s) Model Railway News
Why DCC? Why not "Linked Section Control" instead? One of the advantages given for DCC is that any locomotive can be controlled anywhere on the layout. While this is technically marvellous, it opens the door to the possibility of what is known in North America as, "corn-field meets!" Linked-section control permits an operator (driver) to control his locomotive anywhere on the layout, but only by express (pun unintentional) permission of the owner of that particular section of line. The possibility of two locomotives competing for the same section of track exists, but it would be the result of a deliberate act of mischief!
So, what is linked section control? It is a concept first published in the magazine Model Railway News, some sixty years ago in October 1953 in an article by L E Carroll. His aim was to simplify the wiring of a model railway layout, and do away with the need for a complicated set-up of switches, relays and interlocks, to ensure that power went to the railway sections that needed it. Not only was this not very railway-like, but it could be very intimidating to the stranger faced with, as the author said, "the impressive array of switches needed at every cab!"
Mr. Carroll set himself four criteria. It must, he said: be comparatively simple to install; permit realistic and interesting running of a varied nature; be equally capable of operation by one, two, or three people; and, be easily grasped by casual visitors.
His solution was not only simple, but elegant. In my opinion, it remains as elegant today as when he first proposed it all these years ago.
I can do no better than to let the original author describe the simplicity of his concept: "The basic system is this. As in real practice, the whole layout is controlled through a number of 'block-posts' (usually at stations) and each of these is linked electrically to three sections on each track - 'block-section in rear', 'station limits' and 'block-section in advance'. Each block-post (under the control of one operator) is equipped with a controller, a reversing switch, and a signal box containing the levers for the signals and any points in the area. That is all. But, each home or starter signal lever also actuates a two-way switch so that when the signal is pulled off, the track which it protects is linked electrically to the track immediately in rear of it."
The four original articles appeared between September 1953 and January 1954, with follow-up articles in June 1956, June 1958, and January 1969. Unfortunately, Model Railway News is no longer with us, but if you have a colleague who has any or all of these issues in his possession then I would strongly advocate that if you want to investigate the concept more thoroughly you try to get hold of them and take a copy.
My Kyle of Auchendammit Branch is single-line, incorporating a terminus and two through stations, and includes a reverse-loop and the option of continuous running around an oval (within a space of thirteen feet six inches by five feet six inches).
The October and December 1953 articles cover these single track elements. [The added complexities of double track are covered in the other articles cited above.]
I will liberally paraphrase Mr Carroll's articles, amended with special reference to the Kyle line, which, in fact, uses REVERSED Linked Section Control, where the train is driven by the destination, rather than the originating controller. This is more true to life and the intermediate block sections can be recovered by re-setting the appropriate signals to 'on', enabling normal operations to be resumed once the train is 'out-of-section'.
the basic switchgear required at Kyle and Loch Dubh is shown in detail in
(Basic wiring at the next station, Tannochbrae, is similar to that at Loch Dubh. Tannochbrae can be considered another terminus, or, when connected to either the "oval" configuration, or to the reverse loop, another through station.) The terminal wiring is arranged to permit the following links with the four possible signal settings:
(a) Home on, starter off. Kyle linked to Drim, for permitting departure of down trains on downstream controller;
(b) Home and Starter on. Kyle linked to Kyle controller, for local shunting;
(c) Home off, starter on. Kyle linked to Drim, for accepting up trains on Kyle controller;
(d) Home and starter off. A false setting which merely links Kyle to Drim, as in link (a). Notice that the interposition of the home switch KH in the starter circuit prevents the threatened "short" between the Kyle and Loch Dubh controllers.
At Loch Dubh station the section break is far enough beyond the points to allow a loco and two or three vans to draw ahead of them without entering the next electrical section. An unswitched shunt arm below the starter would cover this movement. The standard signal settings at an intermediate block-post such as Loch Dubh enable an operator on setting up link:
(a) to accept a train from either direction on the destination controller by pulling off the appropriate home signal (at the same time, if need be, holding a train on the other track);
(b) to carry out shunting or running-round in his station without affecting the adjacent block-section;
(c) to despatch a train in either direction on destination controller (still holding another on the other track if required), and to allow it to proceed on its way to the way to its next booked stop by getting "line clear" and favourable signals in the usual manner;
(d) to pass "through" trains.
With Kyle and Tannochbrae boxes closed Loch Dubh can secure control of the entire line merely by pulling off his home signal(s). Single-handed operation from Loch Dubh is quite feasible. Naturally, as in other sectioning schemes, auto-uncouplers and the remote control of the points is needed for running round trains at the other stations.
The only source of potential short-circuits is the pulling-off of the home at one box while that at the box ahead is also off for a train in the opposite direction (!), thus linking the same piece of track to two controllers. Simple electrical interlocking would prevent this, but I agree with Mr. Carroll that it is not worth while, because:
(a) the operator at Loch Dubh should not pull off his home until receiving "line clear" from Kyle;
(b) Loch Dubh would not ask for "line clear" if he had just given it to Kyle himself;
(c) Kyle would hardly pull off his home if he had just given a "line clear" to Loch Dubh;
(d) Even if the home were off, the pulling-off of his home signal for Loch Dubh's train would, as already shown in the terminal wiring, nullify the effect of his starter;
(e) If the short did occur the overload cut-out would trip and draw immediate attention to the error.
As noted earlier, the Kyle Branch layout includes an optional oval configuration to permit continuous running for testing purposes (read, watching the train go around!). As Mr Carroll said, "This formation has come, of late, to be somewhat despised as a layout form, but it certainly has advantages for exhibition and testing purposes, and it does allow a good length of run in a small space. Moreover, it cannot be denied that the oval has prototypes in real practice, though perhaps with rather more numerous stations. It has, however, the great failing that it is apt to pall rather sooner than a more realistic line. That is where section-linking is of help, giving each operator a real twofold task and introducing signals that are an integral part of the operational, as well as the scenic, side of the layout.
"It is just possible to run block signalling - and hence linked-section control - on an oval with only two block-posts but this does mean receiving your up trains back again from the same box to which you have just passed them on - and the same applies to your down trains. The use of bell code could disguise this shortcoming, and some sort of scenic break between the two halves of the oval would also help".
The Kyle Branch two-station single track oval configuration is shown in
The main line is the absolute minimum, but the storage sidings at Tannochbrae (which do not complicate the sectioning) should allow quite varied and interesting running to be obtained. Two trains can be operated at once on the main line, and they can be terminated at either station.
Mr Carroll goes on to say: "On such a small layout realistic timetable operation will not he achieved unless all trains stop at both stations on each journey, but nevertheless continuous running is bound to be needed at times for testing, or for amusing the uninitiated who can see no interest in proper working because it looks so slow."
But, as he says, if you use the wiring given in figure 3., and you were merely to set up link (d) at both stations, we would have an "After you, Cecil" state of affairs - a floating chain of control with no connection to either controller. To overcome this snag we must make a slight change in the wiring at the principal station, from which any single-handed operation will be carried out. As shown in figure 3a, the feeds through the home-signal switches are connected not to the station feed as usual, but direct to the controller, thus anchoring the "floating chain" to the principal controller when either home signal is pulled off.
For single-handed operation, the box at the smaller station is closed in the usual manner, and given remote control of the points and auto-uncouplers - the normal service could still be run without leaving one's seat.
However, on the Kyle Branch specifically, matters are complicated by the need also to cater for the terminus-to-terminus, terminus-to-loop, or continuous run configurations. The connections needed to permit either Kyle itself or, in the case of continuous running, Tannochbrae, to be linked to Drim Junction are routed through contacts on a relay activated by the lever controlling the Drim Junction points. This relay also controls Loch Dubh's advanced starter bracket signal as reflected in figure 4.
The reverse loop (Dounesuth) is wired through a DPDT connexion on the lever that switches the junction switch at the east end of Tannochbrae. When aligned with the main line, Dounesuth is effectively a western extension of Loch Dubh. When the loop line is selected, then Dounesuth is connected to the eastern section of the main line. To prevent over-running the loop section, and possibly having the embarrassment of having a train start up in the wrong direction, loop power is further fed through a momentary contact push button switch
Although I have long admired the concept of linked-section control, it was only with the re-incarnated Kyle of Auchendammit line that I took the plunge and actually implemented the system. There are several reasons for this, one reason being that during the long hiatus since the previous incarnation of the line, I acquired twenty-five Japanese lever four-pole double-throw switches which formed the guts of the lever frames. The signalling and point control and configuration were designed around the finite availability of these switches, which instilled a healthy discipline in the design! However, the need for more tracks at Tannochbrae, and a desire to more accurately portray the track plan at Kyle, meant that I needed more 'levers' and abandoned the key switches for less bulky replacements.
My previous layouts in North America were designed to exploit that marvellous suburban gift to civilization, the basement, and took up a great deal of real estate, external to the railway room, and the signalling of a large layout intimidated me! The basement in my current house unfortunately does not lend itself to punching holes through walls or concealing hidden sidings in the furnace room, but is more easily signalled.
Household authorities granted permission to use one room fifteen feet x twelve feet, and the resulting plan has been designed with a eye to exhibiting at local shows, where 3mm scale is virtually unknown. A final size of twelve feet x five feet six inches was settled on, with an eighteen inch extension for the end of Kyle quay. "Hidden" sidings are on show (Tannochbrae), and signalling is simplified.
Modern technology has provided mini LEDs which will permit even me to implement colour-light signals, and two-colour searchlight signals, with reasonable ease(?). I However, as the layout will reside against one wall, there will be some situations where the aspects of the signals will he facing away from the operators. Some semaphore signals will therefore be needed to allow the operator to read the signal condition from the rear. Model aircraft radio control servos will provide the motive power for both points and signals using appropriate servo control boards.
Much reliance has been placed on computer applications such as VISIO that permits such things as lever frame configurations to be easily modified to reflect signal settings sequentially, and verify that the power is going where it should! As noted above, the lever frames are made up from double pole double throw (DPDT) switches with long-bat handles, look acceptable, and work as advertised!
The diagrams illustrating this article were prepared using VISIO 2007.
Kyle is very much a "work in progress" and further reports will be forthcoming.