What about signaling?!?

I think I can add a _totally different_ perspective on the topic
"trains and mechanical music".  If you look at a mechanical music
instrument, it has two different sub-systems:

* One might be called the "player subsystem".  This is the instrument
proper (pipes, strings, etc.) _plus_ the mechanics required to induce
a sound from it (main pallet valves in an organ, striker bellows or
magnets/solenoids in a player piano, etc.).

* The other is the "logic subsystem".  Loosely described, this is what
"can be replaced with a computer today".

Now, railroads contain "heavy logic" at one place, which is signaling.
From Armstrong levers and the mechanical interlocking machines to
electro-mechanical and relay interlocking machines to today's computers,
all this is a huge logic machine -- and this was my almost sole
"railroading hobby"!!

Have you ever seen the mechanically involved systems used to control
switches (turnouts)?  Do you know about the seamless arguments from

* the lever actually moving the points of a switch
* through the switch mechanics,
* the wires or pipes to the signal tower (signal box),
* the mechanics of a "simple" Armstrong lever (what happens if an
  engine runs a switch [correct term? going over a switch from the frog
  end when the points point in the wrong direction])
* to the actual interlocking which enforces precisely defined operation
  sequences to
* the electrical network connecting various interlocking machines
  (e.g.  for block signaling), again with a huge investment into
  security (e.g.  against lightning)

This technology originates mainly from around 1880-1900 and continued
into the 1930s, i.e. it covers exactly the time when sophisticated
mechanical instruments where invented and mass-produced.  It appears
that this was the time of "mechanical logic" in many fields.

Of course, clock making had invented the necessary technology before
(astronomical clocks have been around at least 200 years earlier than
1880), but these were always specially-made items.  Only the precise
machining technology of the second half of the 19th century allowed
mass-production and mass-deployment of "mechanical logic".

I fell victim to this hobby (signaling equipment) when I was 14 years
old.  When I became 18, I had already travelled a few parts of the
Austrian Federal Railroads and wedged myself into quite a few signal
boxes.  But then, the real thing only started: Through various
"connections", I got an official permit to visit all railroad locations
(except locomotives - these were only for "very special" people; but I
was not really interested in locomotives anyway).

One must stress that the Austrian Federal Railroads were the usual
bureaucracy then, with huge sets of rules why no-one could go anywhere
he or she wanted, so this was really "something".

In the following 10 years or so, I travelled to all places in Austria
and visited and photographed quite many signaling machines, from

* the simplest boards on the wall which just had to hold keys which
  locked switches, to
* various types of mechanical interlocking machines, to
* the electro-mechanical interlocking machines (the first of these,
  by the way, was constructed in the 1880s for the then-Austrian
  station of Prerov, now Czechia), to
* relay-operated machines (have you ever stood in the relay room of a
  large interlocking, when routes are set up and trains are travelling?
  especially in a modern relay interlocking, where relays are located
  akin to the location of their train-side counterparts - switches,
  signals, barriers, axle detectors etc. -, you hear cascades of relay
  clickings rushing though the room from one end to the other ...);
to
* - finally - the computer-operated signal-boxes - which are (for me -
  professional Software engineer) at once the most boring methods of
  "doing logic" and one of the most interesting (because they have to
  connect computers and security, an inherently fatal connection ...).

In the U.S., I only was on the tower in Buzzards Bay (when this station
only had a single track - but the interlocking machine still had the
mechanics for the original 6 or 7 tracks); and in a tower in Oakland,
California.  The only time I was in England (at the age of 16, with my
parents), I edged myself into the signal box in Dover Priory (I think),
then a completely mechanical interlocking with telegraph.

In New Zealand, I was at the Greymouth signal box; and at the station
where the line to Mission Bush leaves the main trunk line south of
Auckland, an electro-mechanical interlocking built by (I think) an
Australian subsidiary of an English company.  I was in Sweden,
in Denmark, in Italy...  In Switzerland I was given quite a few tours
by railroad people, just when I asked.  But never in Germany -- they
are so rule-fearing that they never allowed me to visit a signal box.

Of course, I do have an interlocking machine at home: a 4-meter long
old Austrian machine from the 1890s.  (I have forgotten the type -- I
cannot believe this!  Was it SA 12?  I have to look it up when I'm at
home again.)  It controls 4 switches, 4 signals and 2 of the most
primitive train detection devices: "Fuehlschienen" or "feeling bars" --
mechanical bars which are lifted above the rail surface when a route
is to be cleared.  If a wheel is standing on or travelling over a
switch, the bar cannot be lifted, the route remains locked, and the
switches cannot be thrown under the train (one of the main reasons for
accidents in stations).

Of course, railroads also had their punched-paper controls, e.g. in
the communication system (telegraphs were widely controlled by; and
punched papers strips) or in hump yards, where punched strips control
the switches to sort the cars rolling down the hump.  They also had
pneumatic controls (almost none in Europe, but in the U.S.).

They also had to deal with "safety ergonomics" (how do you prevent that
the signalman inadvertently presses a button which sets a signal on
"go") just like the organ builders had to deal with ergonomics (the
ideal placement of stop buttons, the layout of a pedal).  But now I am
leaving the realm of mechanical music -- and, of course, I could go on
and on and on...

Did anyone say "steam engines"?  Well, they are nice machines, but
definitely not really interesting ;-)  (Except for the mechanical and
mathematical details of the Walschaerts/Heusinger gear, which is
*again* a "mechanical logic system" -- and also pneumatic, if you
consider steam to be a gas "logic system" -- only that it uses up to
100 hp to move (on large engines...).

Harald M. Mueller
Grafing bei Muenchen, Deutschland
http://www.haraldmmueller.de/

 [ Harald polished his English language while working in San Jose,
 [ California, for a few years.  I think all of his spare time was
 [ spent chasing trains!  ;-)  -- Robbie