Lock iconRectangle 1Rectangle 2 + Rectangle 2 CopyShapeRectangle 1

A CATS Tutorial Using the Dr. Bruce Chubb Examples in Feb 2016 RMC

R.S. Johannes, Signal Master HUB Division Modular Railroad

Once more, Dr. Chubb in his own inimitable manner has provided us with yet another marvelous teaching example.  In the Feb 2016 Railroad Model Craftsman [RMC], Dr. Chubb provided a clear and highly visual description of speed signaling.  The HUB Division of the NER is located in Massachusetts and our modular railroad is heavily invested in JMRI, the C/MRI and the CATS extension to JMRI.  Being an “eastern road” we use speed signals in form of NORAC Version 10 rules for signaling on our modular railroad, the Hoosac, Upton & Boston.

As we are constantly looking new and better for ways to teach ourselves, new members and other interested model railroaders, we keep our eyes peeled for teaching opportunities and Dr. Chubb just provided us with an outstanding one to demonstrate speed signaling.  I reasoned that recreating the track segments using CATS would provide a useful interactive extension to the graphics provided in Part 3 of the Signaling series and found on page 56 of the February RMC issue.  By creating a CATS panel, people wanting to learn could use “simulators” as opposed to real hardware and see multiple “what ifs” beyond the settings shown in the article.   It also would provide a good test for the functionality of CATS as well.

CATS has two features that are useful to teach Dr. Chubb’s examples using speed signaling.  First, for each signal type, such as 1-headed, 2-headed and 3-headed searchlight signals, CATS creates an easily edited Signal Aspect Template specific to that signal type.  This template allows one to edit the actual head color settings across the range of aspects/rules.  All you need is a copy of a rulebook to find the settings you want to use on your railroad.  The NORAC rules can be easily found online and I have a printed version used the former Guilford Rail System, now PanAm Railways.  The settings I used to define this table for a 3-headed searchlight are shown in Figure 1 below.  The settings are simply selected by using ordinary window’s dropdown menus. 

Basically, one goes rule by rule through the NORAC rules and edits the signal template to match the rulebook.  Several of the settings did require editing from their default CATS settings to bring them into alignment with NORAC 10.  It is not the default settings but rather the simplicity with which they can be edited that matters.  For example, Rule 290, restricting, defaults to red over red over red but in both NORAC 10 and Dr. Chubb’s figure on Page 58, restricting for a three headed searchlight is red over red over yellow.  [See Figure 1]  Also note that NORAC 10 uses red over yellow over green for medium approach medium as opposed to red over yellow over blinking green used in Dr. Chubb’s article.  This just goes to further emphasize the variability in aspect/rule definitions on a railroad-by-railroad basis.  The strong advantage of the CATS signal template is that instances of a particular signal type and their associated template are easily planted onto the track diagram at any arbitrary location where they are needed. 

FIGURE 1.  The Signal Aspect Table for Chubb Examples A, B, C

The second of the two important CATS features is that each track segment is assigned one of four speeds, normal, limited, medium or slow.  A block uses the minimum speed found in all of the track segments making up the block to define the overall speed of that block.   At turnouts, each leg can have different speeds associated with each of the two legs.  For example, Figure 2 below shows that Turnout 3 in Chubb’s track diagram is set to “normal” over the straight leg but to “medium” on the diverging leg.  This will have immediate implications for signal aspects as a function of which pathway is selected as the route across that turnout.

FIGURE 2.  Setting Speeds on Track Segment in CATS

For four of the six examples covered in article, only three signal types and their associated Aspect Templates are needed, namely one, two and three headed searchlight signals.  For examples D, E & F, a fourth signal type is needed as there are two two-headed dwarfs used in these examples.    Their definitions are found on Page 59 in the RMC article and also in the NORAC rulebook. 

So now it is time “test” these “layouts.”  This is another reason I sought to create these examples.  Using JMRI simulators, you can be certain as to how the layout will respond to various states of occupancy and route settings before ever dealing with hard wiring the layout.  This is a tremendous advantage and keeps you on top of the layout instead of under it for the vast majority of the time.  CATS track diagrams are produced by a separate application called Designer.  Designer allows you to save its output as an XML file that is then loaded in the CATS runtime application.  However, when you are in the learning or design phase, it is quite useful to be looking a larger array of data than you will when actually running your railroad.  Here are the tools I tend to use and the sequence of events I use to open them.  First of all, you launch CATS.  Be sure to create preferences for devices you will be using.  In these examples, I actually created simulators for NCE, Lenz and C/MRI.  In the case of C/MRI you will also have to create virtual nodes such as SMINI 1, SMINI 2, etcetera to allow you to address them.  Once your panel is fully debugged, changing them from virtual nodes to real nodes should leave you with an operational signaling system but the lion’s share of the work was done at your desk not below the railroad. 

The three most useful debugging windows I open are the JMRI system console, the JMRI sensor table and the JMRI signal head table.  For this simulation the sensor table is not necessary but it is valuable during debugging as well.  After loading a CATS xml layout file but before opening the system console, I use a feature in CATS to set what events I want to be traced under the appearance dropdown and select “Signal Changes.   See FIGURE 3 below.

FIGURE 3.  Selecting Signal Aspect Changes for Tracing

With Tracing set, it is now time to open the System Console.  This is found under Help Dropdown on the CTC Panel window as can be seen in FIGURE 4 below.  The system console comes up as a simple scrolling computer terminal text screen.   It is well worth looking at this when first loading a CATS file as it should be free of warning or more importantly error messages.  These should be tracked down and fixed before going further.  The reason I like having this screen open is with signal changes set to be traced, every time a signal changes its aspect for any reason either as a consequence of occupancy changes, a routing decision by the dispatcher or as a consequence of change in aspect from a nearby signal with a dependency, you will see both the rule that was invoked by number and actual colors of all the signal heads involved.

FIGURE 4 Starting the System Console.

The final step is to open JMRI Signal Head table under tools in CTC panel window.  When all of this is done, your screen should look similar to the screen shown below in FIGURE 5.  The CATS dispatcher panel is shown in the upper left hand corner.  By default, CTC signals will be shown as white but their actual setting on the railroad will be all red or stop until the dispatcher opens up some routes for moving trains.  Below the dispatcher panel is the system console, free of error or warning messages.  Below system console is the CTC Panel window housing access to the myriad of useful JMRI tools including operations and WiThrottle.  Lastly on right hand side is signal head table.  I used the naming conventions from the diagrams in Dr. Chubb’s diagrams to name the signals.  These collate in an order that makes it easy to see the setting of all the heads on any particular signal.  Heads are numbered and the convention is that Head 0 is on the top, Head 1 is the middle head and Head 2 is the lower head.  As shown in FIGURE 5, signal head 4R_0 [top], signal head 4R_1 [middle], and signal head 4R_2 [bottom] are all red consistent with the dispatcher panel display that shows Signal 4R as a three headed white graphic.

Of note, clicking on the color buttons in the signal head table allows cycling through the colors available for that head and once you have actually attached real signals, you can confirm your wiring by seeing them change correctly as you select the colors.  Both solid and blinking aspects can be tested this way.

Notice that there are four signals that are visible and are not white in FIGURE 5.  That is because these are intermediate signals not CTC signals.  Many modern panels omit these signals from the display.  This is easily done in CATS by asking CATS to remove the signal from panel but retain the location and definition of signal. 

FIGURE 5 The CATS Debugging Environment

Right clicking any track segment allows the dispatcher to alter the state of that track segment.  For example, besides setting occupancy, one can also take the track out of service or perhaps more importantly grant track authority to a block.  Granting track authority will make the track segment appear blue and set signals protecting it to restricting.  On modern panels this is much like the call-on sequence found on US&S panels.  We need this as all of Dr Chubb’s examples have a train/locomotive, Train A, on the west-most track block on all six diagrams.  Three of the diagrams use a single track diagram.  These are diagrams A, B, and C.  Diagram D extends the crossover at turnout 7.  This diagram also adds two dwarf signals (4LB and 10R).   The final diagram F has the lower trackage replaced with a “dark” yard to show signal 4R changing to restricting on allowing the train to routed into that yard. 

So let us examine what happens as we set the panel in FIGURE 5 to the state shown in Example A in the diagrams on Page 56.  This can be seen in FIGURE 6 below.    Here are some of the salient features.  First of all, the track segments change to red in order to show occupancy.   The west-most red segment corresponds to Train A in the RMC Figure 1A.   On a modern CRT CTC panel, cleared routes are show in green with arrows in the direction of travel.  Note that as routes are opened, some of the previously white signals now show yellow.   This indicates a signal less restrictive than red but does not show the actual state of the three heads.  Also, some of the previously white signals are red.  This is because, I think Dr. Chubb intended the track segments shown in Figure 1A to be under APB discipline and signals opposite to the direction of travel would “tumble down” to protect against opposing direction head-on collisions.   Refer to FIGURE 1 above, as this defines the signal aspect template for all the three headed signals, 4R, 4LB, 8L & 8LB.  A similar signal template exists for two headed searchlights defined for signals 245, 246, 4L, 8R, 4LC and 275.   Note that Train A in west-most block is cleared easterly to signal 8R.

Finally, let’s compare the state of the key signals in this example, signals 246, 4R, 8L and 275.  We can see this by examining the settings in the JRMI signal head table and those signals that have most recently changed state can be seen in the text at the bottom of the system monitor window.  According Dr. Chubb, signal 246 should be yellow over green [Approach Medium/rule 282], signal 4R should be yellow over red over red [Approach/rule 285], signal 8L should be red over yellow over red [Medium Approach/rule 286] and finally signal 275 should be yellow over red [Approach/rule 285].   All of these can be verified as correct from the signal head table and the final three lines in the system monitor show the recent changes for the heads on signal 8L.  These are in the form:

  • 8L HEAD 1 indication is ARA 286 – Medium Approach color is red
  • 8L HEAD 2 indication is ARA 286 – Medium Approach color is yellow
  • 8L HEAD 3 indication is ARA 286 – Medium Approach color is red

FIGURE 6.   Example A from Dr. Chubb’s Article, Page 56 Feb RMC 

All that is necessary to test examples B and C is for the dispatcher to alter the routes that are cleared across the track patterns to match Dr. Chubb’s diagrams.  All three do match correctly with the RMC article.  To handle examples d, e and f, the track diagram had to be altered.  Example f replaces the southeast-most track with a yard that is dark.  That is, its occupancy is not detected.   This shows up in CATS as grey as opposed to white track on dispatcher panel. 

Example f has is the one case where a work-around was necessary within CATS.  The current CATS version does not set signals to rule 290/restricting upon movement into dark territory that are also yards.  Rodney Black, the author of CATS, knows this and it will be rectified in an upcoming release.  For now creating a special signal type and associated signal template that I called a SA-3 Yard type is the work around.  This signal template sets Rule 288 to same pattern red over red over yellow as used by rule 290. 

This signal type is placed only at the entrance into dark territory.  With this nuance, example f shows that routing into the yard will occur by signal 4R showing red over red over yellow [restricting/rule 290].   However, signal 246 in approach of signal 4R is also affected by this and displays the aspect for being in approach of slow approach slow rather than approach.  This is a limitation to be sure but there are plans to address it in future CATS releases.  To me, this is a small price to pay for the ease of implementing signal aspect definitions using the signal aspect template.   Finally, the aspects for signal 4LC on setting the route to exit the yard are either Slow Clear/Rule 287 or Slow Approach/Rule 288 as described in the RMC article.

Examples d & e both require adding a signal temple for two headed dwarf signals and placing them as signal 4LB and signal 8RB. As well, in examples d, e and f, signal 245 changes from a two headed to a single headed signal.

All of four of these CATS XML files are available on request.  Additionally a fifth file for examples A, B & C with the intermediate signals made non-visible on the panel is also available. The appearance of that dispatcher panel is shown below as Figure 7.

FIGURE 7  CTC Panel Displaying Dispatcher Controlled Signals Only



Leave a comment

Please note, comments must be approved before they are published