Operations 

Overview

The MVL operating scheme involves several components that work together:

 I next spent a considerable amount of time and probably too much effort before settling on a sensing and control system that provides basic point occupancy control for the layout. This means the computer-in-charge of operations has no idea where on the layout individual locomotives are located (as with RFID transponder systems), only that a loco has reached a specific sensor point on the layout.

 In other words: It's like being in the military: No one remembers your name, but they know you're around here somewhere.

 It's not a perfect machine system, but the software makes sure that things go as right as possible. To make this work on your layout, you probably need to give up the idea that all trains should have access to all parts of the layout (via turnouts and crossings) at any time.

 For the most part, each train on the MVL has only a single path over which it normally runs, and that train is usually the only train assigned to those tracks. It's not exceedingly elegant, but works reliably.

.

 The layout includes forty-two (5 and 11 mm photoresistor) optical sensors buried beneath the track ties at strategic points. The sensors are of different resistance values because extra sensitivity is required in some areas where relatively low lighting conditions exist. 

Electronics Hint #1: I've found there's no reliable way to test a photoresistor to see if it is going to work prior to purchasing it, and come away with the exact result you desire. To get the correct ones for your layout and lighting, simply buy a bunch of 'em in different sizes and Ohm values, and then try 'em all until one stands out the clear winner. Radio Shack sells a photoresistor variety pack that works good for experimentation. 

Tip: You can always add resistance to your optical circuit (by adding additional resistors), but it's pretty hard to take away resistance from the circuit when you have way too much of it.

 Each commuter station on the MVL is provided with up-track and down-track approach sensors, and a stop sensor located in the middle of each station platform.

  Because the sensors (right) are monitored by the computer in real time (about 1000 times per second), there's no immediate need for a traditional signal system, even with 10 commuter trains running at once.

Electronics Hint # 2: Because my locomotives don't have RFID transponders installed, it's hard to make an optical sensor system also run a signaling system without building a lot of extra circuitry. If you plan to add a traditional signaling system to your layout later on, you're probably better off to chop the track up electrically into short blocks from the outset, and then use current sensors to detect your trains. Both types of sensing systems (optical and current) have advantages and disadvantages.

Turnout Control 

I use off-the-shelf Atlas turnouts and twin-coil switch machines, primarily because the electrical frogs and points are inherently DCC friendly. 

Lesson Learned: The overwhelming number of electrical problems (short circuits, smoked decoders, etc.) I've encountered in the past have been as a result of great looking turnouts that are not DCC friendly causing electrical headaches when locomotives derail at the wrong place, or at the wrong time.  

All of the switch machine motors are individually wired to the central CTC panel (below) using terminal strips, barriers, and 8-conductor twisted pair phone cable. A total of 13 turnouts (Atlas #4 and 6) are controlled remotely using Atlas DPDT slide switches, which are physically located on top of the CTC panel. Each pushbutton is connected to an Atlas Snap Relay that drives red/green LEDs to visually indicate the turnout position.

 A second set of snap-relay contacts (below) are used to drive the inputs of a dedicated NCE AIU-01 board. These inputs provide turnout position information to the computer for the purpose of computer-assisted train routing. 

The computer also displays a turnout map which graphically indicates the turnout position. Additionally, NCE Snap-it Stationary DCC Decoders are used at places throughout the system to provide automatic turnout control via computer software.

Computer Software 

A total of three NCE AIU-01 input boards (right) read the turnout  contacts and occupancy sensors many times each second, and report to the computer (via the DCC command station) concerning the current status of each commuter rail line. 

The custom software program also reads the NCE Fast Clock, and then checks a database file to determine when the next train is scheduled to roll, pause, stop, speed up, or slow down.

 A prototypical station stop is performed at each commuter station, in accordance with the schedule. Train inertia (momentum) effects are produced by programming the onboard locomotive decoders.

The control program and sensors provide basic point occupancy control of the layout. This means the computer has no idea which locomotive is where. 

All the computer knows is that something (not a hand I hope!) has covered an optical sensor at a known place on the layout, and that a train is scheduled to be at that sensor at a specific point in time. The rest of the control scheme must be interpolated entirely through:

In my way of thinking, a model train layout without sound is like a day without sunshine. Cloudy days are o.k., but sunny days make you want to sing.  And even though I can't carry a tune in a bucket, I love to sing!

I purposefully waited until the train control software was almost complete before deciding to tackle the inevitable. I had to either buy or create a locomotive sound system. Of course, my first thought was to purchase a basketful of those wonderful Soundtraxx DSD150 sound/power decoders to install in every loco.

 The problem with this was two-fold. First, that bushel of new decoders would cost well over $1,300, not including the ones I would probably give the old smoke exit test during installation. Secondly, almost all of my locos already had NCE and Digitrax  decoders already installed.  Lastly, most of my motive power simply does not have room inside the shell for a large decoder and an onboard speaker.

 I decided that It would probably take some time, but I was going to try to create a real-time computer software-based locomotive sound system, and then pipe the sounds through speakers located under the layout into each room.

I started by building the computer (.wav) sound files from a high quality sound effects CD I  purchased for a different project a few years ago. The sounds of the locomotives would not be mere recordings that run in a tape-loop, but computer sound files tied directly to the actual track speed of each locomotive. 

 I first grabbed a short sample of a modern EMD diesel locomotive running at track speed (79 MPH) and extrapolated audio for each of the 28 DCC speed-steps using the Hall effect and pitch generator features of a Windows  XP audio editor called Goldwave.EXE .

 Thus, after all of the bells and whistles (literally) were added to the final audio mix, I ended up with a 16-bit stereo  .wav file that contained over 40 MB of sound data.

  I next used a sound editor to mark the byte start marker of each DCC speed step and sound (bell clang, chime, horn,, brake squeal, motor start, etc.) . Later my program would use this information to locate each specific sound byte on an electronic cue provided by the computer program at the appropriate time.

Next I spent about a week learning to program the various data registers of my computer's audio sound card. I then created and tested the sound control routines and threaded the resulting code into my existing train control software.

 Finally, I ran an audio cable from the line connection of the soundcard to a stereo audio amplifier, to which I attached speakers (a 12-inch woofer and a 4-inch tweeter) under each layout table (above). As an unexpected benefit, I discovered the camcorders I use to capture video from each room pick up the sound coming from the speakers and throw it back to the speakers in the main room. 

The unintended result is a poor man's surround sound processor, After several weeks of tweaking the speaker positions and computer code, I'm now pretty much satisfied with the result. Of course, there's always a perfect sound file that can be created as the layout progresses and the locomotives change. 

Audio/Video Feedback

The Metro Interlocking CTC Panel shields the NCE AIU-01 cards, computer and stereo speakers. Also buried back there is the heart of the A/V feedback system, an 8 X 4  (8-channel 4-output) serial matrix video switcher. The box controls the pictures streaming in from six remote color cameras, and routes the currently selected picture to one of the overhead video monitors assigned to each room. 

The computer auto-switches from camera to camera (via software) as it monitors the optical sensors over which the trains pass.

 The software guts for the two networked HP computers are supplied by MS-Windows 98, operating in MS-DOS compatibility mode .

Click here to return to the Home Page