Signal control systems?

[JRP]

I'm in a modular club and we want to install and operate searchlight signals on our boards.  We have the signals (BLMA 3 color LED single and double targets which operate on 2.2 volts).  I think we want to have a system that will incorporate infrared(?) controls that will automatically change the light color when the locomotive passes the signal, much like the prototypes.  We use all DCC.  Any recommendation on a control system?  Have you had personal experience with one of the dozen or so control systems that are out there?  Any suggestions are very appreciated including those systems we should stay away from.

JRP   

[7j43k]

Yes.

I'm in a modular "club", too:  Free-mo

 

We have an excellent signal system.  It's called MSS.

Do a search on it.  If you have questions, just ask.

 

There is nothing better.

 

Ed

[RR_Mel]

I built my own signal system using optical detection and a Arduino MEGA.  I doubt if anyone on the forum will acknowledge optical detection but it works great for me.   One advantage of optical is it will detect anything on the track at coupler height with no rail power applied.

 

The MEGA will drive 16 single head three color signals.  It is a simple truth table driver so it will work with any detection system.

 

Originally I built up a Rob Paisley system and it worked great for many years.  Early on I used Twin T detection on my DC layout and switched to Rob’s DCC detectors, they also worked great.

 

For my new Arduino system I went with IR optical detection mainly because I didn’t like having to put resistors on the axles of my rolling stock for current detection.  I use Vishay TSOP1738 and TSOP1756 IR receivers (about 55¢ each) and 100ma IR Emitters (about 20¢ each) modulated at 38KHz and 56KHz (the Emitters are running 50ma).

 

The Arduino MEGA (about $10) will except any switched low input to drive the signals.

 

I didn’t track the cost but I’m sure my total investment for 16 signals is under $100, including the signals.  I made the signals with K&S brass and cast epoxy signal bases, about $3.50 each including the three color LEDs but that’s another story.  Many hours building the signals.

 

Here is a link to my Arduino MEGA Sketch.

 

http://melvineperry.blogspot.com/2017/02/february-3-2017-arduino-mega-signal.html

 

 

Mel

 

Modeling the early to mid 1950s SP in HO scale since 1951

 

My Model Railroad   

http://melvineperry.blogspot.com/

 

Bakersfield, California

 

I'm beginning to realize that aging is not for wimps.

 

[JRP]

Ed, yes, thanks for this information.  I found MSS on site and will look into this. 

Regards,

John

 

 

 

[JRP]

Mel, this is very interesting and sounds like it has worked for you for years.  One of our members use to be an electronic enginner for Lockhead Martin (did aircraft wiring).  He does all our wiring now and I will pass this along to him.  I like the fact that you don't need track power with optical to sense movement at coupler height.  Plus not having to install resistors on your rolling stock is another great advantage.  Thanks for sending me this information. 

Regards,

John (with Green River Junction Modular Railroad)   

[RR_Mel]

John, If you haven used optical detection it isn’t a simple task but for me well worth the time invested.  Straight track is a piece of cake, curves on the other hand are complicated.  Several of my dual track mainlines took up to five beams for each curved track.  I didn’t use signaling in my yard, too complicated, just my mainline.

 

My first attempt was using the overhead lighting as the light source and regular photo sensors as vertical beams. They worked very good with the lights on but I do a lot of night time running.  I tried using the photo sensors for the long horizontal beams but had problems with crosstalk on the dual tracks.  Thus I went to IR beams, to stop the crosstalk I went with modulated IR detectors.  38K for the outside track and 56K for the inside track.  This worked very good, it took a lot of time and patience but it was worth the trouble.  You could use regular photo beams for one track and IR for an adjacent track to prevent crosstalk thus avoiding the need to modulate the IR beams.

 

I finished my project in May and it has never miss fired or failed in any way.

 

The only real problem was hiding the emitters and receivers.  I have emitters and receivers in trees, rocks buildings, signs and signal cabinets all over my layout.  The modulated IR beams easily go 30’ for straight shots, I have two 13’ beams on straight track.

 

You can also reverse the direction of the beams and use photo sensors on dual tracks.

 

You can use an Arduino to generate the 38K and 56K to drive a high current transistor to drive all the emitters from one source.  I’m driving a total 41 emitters at 50ma from one 5 volt power source using 40Ω ½ watt resistors in series with the emitters.

 

Mel

 

Modeling the early to mid 1950s SP in HO scale since 1951

 

My Model Railroad   

http://melvineperry.blogspot.com/

 

Bakersfield, California

 

I'm beginning to realize that aging is not for wimps.

 

[peahrens]

Of course, you will need to weigh the balance between prototypical operation and ease / complexity of installation.

In my case, this year I added Logic Rail Technologies "Signal Animators" and 6 Tomar 3-LED signals to my small layout.  The Signal Animator uses optical and/or IR sensors installed between the ties.  When a sensor is covered, the signal goes from green to red.  After the sensor(s) for that signal are uncovered, after 30 (or 10, depending on a jumper) seconds, the signal goes to yellow, then back to green after another delay.  This makes a fun, easy addition, though far from prototypical.  My layout is small.  This setup works reasonably as a feature, when running in the typical round the layout direction.  It is screwy when making the rounds in the reverse direction.  It was a happy medium for me.  

If I had a layout with much longer runs, I would have considered the "Block Animator".  It requires more sensors, typically four or more, per signal to handle the bi-directional aspects of entering and leaving blocks.  

I used the optical sensors and am quite happy with them.  They are simple to install, just drill a hole between the ties and insert the sensor.  My room has room light and fluorescent overhead floods, pretty normal room lighting, and I did not have to adjust the potentiometer on the circuit board.  The instructions point out that if desiring to operate as a night scene (or in a tunnel), the IR sensors (a pair per sensing spot) are needed.

I know some other Forum members have used the Logic Rail products, as reported in prior threads.  You might do a google search ("site: cs.trains xyz") to find prior threads as well as what you learn here.

https://www.logicrailtech.com/

 

[RR_Mel]

peahrens

Of course, you will need to weigh the balance between prototypical operation and ease / complexity of installation.

 

Paul is correct about ease / complexity of the installation of beam break type of detection.  It isn’t easy to do non straight track, it can be very time consuming as well as many beams to get complete track coverage.

 

I have one dual track long sweep S curve at the edge of my layout that was very hard to hide all the emitters and receivers for both tracks.  Because of the closeness of viewing the S curve it made it even harder to hide everything.  A non rivet-counter wouldn’t see the hidden detectors but even 3mm LEDs are very hard to hide from good eyes at 18”. 

 

Regular Photo sensors and regular IR LED detectors and emitters are between 2 & 5 mm in diameter, the Vishay TSOP1738 receivers are ⅜” x ½” x ¼”.  The biggies can be very difficult to hide.

 

Here is a tipical emitter hidden in a Woodland Scenics tree armature.

 

 

 

After the trunk is painted its hard to see the LED leads but to the good eye they can easily be seen at 18".  The leads are counter sunk in the trunk using a #11 blade.  The emitter ends up at coupler height.  I have 13 LEDs hidden in trees, most very hard to see.

 

EDIT:

 

Optical detection is great for detecting any object at coupler height blocking the beam over the track.  IR beam alignment can be touchy because you need a IR viewer to see the beam, a digital camera viewfinder can see a IR beam.

  

Mel

 

Modeling the early to mid 1950s SP in HO scale since 1951

 

My Model Railroad   

http://melvineperry.blogspot.com/

 

Bakersfield, California

 

I'm beginning to realize that aging is not for wimps.