Turnout Control in 2021

Digitrax DS78V released 09/28/2021.   Operates servos.   Looks like it will do everything the discontinued Tam Valley products would do.  I have Tam Valley,  I have local toggle turnout control, throttle turnout control and LED indicators. 

The Tam Valley personal situation is just a sad deal.  

C/MRI nodes ?

guessing this is not what you want to hear.

Bruce Chubb had a similar problem, being able to control a large layout from a central location and minimize the amount of wire as well as loss of sanity.    Intelligent nodes (e.g. Arduino) distributed around the layout control nearby devices with short lengths (10') of wire.   a single long communications bus (like the DCC controller UTP bus) connects the nodes to a central unit.

the central unit sends commands to the nodes and the node control outputs.   the node could also provide input, for example, turnout position, block detection or monitoring a button for local control of some device.   that input is sent back to the central controller and the controller sends back a command echoing the local request, or denying it.   the central unit could be a PC with a graphical user interface (GUI).   or it could just be panels with switches and leds

presumably JMRI can support the use of such nodes.    yes, there are several learning curves.   of course it would help if you could get someone to help (do you live near cumberland, md)?

arduinos could control tortoise machines with relays (that aren't so cheap).   Chubb also has a circuit for controlling Tortoise machines with op-amps

the benefit of C/MRI nodes is the not only solve a wiring problem, they also make other things possible (e.g signaling)

The PSR has a distributed set of node.    another layout i've been working on has Arduinos controlling interlocks

i believe the other option is lots of wire.   consider 8-conductor wire.  and it you want some feedback for turnout position, there's more than just a couple wires.   could use latching relays and multiple momentary buttons

Thoughts

One thing you didn't cover, although I gather from your leading contender is you do NOT want to DIY any/all it.

I would look into LCC products from Dick Bronson of RR circuits: a href="http://rr-circuits.com/description/index.html"> http://rr-circuits.com/description/index.html,although his website is confusing as to what is the recommended path these days, I'm going to post over on the LCC groups.io forum and ask for clarification.

A solution would be smails from circuitron ( http://www.circuitron.com/index_files/Smail.htm), and SMAIL translator adapter from Berrett Hill Shop. ( https://www.berretthillshop.com/store/products/smail-translator-adapter/), although I think that will get pricey fast, but check online retailers for less than circuitron MSRP.

If you already have tortoise machines, check out the turtle from tamvalley: https://www.tamvalleydepot.com/products/turtlestallmotordriver.html, and if you prefer the sleek touch toggle like design for activation they have:  https://www.berretthillshop.com/store/products/tam-valley-depot/

I have not used these products myself, just what I would research for an off the shelf solution, both Tam Valley Depot and Berrett Hill are MRH Sponsoring advertisers.

A DIY solution becomes more involved and I would use servos as a cheaper alternative to tortoise machines.

C/MRI

As far as I know with the C/MRI solution either the original and/or updated with arduinos, requires a central computer to handle logic i.e what happens when a certain input is read, a dispatchers panel (tower) with switches and leds is just another input/output node.

The I/O nodes are just that input (e.g. push buttons, block detection, toggle switchers) and outputs (e.g. turnout control, leds, incandescent bulbs) which as Greg said are connected with a communication buss (RS-485) to a central unit, but it has to be a PC either running C/MRI software or JMRI.

my way

Hi,

I can show you one extra simple way, that you never understand. As you have tortoise, it is good to use simple DCC decoder for tortoise: http://www.fucik.name/masinky/zelva/ Easy to plug, easy to configure.

As a controller (or throttle if you like the name), we are using "The Cube": http://www.fucik.name/masinky/TheCube/ Again extra simple controller for one turnout. You can have multiple of them configured for same turnout This small device can be easily installed at any place  you need to control turnout. Wiring is extra simple by using model servo extension cables.

On the top the system will automatically communicate with computer by nature and without any configuration.

LCC

As long as you don't want throttle control (which IMO is less than ideal anyway), LCC is the way to go. You can program any logic you want into the nodes themselves without a PC, and use it with JMRI/CATS for dispatching if desired.

The challenge with DCC-based systems is getting feedback from the turnout, currently the only plug and play solutions for that are LocoNet-based. If you have physical switches for every turnout or route on the fascia, then you don't need to control turnouts via DCC.

I believe your using lenz?

I believe your using lenz? which is xpressnet?  with that, im not sure if you want to go this advanced as it would use cheap amazon fire tablets as control panels but..

You could use digikeijs dr4018 switch decoders on your Tortoise and use a JMRI panel displayed on a tablet..

If you want fascia controls, use a dr4088rb-opto (connect thru xpressnet) to send the switch control feed back to jmri to switch the dr4018..  

I believe that should work.

It seems you have several

It seems you have several choices to make. You can go the old fashioned way using switch/relay/light logic which is cheap, simple and reliable or you can go the newest high tech way which is expensive, complicated, and probably reliable once you get it working., or somewhere in between. All my control panels are the first.

The simplest version that would satisfy your requirements break down into three separable components: a Tortoise at each turnout location; a control panel consisting of just momentary push buttons and LEDs at each desired control point; and a relay logic board at some central location. The wiring to each Tortoise (or paired Tortoises in the case of crossovers or double ended sidings) is just two wires for the motor. All position control and display is provided at the control panels. This should minimize the long runs and allows the use of 8 conductor Cat5 cable to serve common areas.

Each control panel consists of two pushbuttons, two LEDs and two resistors for each Tortoise. Wiring from each panel back to the central relay logic panel requires 4 wires per Tortoise or pair plus one supply line and one return for the panel. You can connect as many panels as desired as they are basically in parallel. Adding two more wires from each control panel to the relay panel (presumably near the main dispatcher’s tower) will allow the dispatcher to lock out control from selected facia panels while retaining the positional display on each panel.

The central relay logic board contains one DPDT dual coil latching relay for each Tortoise (or pair). The latching relays allow a single polarity 12V supply and multipoint control via pushbuttons. Turnout position and status are retained during power interruptions or shutdown. The miniature relays cost about $4 each, pretty cheap logic and memory. Route selection can be added but that gets into the diode matrix thing. It gets complicated fast but adds greatly to operational convenience.

The advantages of the old fashioned approach are low cost and simplicity. The disadvantages are work required to build and the fact that it is not easily modifiable once built. Route control, if implemented, cannot be changed with just software or address changes.  If the layout is fairly mature or well planned this should not be an issue.

This sounds interesting but I

This sounds interesting but I am having trouble figuring out how it works.   Are you using the relay to control the tortoise motor?  And then you use push buttons to trigger the relay?  Or what.  I guess I just am not following this.

As a follow up to clarify.  I have the tortoises (bought years ago for this layout or used on the previous layout). I started building this layout with my father who was much more knowledgeable about electronics and such but we never came to a true agreement on the turnout controls before he got sick then the layout sat for a few years while I took care of him.  So now I need to get moving again this time on my own.  Advantage is I only have to please myself, disadvantage is I know less then he did and I will have to do this myself.  And while I understand electricity (ie I can wire a house) my grasp of electronics is… basic.

It is a balancing act between easy of building/cost/knowledge/skill.  I can’t afford a huge amount of money per turnout as ultimately the layout will need a lot.  At the same time that argument holds true for time as well.  At this point in phase 1 of the layout I will need about 7 control points in Meadow Creek About 10 in Quinnimont and I would guess about 6 in Prince.  (These are based on memory not looking at the track plan) And represent the turnouts that are basically the turnouts of the interlocking plants and or the mainline. So for phase 1  turnouts.  The yard tracks and such i will control manually.  So for phase 1 I will need about 20-25 control points (turnouts)  if you assume $10 each that is 200 if you assume $20 it is 400 and so forth,  It adds up fast.

One thing I am dead set on is mot using momentary contacts to run the motors as no one ever holds them down to get the full through and if they do they are typically not designed for that so it often damages them.  If the run a relay or a logic board that is different.

FYI someday I hope to add signaling and I am wiring this layout in such a way that I can add the detection in without having to Rewire and this kind of future proof logic is desirable for the turnouts as well.  That said I have no intention of using routing.  As for changing the layout design,  it is not very likly as this layout is a pretty strict prototype design in that if the prototype had a turnout there so does this layout.  So I am not concurrence about changing the layout design unless I rebuild it from the ground up.

-Doug M

Schematic

Doug,

I think there is a simple and cost effective solution to you requirement. Its late tonight but tomorrow I will gin up a schematic for a single Tortoise with button control and LED position indication. If you don't anticipate route control its really simple and can be copied times N for however many turnouts you want to control and from however many locations. If I can post the graphic I think you will get the idea.

Charlie

Relay specs and source

I am looking for something along these lines as well.  I don't have the long distances to contend with but I would like to be able to control some Tortoise machines from several locations.  Can you post a link to the relays you have used?  

Thanks, David

dpdt switches vs latching relays

maybe i misunderstood the concern for "very long wires"

presumably the turnouts are a short distance from panels controlling them.   toggle  switches are the conventional way to control.

spdt switches can used if separate + and - supplies are used.

.   

however, dpdt switches can be wired in series to control a motor from multiple locations.   dpdt wiring would use a single power supply.   the drawback is the toggle switch position doesn't necessary indicate the position of the turnout.   LEDs can be wired in series to indicates machine position

spdt or dpdt latching relays can also be used to control a turnout from multiple locations.   a pair of momentary switches can be used to toggle the latching relay.   this is a more expensive approach requiring both switches and relays

so if nearby panels are within site of turnouts, wire lengths shouldn't be unconventionally long and presumably any remote turnout control switches are also within site.  no need for C/MRI

turnouts could certainly be out of site of a more centrally located panel requiring long wire runs, but wouldn't it also need wiring for block detection to monitor train positions

The long runs come from the

The long runs come from the Dispatchers panels to the layout,  the local control is relatively short.  But giving the tower operated located in the office control of the turnouts during operating sessions is going to take very long runs (up to 90 feet or so) 

If I just wanted local layout front control that is easy.  A couple leds and a toggle and away you go.  It is getting the cabin operated located elsewhere to be able to control the turnouts that is a problem.  

-Doug M

At this point,

a relatively accurate to-scale drawing of the layout, along with your envisaged locations for control (including multiple control points for the same locations) is in order, Doug.  Everyone has different criteria for "too-long", "simple", etc. etc.  so suggestions for solutions sans real data may not be useful for you.

I understand your $$ sensitivity.  The truth is, if there was a simple, cheap, easy solution out there.  There are less expensive solutions, which require more time, savvy, and effort from the user, and there are more expensive solutions, which still require some time, effort, and comprehension, particularly as you must figure out how to do what you want with their products.  Some suppliers are more helpful with walking you through applying their product, but with any commercial product you're vulnerable to the supplier disappearing from the marketplace - look at the problems some are facing as Tam Valley contract's it's product line as an example.

If you're up for it, you can do what you want with a lot of personal effort using Arduinos, some off-the-shelf hardware, some programming, and JMRI, but do not let anyone convince you this is the route to go if you're not both electronically minded, and programming-aware.  If you're not willing or able to roll-your-own, then you'll have to have a firm grasp on your needs(start with a diagram, then describe your use cases), then seek advice about implementing it with the commercial products available.

Blair

Schematic

As promised, here is a schematic for control of a single Tortoise. The use of a magnetically latched relay for each Tortoise allows a single polarity power supply and momentary pushbutton control from multiple points.

Any number of control panels may be connected in parallel. Also, two or more Tortoises may be connected in parallel for crossovers or double ended sidings as desired.

If the +12V supply line to any control panel is interrupted by a remote toggle switch at a central control site (tower) the control function of that panel is inhibited but the display function remains.

LEDs: Both, either or neither LED may be used depending on your preference. While it is common to suggest a 1K resistor in series with each LED for 12V operation I find them way too bright. The values in the table will provide a more useful illumination.

Some suggested parts and sources are:

https://www.digikey.com/en/products/detail/panasonic-electric-works/TQ2-L2-12V/251780

https://www.digikey.com/en/products/detail/e-switch/ps1024ablk/81742

https://www.digikey.com/en/products/detail/e-switch/ps1024ared/44577

https://www.amazon.com/EDGELEC-Diffused-Resistors-Included-Emitting/dp/B077X95F7C/ref=sr_1_18?dchild=1&keywords=5mm%2Bled&qid=1634218187&sr=8-18&th=1

The relay circuitry can be included on the control panel for single location applications or separated for multiple Tortoise or multiple location applications. Route control can be added but this increases complexity and adds the dreaded diode matrix.

Dispatcher control vs local control

You want for dispatcher and local control of the turnouts.  Do you want the dispatcher to be able to select wether local control works?

Digitrax DS78V

This is a cool piece! I hadnt seen it before, makes it easier then my 2 unit suggestion.

Honestly, with what your talking about I would spend the time figuring out how JMRI's Panelpro works for making a digital panel for the dispatcher.  Then use the ds78v to control the turnouts from the dispatch panel and from the fascia switch.  No need for 90' of wire.

Long Runs

Doug,

With the dispatcher tower located 90 feet away you face a dilemma. The relay approach requires 4 wires per Tortoise (or Tortoise pair) between the dispatcher and the local control panel. You might have to invest in a 1000 ft. roll of Cat5 riser cable.

https://www.amazon.com/1000ft-Unshielded-Twisted-Ethernet-trueCABLE/dp/B01JAVJNIO/ref=sr_1_3?crid=LFQUVGJ47RZO&dchild=1&keywords=cat+5+riser+cable&qid=1634228068&sr=8-3

This would theoretically accommodate 22 individual turnout sets located 90’ away. Messy but simple in concept. Digitrax makes some neat hardware but it’s really geared to the automation enthusiast. The afore mentioned DS78V only supports servos, not Tortoises. Digitrax does make a DS74 for Tortoises  but it costs $60+ for 4 Tortoises and you still need the buttons, lights and LocoNet cables. You’re pretty much stuck with building the control panels unless you go with computer control and JMRI. Again, aimed at the automation crowd. Digitrax is pushing stationary decoders to control turnouts but I don’t see the advantage of entering a 6-7 button sequence on a throttle to activate a turnout vs hitting a toggle or pushbutton on a panel.

I admit I’m a lone operator with a medium sized 22’x28’ layout but it has 82 turnouts, all controlled by some form of switch/relay/diode/wire logic. The control panels are on the facia and the dispatcher (me) has an adequate pair of sneakers.

Charlie

Hmm; Chuck,

The circuit as drawn could be misleading.  There is no -12V, only 0V.  This error is very common, particularly when people are used to dealing with battery systems.  Do NOT apply +12V and -12V (such as one might find in many common power supplies) to the circuit simultaneously; the tortoise will fail.  As for the resistances, YMMV.  I've not personally come across a green LED that gave any light at all with 220K in series; 220 ohms might be expected in a 5VDC circuit, while in a 12VDC circuit one might use 1K ohm, but it will depend on the particular LED you select.

As for the gent suggesting the DS78V, it is specifically for servos, not slo-mo Tortoises like the OP has.  If you follow the Digitrax groups.io forum, you'll note other limitations.

Blair  

As for the gent suggesting

 Your right, oops.   But its looks like the DS74 can.  to bad it only does 4 turnouts, but it still would work.

the dispatcher tower located 90 feet away you face a dilemma ?

  If cost is a problem you probably should consider another layout plan. A layout that large and complicated is gonna cost a lot in every stage of construction. Turnout control will be a minor expense compared to everything else it will require. Keep in mind that it's not just size of layout that determines cost, it's also the complication and the type of track, controls, scenery and details that will be needed. ......DaveB 

Power Supply Notation

Blair,

Re the +12/-12 notation: Technically you are  correct and I thought about it when I was drawing the diagram. However, I'm replying to a non electronics person and I didn't want to further confuse him with reference to 0 volts or ground. I tried to make repeated reference to single supply or single polarity supply but at some point you just have to go with something. As for the LED resistors, the current super bright LEDs operate on very low currents. All my control panel LEDs operate on less than 250 Microamps with the green lights at less than 50 Microamps. Take a look at the links in the previous post. They are really much improved from the very early LEDs I helped develop at Texas Instruments 50+ years ago .  I find a 1K resistor tends to light up the basement and definitely compromises nighttime ops.

The DS78V issue has been noted.

Charlie

So in brief to sum up so

So in brief to sum up so far…

It looks like we have not seen a significant. change in applicable technology.  

So we are facing either a hardwired example with numerous long runs of wire or we are looking at something like a DCC set up with stationary decoders. Depending on the decoder you may or may not get a local control option via switches.  But the remote option would be through a GUI on a computer of some sort.

A distant third option is to network a bunch of micro computers together…

As further answers.  I have a very large layout under construction,  The plan is frankly besides the point.  it is a model of the C&O at a time pre CTC when Cabins (Towers to most of you) controlled sections of the railroad.  Such as the interchange and passing tacks and crossover at Prince WV.  on this layout is 7 such. locations.  3 of. them are related to yards that will have yard masters that can play operator and align the mainline/passing tracks as needed.  The other 4 will need to be remotely operated.  And being as i doubt. i will find 4 people to play tower operator it means i will need to locate these 4 cabins in the same place so one person can do the work on all of them.  The only place to locate the panels in the basement is located far enough away that it will take 90’ of wire (as the wire is routed) to get from Dispatchers office to the  the location in question.  This is. other a lot of work and not inexpensive in wire alone.  Thus my looking into options.

As for locking down the turnout control,  I am not that worried about it for ops sessions but it would be nice.  I do however want a way to lock down the fascia mounted controls during open houses.  This layout is such that i cant see the whole thing.  And i have seen to many people throw a turnout (accidentally or otherwise) at an open house to trust that something out of sight wont get thrown and thus cause a train wreck,  This system (if all else fails) could be as simple as a concealed switch that disconnects the common on the turnout controls.  The layout is a double tracked mainline with reverse loops at both end so is for show purposes a big dogbone loop and in general not turnouts need to move during a show loop display.

Please note if you are on one far end of the walk ways and want to go to the farthest (walking distance) away it is about 200’ end to end as the walkways form a bit of a spiral.  So i nor my helper can be everywhere during an open house.  

And yess cost is a bit of an issue.  i note that some electronic. control units/stationary controllers go for 12-20 per motor controlled.  And if. that can save time and money on wire that is not horrible.  but if we need rto add in monitors and such and still have toi run almost as much wiring then what is the point?  
 

My biggest problem with more complicated (electronicly) systems is that. in 5-15 years when it goes wrong it will take me a while to figure out what the heck i did.  I already have a binder with about 30 pages just showing the basic layout wiring in detail so when i forget it i wont have to take it apart of something goes wrong,  On large layouts you have to document everything…

-Doug M

One item to look at, Doug, might be

the CPNode from MRCS.  Has some of the features you're seeking, and it's customizable because MRCS has made the Arduino source available.  Also has the benefit of being 'networkable'.  Not sure where MRCS is going in the field of new add-ons, but hey, you could ask.

Blair