Gopher Valley Central
GmtJqS1.png
GVC_diagram.jpg

Built By: Jeff Faust

Description

Self-contained, 24" x 48" N scale layout which may be used as a reverse loop in a T-Trak layout. Twice-around oval with reverse loop; track plan based on Plan 33 from "101 Track Plans." Minimum curve radius 8", maximum gradient 4%. Atlas Code 80 track on cork roadbed with Unitrack transition at interface (Unitrack bumper attaches for self-contained operation). Turnout at reverse loop modified to function as a spring switch. Reversing section is powered via Digitrax AR-1 auto-reverser circuit for DCC operation. Power feeders may be attached to terminal strip on underside. Built-in transistor throttle (currently disconnected) for DC operation. Natural-finished oak fascia.

Evie-Strong_321.JPG

T-Trak Application

Connects to T-Trak modules via South Sinnemahoning transition module. When used in a T-Trak layout, the Corrugated Curve reverse-loop module is typically used at the other end. (Here's a video of the GVC in action, taken at the RIT show, December 2010.) The bi-directional, single-track operation, when used with DCC and multiple trains, opens up a whole new world of opportunities for head-on collisions…so I'm installing signalling! (Details below.)


Reversing loop in action:


Signalling

semaphore_0437.JPG

When the GVC anchors one end of a linear T-Trak layout, it does create a bottleneck. A train passing South Sinnemahoning travels a long stretch of single track, traverses the reverse loop, and retraces its route back to double-track territory at the T-Trak modules. Meanwhile, other trains have to wait—or collide head-on if the engineer gets distracted (which does happen sometimes at a show, believe it or not). "What I need," I said to myself one day, as I separated two opposing locomotives, "is a signal…"

A signal! Ah-hah!

So, here's how I'm doing it:

Block Detection: Since our club uses NCE PowerCabs, NCE's BD-20 block detectors seemed the obvious choice. (The fact that NCE's headquarters is just across town was a factor, too.) A feeder to the block gets three turns through the "donut," which yields enough sensitivity to detect my 44-tonner parked with the headlight off. The BD-20's logic output is connected to an input pin on the Arduino with a 10K pullup resistor, as per NCE's instructions. The output goes "low" when a train is in the block, which seems counterintuitive to me, so I invert it in software.

Servo: If you think it takes a Tortise to move a semaphore blade, you've been searching the wrong end of the hobby shop. I strolled over to the R/C counter, and asked about servos. They handed me a HiTec HS-55; it's small enough to easily fit under a T-Trak module (even mine!), and only $14. Just about any servo should work, however. Three wires connect it directly to the Arduino; no limit switches or other components are needed. I positioned the servo directly below the hole for the throw rod, glued two small pine blocks into position for mount points, and attached the servo with the provided screws. That was simple!

servo_0439.JPG
Semaphore: I installed NJ International's #2000 (square-end blade for absolute stops). It's all-brass construction, which is reassuring when small kids are around. Since the throw rod was too short to extend much past the module's subroadbed, I made a longer one from 0.020 brass rod. Easy.

Logic: I'm using an Arduino Uno microprocessor from RadioShack. The online documentation, resources and code libraries for these things are extensive. All the necessary signalling logic happens in the code. Want to slow down blade movement? Delay a clear indication by half a second? Add a buzzer for the approach circuit? A flashing occupancy lamp? Just change the software and plug in the desired gizmos.

The components were rigged together temporarily with a breadboard for the LOTS Convention on July 14, 2012, and worked beautifully all day long. By mid-morning, I had gotten the approach buzzer to work, but the piezo element on hand was not quite loud enough to command attention over the ambient noise of the exhibit hall. Nonetheless, we managed to hold down the head-on collisions to just one—that's quite a drop!

A second semaphore, facing eastbound traffic, was installed and operational for the Strong Museum exhibit in mid-September 2012. Wiring remained a breadboarded temporary affair. The following video shows both semaphores in action.

For the Diplomat show in November 2012, permanent wiring was installed and (mostly) working. An occupancy LED was installed on the back of the South Sinnemahoning module for operator convenience, and the approach buzzer was upgraded. This show also marked the debut of the new Sunset Valley module, with two more semaphores and its own Arduino. The connection between the two Arduinos (and their attached BD-20s) was not yet functional, but some quick re-coding allowed one of the new signals to be triggered by photocells installed between the rails.

The Sunset Valley module, with both signals working, was used in the RIT show in December 2012, running in self-contained (photocell train detection) mode. The semaphore lamps were connected, and were both burned out by the first hour of the show, but everything else worked flawlessly. The I2C connection between the two Arduino boards was functional for the November 2013 Diplomat show. Now an occupied block can trigger an "approach" indication up the line.

Speedo_033.JPG

I made just one addition for the March '14 Diplomat show: a digital speedometer. Some additional code takes a timing of the block's occupancy, divides it into the length, and displays the speed on a connected SparkFun Open Segment Shield. Next up: LEDs for the signals.

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License