Design procedure for yard ladder control using slow-motion switch motors

Introduction

In an article in the August 2011 Model Railroader, I described a method to control yard ladders with a single rotary switch. By simply turning the rotary switch, all turnouts would automatically line up for the selected yard track. The method was based on using a bipolar power supply, Tortoise switch machines, and one SPDT contact on the Tortoise switch motor. A small number of diodes were also required, under certain conditions.

The article described the problem and then presented the solution for a few yard configurations. Although any yard pattern can be controlled with this technique, it isn’t readily obvious just how to arrange the switches and contacts for every situation.

So I came up with a simple six-step method that will guarantee a workable design for virtually any yard configuration. This method requires no formal training in Logic Design, so it can be used by any modeler who has a need or interest. I’ll describe the steps in detail and give examples of more complex yard configurations. But first, let’s see how the technique originated.

Background

It all starts from the basic series yard ladder. Figure 1 shows a five track version along with its truth table. The labels on the yard plan are arbitrary – you can put the plus and minus signs in any order, as long as there is a plus and minus for each motor. Any two-state symbols can be used. We can use a1 and 0, true and false, or + and -. Here the plus and minus make more sense, as it directly relates to the polarity we will apply to pin 8 of the Tortoise switch motor.

Constructing the truth table is critical, so you have to understand how to do this. We simply make a blank matrix and label the motors across the top and the track numbers down the left side. Now we fill it in based on how the track plan was labeled. For example, to get to track 1 we need plus polarity on all four motors, so we fill in row 1 with all + signs. The third symbol used is the X for a don’t care situation. When going to track 3, we don’t care which way turnout A is heading, thus the X in box 3A.

A brute force, full diode matrix can now be drawn directly from the truth table. Incidentally, diode matrices have been around for a long time. If memory serves, they first appeared in Model Railroader some 30 years ago and used to control ladders with double coil switch machines. Figure 2 shows such a design applied to Tortoise switch motors and using a bipolar power supply. Any plus in the truth table requires a diode in the upper deck of the rotary, and a minus requires a diode (in reverse direction) in the lower deck. So diodes D1 thru D4 perform the logic from row 1 in the truth table. If you are not familiar with this, study fig. 2 for a moment and you will see how it is formed.

If you actually constructed the circuit in fig. 2, it would indeed control the yard ladder as desired. But note that 14 diodes are required. It is not so much the cost (as diodes are only about five cents apiece) as it is the time and aggravation of wiring up a full matrix. Worse yet, the number of diodes increases rapidly with more tracks in the yard. To figure out how many diodes you need, you sum up the numbers from 2 to the number of yard tracks. For example, a seven yard track would require 2+3+4+5+6+7 = 27 diodes! Mathematicians call this a Fibonacci series. This is the main motivation to look for alternative methods of control.

Now we need to make the first mental leap. If you stare at fig. 2 long enough, you will see that some diodes can be shared if another logic element is introduced. If you were to place an SPDT contact at point X that switched diodes D2 thru D4 to position 2 of the switch when A moved to position 2, you can eliminate diodes D5 thru D7! Similarly, placing contacts at points Y and Z allows the elimination of diodes D8 thru D10.

The purpose of the diodes is to block current from flowing in the wrong direction, causing an invalid logic condition or a short circuit. Once diodes D5 – D10 are gone and the contacts introduced, a little circuit tracing reveals that NONE of the diodes are really required, and the circuit will work. The circuit was presented in the MR article and is also shown here as fig. 3. I have inserted the three diodes that are in series with the wipers on the contacts, for reasons discussed in the MR article.

Now we have arrived at a method that requires only 3 diodes and one fourth of the solder connections (compared to a full diode matrix) to accomplish the task. Note that indicator LEDs can be placed on the panel and easily hooked up directly to one deck of the rotary switch. They can actually share one resistor, as only one lights at a time. The resistor should be somewhere between 560 and 1k ohms. This option is shown in fig. 4.

I do need to discuss the more straight-forward option. Here we just place four SPDT toggle switches on the panel and wire up each motor separately. From a wiring and conceptual point of view, this is certainly simpler. But from an operational point of view, it leaves a lot to be desired. I have operated layouts with multiple yards of seven or more tracks that were wired up with the simple option. It turns out that new operators constantly select the wrong route until they become familiar with the track plan. Invariably the owner of the layout has to station an experienced operator at the yard panel to insure smooth operation. It is your choice, but remember that wiring is a one-time effort but operation is forever.

Now that the preliminaries are over, let’s discuss the six-step design process.

Designing the control circuit

Step 1: Sketch and label the yard track plan.

As an example of this process, I will use a compound yard track plan shown in fig. 5a. Although the plus and minus assignments are arbitrary at this point, the process will work out easier if the routes to the farthest tracks are labeled with plus signs.

Step 2: Fill in the Truth Table.

In fig. 5b I have created and filled in a truth table matrix based on the yard track plan and assignments generated in step 1. From now on, I will leave the don’t care entries in the truth table blank for clarity instead of using an X.

Step 3: Make the truth table upper triangular with respect to the plus signs.

This sounds like a mouthful, but I’ll explain the terms. Take a closer look at the truth table in fig. 1, and note that it has an interesting pattern. Minus signs are only located on a diagonal going from box 2A to box 5D. Everything above this diagonal are plus signs, and everything below is a don’t care situation. When a matrix is in this form, it is called Upper Triangular, for obvious reasons.

Here is an interesting fact about truth tables. You can swap rows and columns, at will, and it does not change the overall logic. But you do have to swap the entire row or column, along with its label. You cannot swap individual box entries. Here is the key point: we know the circuit realization for an upper triangular truth table. It is shown in fig. 2. All we have to do now is to re-arrange our truth table, by swapping rows and columns, until it looks as close as possible to upper triangular.

In fig. 6 I have re-arranged the truth table in figure 3b, by swapping columns only, to get it close to upper triangular form. To do this I also had to change some of the don’t care entries to plus, which is perfectly legal. With the exception of the 5B minus entry, it is upper triangular. We’ll take care of this in a later step.

In case you are wondering, this step can be done with any yard pattern. You should end up with all plus signs above the diagonal and only a couple of stray minus signs below the diagonal. If you cannot, try swapping the signs on the turnout motor that is giving you problems and repeat from step 2. You will have an easier time if you enter the motor labels in the truth table starting from the far point in the ladder, and working towards the throat. In fig. 5b I deliberately did it backward to illustrate this.

Step 4: Draw a preliminary schematic.

We now make a preliminary schematic that is exactly like the pattern for the simple series yard. Of course, you have to extend the pattern to match the number of Tortoise switch motors in your particular ladder. Now here is another key point. You label the motors, left to right, in the SAME order determined by your upper triangular truth table. So now you see the purpose of step 3. It determines the order in which the motors have to be connected, and which contacts go where.

So I don’t get things confused when it comes to wiring, I label each Tortoise switch motor according to the plan in step 1. Then I carefully associate the motors on the schematic with the correct motors on the physical layout. We are almost home free now.

Step 5: Add diodes to handle the missing logic.

Remember that oddball minus sign in 5B? To handle this, we need one more diode that will force motor B negative in switch position 5. The final schematic is shown in fig. 7.

Step 6: Wire it up, test, and install.

As I do my wiring on a fixture, I can test it before actually installing on the layout.

I will work another example that is a little more complex. In fig. 8, I have drawn a compound yard in parallel with a simple series yard and its associated truth table. This configuration actually exists on a friend of mine’s layout. In fig. 9, I show the upper triangular version of the truth table. Now we have three stray minus signs to account for. Fig. 10 shows the final schematic.

Final comments

The design algorithm presented here was chosen for its simplicity. It does not require a college level course in logic design to understand or apply it. Trained logic designers will note that it does have a small flaw. Namely, it doesn’t always yield the absolute minimal design. But it is only off by a couple of diodes. This is not too much of a price to pay for the simplicity.

If you like logic puzzles, you could see if a contact configuration can be found that eliminates some or all, of the extra diodes. But by the time you do that, you could have wired up a couple more diodes ten times over. As a challenge, apply this procedure to the second example in the Model Railroader article. You will come up with an extra diode. Compare this to the schematic in Model Railroader, which has no extra diode and is minimal in configuration. Let me know, there is always a better way to do something. I hope someone can find it.