Wednesday, May 13, 2015


Now that the benchwork has been built, subroadbed and roadbed installed, and track laid on the mountain grade, it has been time to wire the new track.  I began with a pair of stations that have been sitting dormant for some time—Cascade Summit and McCredie Springs.  Cascade Summit actually has a fair bit of track that demands more wiring and switch machines.  I spent two weeks installing switch machines and wiring and have yet to complete Cascade Summit.

As I operated on another large, under-construction, nearby layout this weekend, I had an opportunity to reflect on differences in wiring philosophies of the two layouts.  The other layout is “dark territory”—no signals, so the wiring is fairly simple.  It really does approach the old (disproven) saying that with DCC one only needs to connect two wires and go.  My layout is not so fortunate.  It has a wiring plan that rivals a complex DC-analog layout with many separate blocks.  Contrasting the relatively simple wiring of the other layout to my own monster led me to reflect on design and construction choices I have made.

One big design difference already alluded to is that my railroad is designed to have a full signaling and Centralized Traffic Control (CTC).  The prototype SP Cascade Line had CTC installed in 1955.  Prior to that, the line was signaled using Absolute Permissive Block (APB) signals.  During the design process for my layout, I recognized that CTC would be a significant labor saving control system, just as the real railroad found.  The earlier system used Timetable and Train Order (TT&TO) control, which depends upon multiple trackside train order operators.  I have observed that model railroad TT&TO operations run “best” when multiple train order operators are used and most train crews use two persons (engineer and conductor).  That certainly is true of the other layout I operated on this weekend.  CTC allows me to eliminate the train order operator position and staff road crews with just an engineer.  This would be critical for operations with a modest number of crew members present, hence my design decision to prepare for CTC installation.

Wiring for full signaling, especially using CTC, greatly increases the number of separately wired track blocks.  A simple mountain siding on my railroad—the basic building block—has seven separate track blocks:  main, siding, “OS blocks” at each end (basically the switches), the mainline east and west of the siding, and a house track—the company spur for maintenance of way and other railroad uses.  By way of contrast, that simple “dark” railroad could include all of that trackage as part of a single power block.  A simple automatic block signal system would still need four blocks.  The result for me is a lot of separate wires running underneath my roadbed.

Underside of a piece of McCredie Springs.  Separate track block lines are for (front to rear) house, main, siding, OS-West and mainline west.  Powered frog and switch machine control wires still need to be connected to the switch motor terminal block. 

Another design and construction choice I have made is to power the switch frogs.  This is the area of a track switch where one side’s rail crosses over the other rail as the route diverges.  Left un-insulated, this would cause a short circuit.  Some modelers choose to leave the area unpowered, relying on multiple wheels of a locomotive for power pick-up on either side of the “dead” frog section.  Instead, I choose to power the frog, but this requires switching the polarity of the frog with the switch position.  I use a set of contacts on the switch machine for this function.  It is a simple wiring step, but it takes a little bit of time that adds up.  My experience with a variety of small locomotives, especially small steam locos with small tenders, convinced me of the need to power the frogs.

In a somewhat related vein, I chose to build my railroad with a feeder to every piece of rail on the layout.  Actually, a few two-inch sections are connected to another rail through soldered rail joiners, but the basic rule remains.  The reason for this is that nickel silver rail has a higher resistance per foot (more voltage drop) than copper wire.  One needs to minimize the voltage drops on the railroad to get the best performance and to ensure the circuit breakers protecting the electronics can operate correctly and quickly when a short is detected.

I have found as I build, wire, trouble-shoot, and operate this railroad that I need to break up yard tracks into several logical electrical blocks.  I have found it best to wire yard switch ladders as separate blocks, perhaps as two separate blocks for many tracks.  The other end of the yard needs the same treatment.  The body tracks can be one big electrical block, but I have found I need to wire no more than four tracks across to a single wire bus line.  Having separate blocks helps with trouble-shooting.

Taking the many separate electrical block construction philosophy another step, I have found the logical groupings of blocks also helps with “short management.”  DCC is much less tolerant of shorts than DC-analog.  The circuit breakers we use for DCC short protection nearly instantly shut down the track they feed when they detect a short.  I find myself thinking somewhat the way I did with DC-analog by providing a separate circuit-breaker-protected set of tracks for each switch crew in a yard.  Yes, there are times when two crew will share such a power district, but having switch ladders at opposite ends of a yard separately protected keeps one crew working while another deals with the short they just created.  I have yet to feed the two ends of my Eugene classification yard through separate circuit breakers, but the basic track wiring will make that a simple change.

Eugene Yard Panel.  Currently, the depot tracks are fed from one circuit breaker and the classification yard has one other circuit breaker.  I plan to add two more circuit breakers for the yard ladders at each end, leaving the body tracks on a third circuit.

DCC signal conditioning can be an issue, particularly for long wire runs.  A rule of thumb is that DCC signal condition becomes an issue for wire runs greater than thirty feet from a booster.  I addressed some of this concern by distributing boosters around the layout.  This reduces the length of wire running from any given booster to the track it feeds.  It also means I have to walk around the layout turning on the boosters when I turn on the railroad.  This also means I must run booster command bus wires to the distributed boosters and run a grounding wire among all of the boosters.  More wire. 

Even with my distributed boosters, I still have some long wire runs.  An answer to some of this length is a “snubber”—a simple resistor-capacitor circuit across the rails.  This is fine for “dark” territory, but the snubber defeats track occupancy detection.  A partial answer for me is to install the snubber for a long wire run just before the pair of track wires needs to separate for detection.  The snubber is on the booster side of a track detector.  This means I need to distribute a number of track block detectors—close to the detected track—rather than collecting them all at a central station panel. 

All of this adds up to a lot of wiring work.  This most definitely is NOT “run two wires and you are done” connection!  I could have simplified a lot of wiring just to get trains rolling, but I still would need to wire the way I am to support signaling.  Since signals are an important part the prototype railroad I am modeling, I just need to get on with it.  Wiring for signaling now will save a lot of effort and heartache later.

No comments:

Post a Comment