What exactly makes a good layout? Model railroaders seem to have some sort of Gestalt understanding of this, as evidenced by certain layouts being more universally popular than others, but there are probably more factors than you expect. Here I have tried to enumerate and exemplify the complete list.
I have divided the various factors into five categories, for reasons that will be clear soon.
This page is subject to occasional update; indeed, comments are welcome and additions will be made if I find I have missed something, or if informative images become available to illustrate anything herein. I welcome your contributions to jbs_remove@scottpages.net, though you need to remove the "_remove" that I have added for spam deflection purposes.
Viewed in this way, a layout is a connection of a number of "elemental" forms. Not all layouts have all possible configurations. In fact, very few layouts have more than a small fraction of the possible forms. This is best explained by listing the forms and giving examples.
The "Inglenook" is nothing more than a siding that branches into
three sidings, and may as well stop itself to form a siding looking
backwards.
I take the name from a type of switching layout that consists
of nothing but a set of sidings branching out from a dead end line.
Three tines on the fork is the official number as this number provides the
opportunity to pose and solve switching puzzles that involve
assembling a train from cars on the sidings in a way that
resembles the game "Towers of Hanoi" (which has three towers).
Again strictly, the capacity of the three sidings are fixed, though
such a limit is readily self-imposed in shunting puzzles.
I use the term to denote a "usable number of sidings in a
yard off a main line".
A Passing Loop is an arrangement of two switches that allows
two trains to pass each other on a single line.
More usefully in switching layouts, it allows a locomotive
to park a car, uncouple and move from one end of it to the other,
so as to reverse the order of the two on the line.
This operation is the same as would be obtained by rotating
the pair on a turntable, except that the car and loco
wind up pointing the other way in the turntable case,
but reverse order without themselves reversing direction in
the passing loop.
A continuous Loop is a piece of track around which a train may run
indefinitely. It is topologically equivalent to a simple circle,
though it may be an oval, or even a figure-eight if it folds back on itself
and crosses with a bridge.
Railroaders who derive the most pleasure from switching puzzles and are disparaging about layouts that simply run, might refer to this topology as a "tail chaser". Even the most simple form-the circle-requires an area of one "scale diameter" in each direction. (A "scale diameter" is the distance across a circle constructed with the minimum permitted radius of curvature in the selected scale. This is an important factor in the design of compact or "micro" layouts.)
As an example, the layout below is a Continuous Loop with three sidings, subtly different from a "Continuous Loop with an Inglenook", shown next:
A switchback is an arrangement as shown at right that in reality is
used to climb or descend a steep hill. It is the railroad equivalent of
a road that zig-zags up a mountain with hairpin bends at the end
of each zig or zag. Since the train must reverse in the dead end sections
rather than turn around, the train can be no longer than the dead end sections.
A switchback allows ascent up more steep an incline (travelling up the page
in the diagram at right) than is possible with curves.
In layout terms, just as in real life, it saves space, since the minimum
turn radius of a train us usually relatively huge.
A reversing loop is the arrangement shown here that causes a train to
return down the line from which it came.
The effect is logically the same as that of a turntable that could
fit the entire train on its platen at once: Reversal of the train without
uncoupling.
In 2-rail model railways, the reversing loop causes some difficulty since it would short the rails together, and various precautions must be taken to prevent the arrangement shorting out the power.
The reversing loop is not a compact structure, needing the better
part of two "scale diameters" in one direction and at least one
in the other direction.
A Wye is the arrangement at right. Like the Reversing Loop it
enables a train to manoeuvre so as to reverse itself
without any uncoupling, and like the reversing loop it
is electrically tricky to implement.
It is typically no more compact, either.
The turntable, beloved of steam enthusiasts, carries out two
functions: It has the capability of reversing the direction of
the locomotive (or very short train) that stands on its platen,
and it provides a multiway switch function typically used to
select from a large number of storage sidings.
The traverser is very similar to the turntable but cannot reverse
the direction of the train. It is typically used in yards to select
storage sidings.
It can also be used to increase the shunting possibilities on a
switching layout.
The traverser can be thought of as the "diesel age" equivalent of the
turntable, diesels and electric trains being often able to go as easily
in one direction as the other, unlike many a steam engine.
A sector plate is a multi-way switch that operates a little like
a turntable, pivoting at one end of the "holding" rail.
(It is a "sector" of a turntable, in the geometric sense of the
word.) The result is that it is very compact. It mainly finds use in microlayouts.
A double slip is not strictly a topological element, but a single
piece of track. Nevertheless, it is so functional that it perhaps
deserves a mention:
Looking like a diamond crossing,
the double slip acts as if it were a pair of points back-to-back,
so that a train arriving on any line has a choice of two lines
from which it can leave the junction.
The plan at right, as an example of using a double slip,
shows a pair of crossed sidings; if made with a diamond crossing,
this configuration would be no more than that, but made with
a double slip it provides also the function of a passing loop,
and can thus be much more useful for shunting operations.
There exist "single slips" also. These act like diamond crossings with the possibility for a train to avoid crossing but only on one side, not both (hence the names "double slip" and "single slip"). Most track manufacturers do not bother to make single slips.
This layout is a switchback, with a single siding at one end,
and an "Inglenook" set of three sidings at the other.
Note that where the tracks cross, they are at different heights,
so that there is a bridge not a level crossing.
This layout strongly resembles the oft-copied "Gum Stump and Snowshoe" plan.
Consider this layout plan: It consists of a Continuous Loop that runs through a Wye, whose branch line leads to a Reversing Loop, within which there is a single siding.
This layout below is topologically the same as the one above: It is simply bent and twisted so that it overlays itself. Some sections have been coloured to make it a little easier to trace out.
Geometric Design is another aspect of a layout. It describes properties like how well the track is squeezed into available area, how good a use is made of space, as well as how big it is---something that depends to a very large extent upon the scale and how tight a radius or curvature or how steep a gradient is allowed. The general concept of Geometric Design can be broken down into more detail, and these detailed aspects will be defined here.
Density is a measure of how much track---in running length or topological features---is
present in a given area at a given scale.
Consider again the example above, repeated at right.
A large number of features and considerable running length have been
put in a small area by means of folding the track around in tight circles,
using curved points, and running on two levels.
Complexity is what enables or limits operations. The more complicated is a layout, the more things you can do on it. An Inglenook permits more shunting operations than a branch with a single siding, and neither allows for continuous running as neither alone has a continuous loop.
Of course, an arbitrarily large layout can have ample topologies; complexity is free to increase with increasing size. By analogy, liquid mercury is very dense, and a small bottle of it weighs a lot, but if you have a big enough bottle you can still make it weigh more filled simply with water. Density is how heavy is the layout design, how many features per unit of scaled area, while complexity is how many features altogether.
Length seems straightforward: What distance of track is there? It is not independent of complexity, of course, nor of scaled size, but I associate it with the railroader's desire for train trips to take a realistic duration. Think of how long it takes a locomotive to traverse the whole layout, and you think about the length of the layout.
Length, thought of in units of time, does not exactly scale with the gauge of the model. This is because a smaller train tends to travel faster than its scale might want. A real locomotive at shunting speed might take several seconds to travel its own length, if it is manoeuvering itself carefully. A G-scale model might take half as long, but a Z-scale locomotive will demand a fine controller to travel slowly enough to take more than two or three seconds to travel its own length.
Automatability is hard to follow if you have never needed to divide a layout up into sections as is a real-life train line. Indeed, many layouts are quite deliberately not automated at all, as the user wants to do everything manually. Automation is the provision of those things that would make a layout controllable by a computer: Sensors that detect a train traversing a section, powered uncouplers, and the like.
These things come in many forms. Solenoids operate points, move signals, open and close level crossing gates. Current or optical sensors track the location of trains. Uncouplers disconnect cars. There are also a range of more frivolous and typically gadgety things such as bridges that can be raised, cars that pick up and eject mail bags as the train passes, etc. There was something of a fad for such automations in the middle of the 20th century, and there is a certain kind of model railroader who likes a layout covered in such gadgets.
The advent of DCC enabled the spread of automation, and is well suited to implementing hands-off automation, but does not help in collecting information in reverse for full computer control.
This one is rather self explanatory: Is the layout movable, and how easily? There is a whole class of layouts---microloyouts---that are especially small, and potentially portable. Z-scale layouts often appear in suitcases and similar, perfectly portable. These are in strong contrast with the more usual layout that consumes a room or attic space, and whose formwork is firmly attached to the building, or the garden variety that is huge and conforms to the land and may well contain much brick and concrete. A fixed layout might be ultimately movable, with effort. A layout may be modular, portable "in pieces". You may do well to consider this aspect of a layout when in the design phase. Modularity is discussed explicitly below.
Dimensionality of a layout is similar to the scale in which it is built. To understand it, you must understand the scale of a layout first.
Initially it appears that if a layout made of standard track sections is, say, 72"x24" in HO scale at 87:1, it should become 39"x13" in N scale, or 278"x93" in G scale. However, we know that tighter turns are possible in G scale or that we may use flex track and "break the rules" in N-scale, so that the result could be smaller than expected.
Dimensionality is the quality that describes how squeezed-down the layout is, compared to scale expectation. The size of a layout is set partly by it being modelled at some scale (N or whatever), but partly also because it may have especially smaller or larger curvature. Dimensionality is the measure of "unrealistic" sizing, usually unrealistically small radius of curvature.
It is worth saying that obsessive detail is not always sensible, worthwhile, or attractive. Few people noticed the bloodshot eyes of the driver of a G-scale car I saw a friend paint once years ago, while overly simple yet overly lush trees can make a layout look better than a botanically exact alternative. Detail should be sensible, and attractive, and can be in the eye of the beholder. I have met people who actually liked Thomas Kincaid paintings.
The most obvious things that join cosmetic appearance and track work are tunnels and bridges. Tunnels in particular can do wonders for a layout.
Consider the layout pictured at left.
It is the layout whose track plan appeared above under the discussion of
Density.
Notice how much track is in tunnel... so much so that it is not
at all clear at once where a train might emerge after it disappears.
The tunnels add considerable interest, although they conceal much track.
(See the discussion of concealment below.)
Bridges, both visible and invisible, allow tracks to cross,
increasing density for a given length, and making the path of a train
more interesting.
It is possible to have bridges without crossing tracks,
where they fulfill a largely cosmetic role, too.
Astute readers will have noted the next section
on levels in a layout; the two are closely connected,
since bridges of one track over another demand having trains
on multiple levels, but either can exist without the other.
These pictures, from their boxes, show Faller bridges used to make a layout
attractive without demanding rails on more than 1 level.
They also serve to remind you that bridges come in various sorts,
from stone viaducts to iron bridges to wooden tressles.
Bridges can make a layout consisting of a continuous
loop seem quite interesting.
First, think of the classic Marklin Briefcase layout... much more interesting
than a simple circle.
The same principle can be taken further...
...this track plan contains numerous crossovers, some visible as
bridges, some inside tunnels,
The thinner lines indicate that the track is in a tunnel.
The layout has three levels, and so crossovers can occur
easily.
Wherever they occur outside a tunnel the bridge is visible,
but the "bridges" are no less bridges when inside the
mountains.
The layout is quite intricate on account of the frequent
crossings.
The actual topology of the layout plan shown is rather simple, a fact belied by the
used of tunnels and criss-crossing complexity: It is made up of
one continuous loop, a passing loop, three sidings, and no more!
The above layout has three levels of height; a multiplicity of levels goes
hand-in-hand with bridges, hidden and visible.
Here is a view of the layout from above, so that the heights
of the various track sections are not perceived. The layout
seems intricate despite the fact that the great majority of it is
one continuous loop.
Viewed at an angle, the interest imparted by the levels is more easily
appreciated. In fact trains appear and disappear through tunnels on at
least five different altitudes in this layout, although this is not
easily portrayed in photographs.
In contrast, consider the demonstration layout at the Sacramento Train
Museum. This layout has three levels, but each is separate and there
are no grades, and no bridges that enable the crossing of lines
(although there are a couple of cosmetic bridges, irrelevant to the
levels).
Concealment virtually demands tunnels and often goes along with bridges,
and so barely deserves mention as a category itself, though the
two are different.
Trying to find and present an example of concealment without tunnels
is difficult. Other possible methods involve "concealment by confusion"
as can occur with a huge, busy layout with a large interchange,
such as this layout at the Golden State Toy Museum in the SF Bay area.
It is simply not possible to take in the whole thing at once.
A "fiddle yard" is a portion of a layout that exists solely to allow an operator to prepare and route trains. It is not decorated, and is generally hidden from the observer's view by some sort of wall or partition. It is mostly included when a layout is intended to be displayed for public viewing.
At right you see a view of a train exhibition setup that epitomises the modular scheme---the center "module" here is simply that gate to allow operators to enter and leave the central area!
Modularity can also be "future expandability".
In this form, a layout has a track that ends at the edge. When
not connected it is regarded as the route leading "off the map",
especially if the model is a replication of a real scene (see Reality
below).
The piece of the layout off the edge need not be connected for
operations, nor need it even exist.
The Val Ease Central Railroad
is constructed in separate suitcases and linked at exhibition time,
yet each section can operate alone.
Does the railway model an actual location, or does it permit shunting operations with the same constraints as a real-world yard? Having a "prototype" in actual use can be a strong appeal factor for some enthusiasts.
This has recently become a much more tractable proposition, as a result of remote sensing technology. Companies such as Keyhole allow you to pick any city and view it in detail. (Of course, that's Google Earth now, kind of dates this study!) With this, you can map its rail yards. For example, zoom in on Havana, locate the train yards near the main docks area, in a short while you can copy down the track layout, and subsequently replicate in a model. (Try it!)
More common is a layout with two scales in separate locations (so that standard, ordinary track can be used). The two gauges are justified through some logic such as the smaller being a model within the reality of the model itself, or the smaller being a narrow gauge installation, say as a mine rail branch that meets the main line to enable the mined ore to be carried away.
Perhaps the most fundamental difference in the control of layouts is whether that are DCC or Analog.
Digital control requires that locomotives carry small decoders. These receive signals from a control system, and drive the train with a direction and power level set by the remote controller. The track, meanwhile, simple carries power continuously. This allows many locomotives and accessories to operate independently with no more wiring than the rails to deliver all the power and signals throughout the layout. DCC is best suited to large layouts where a number of operations can occur simultaneously.
Analog systems supply power to the rails in the age-old fashion, removing the applied voltage to stop the train, reversing the polarity to reverse the train. This arrangement suits small layouts with one or two independent train operations occurring simultaneously at most. Analog controllers can be more sophisticated, give more realistic control to the driver, and can report such things as current drawn or present motor speed. However, they require that the track deal with isolation between circuits.
How many trains can run at once on a layout? With DCC the number can be large. If the control is analog, a layout needs work and attention to support multiple trains (operators). As a rule, multiplicity is not important for the single owner-operator, but if you want to "share the experience" with another member of the family such as a son or daughter, the option can be appealing.
The effort is reduced is the track plan takes the desire for multiplicity into account. Two parallel rails can share a shunting yard with only a couple of isolation points.
The simplest good quality train control is an autotransformer.
After that, a low-impedance voltage source generated electronically
is the best.
There are numerous train control designs that have appeared in
magazines and shops over the decades since the introduction of
the transistor, some of the more outlandish (for example the ETI1508
pictured from a 1982 magazine article at right) from the hands
of your humble author.
In the end, the type that applies pulsed power and some measure of feedback based on Brushed DC motor back-EMF speed sensing is the best, and optional inertia simulation, is about as good as it gets. The latter is excellent for shunting work. If you go analog, this choice can affect your running pleasure quite a bit.
I am collecting technical details of various controllers here.
The use of solenoid control of switches is the second electrification stage after the train control. If you want to automate a layout, or the layout is larger than is convenient for you to reach across or walk around, this becomes very appealing.
Point solenoids are only rated for intermittent operation, so a fault that powers them continuously will likely destroy the motor. Most serious builders use a capacitor discharge actuator circuit that reduces the power dissipated in the point motor and allows it to withstand a short circuit. This level of technology is easily within most constructor's capability.
Powered uncoupling is supported by some track manufacturers. This is generally only worth the effort if you want full computer control of the layout, or the layout is large and intended for exhibition.
There are two decision to be made once you have decided to add signals to a layout. The first is whether the signals are modern ones that resemble road traffic lights, or the old-fashioned but rather charming mechanical arm type. The former are usually cheaper, especially with the use of LEDs in place of incandescent globes.
The second decision is whether the signals are to be manually controlled, so that effectively they reflect the desires of the signalman (you the operator) or whether they automatically respond to train movement so that it always looks as if the train is doing what the signals command when in fact the signals simply reflect what the train is doing. This latter alternative takes a little more work, and is appropriate if the signals are intended as a cosmetic addition, or the layout is mainly for exhibition. The former choice is the one that is more appropriate if your layout is to be a realistic recreation of operations designed to occupy the operator's mind and challenge his or her organisational ability.
In a real railway, the signalman needs to know if a block is occupied. Block detection requires electrics in the real world, even if those electrics consist of devices at various staions that record the release and return of tokens from machines at stations. On a model layout, the operator can usually see the whole layout, or there is not sufficient layout to support enough trains for there to be serious risk of having a train on a bit of track without the operator knowing about it.
However, if you want ultimate realism, you will have a board that indicates where trains are, requiring block detection. Also, if you wish there to be any degree of computer control of the layout, the computer will need to have some sort of train position detection, which amounts to block detection.
There are two popular technologies to achieving block detection on model railroads: Current and optical. Current detection operates by sensing the flow of current in the rails of each section of track, with each section or block having been wired separately for power and being isolated from adjacent blocks.
Current sensing is inappropriate for DCC systems, since you would have to refrain from using the rails to supply power to any other devices (points, signals, lights) in every sensed block. In this instance, and in many analog cases as well, optical sensing is the best alternative, although this amounts to position detection (for example at the entrance to a block) rather than actual block detection.
Optical detection is mostly a Do-It-Yourself technology, or is supported by small garage-company enthusiasts, rather than being the province of mainstream manufacturers. This says a lot about its popularity. The best optical detection technology is pulsed-IR, since this is very reliable and reasonably immune to interference from external light. A sensor is typically based on the electronic components from various manufacturers that consists of an IR LED and a photodetector in a single package. They typically give an active-low, open-collector or relay-closure output when a train or part of a train is 3--20mm from the sensor face. This author's offering is documented here.