Model Train Forum banner

1 - 20 of 24 Posts

·
Registered
Joined
·
26 Posts
Discussion Starter #1
I want to install a operating signal system on my ho layout and wondered if any of you folks have had any experience with a particular brand good or bad.Thanks Bill
 

·
Registered
Joined
·
8,573 Posts
Train detection is the most difficult and expensive factor
for the operation of layout signaling systems.

Here is the Google response to detection system request:


One Florida N scale club seems to have hi techies as members who have designed their own
signal system. It apparently senses cufrrent usage. I have seen it...works smoothly...and is amazing to see. For most systems you would need to break up the tracks you want to signalize
into isolated sections just as does the real railroads.

Don
 

·
Registered
Joined
·
318 Posts
I would also love to do this. Walthers has one, as well as a few others with optical sensors. They get pricey though. There is also Arduino plans for one. Just Google Arduino crossing gate and you should get a few hits.

Sent from my Moto Z (2) using Tapatalk
 

·
Registered
Joined
·
2,891 Posts
As mentioned above, the signal system is an expensive addition to any model railroad depending on how extensive your layout is.
Some systems are a plug n play that only do the one function that they are designed to do, while others are software driven and give you endless options to experiment with and design your own signal system.
I'm using Train Controller Silver on my layout. It's made and sold by Railroad Company out of Germany. A little pricey but its a very good Windows based system with easy programming of the signals.
This video clip is in the early days of my layout and signal system. It shows how the signals can respond to the presence of trains in a block or the position of turnouts. Since this video I have changed my operations to dark lights when no trains are present. A pretty cool feature that wakes the signal up when a train comes into the approach block, then to red in the next block, yellow in the second to next block then temporary green then off.

 

·
Registered
Joined
·
940 Posts
I've got signals working on my model railroad. As others have said, it is expensive to implement. I planned from the start to have computer control and to drive working signals, so my railroad signals are fairly realistic. As I see it, you can operate signals in two main ways, they can be "pretty" and just be tied to the position of turnouts or they can be more realistic and also take into account the position of trains. The first way requires that you have some type of electrical switch tied to the turnout position that can switch between red, yellow, and green LEDs in the signals. If you are using Tortoise motors to drive your turnouts, they have built in switches that can be used for this purpose. The second way requires both turnout position sensors as well as sensors that can detect train positions. There are multiple ways to do this, but I only have experience with the mechanisms I'm using. Like D&J Railroad, I'm using TrainController software as well as Digitrax hardware to do the sensing and logic that drives the signals.
 

·
Registered
Joined
·
1,060 Posts
i agree with Mark, a functional vs "pretty" system requires recognizing both block detection and turnout position.

while commercial software is available (e.g. Freiwald Software), if it runs on your PC, you need someway to input the block detector and turnout positions into the PC. A common approach is the use of C/MRI-ish nodes that communicate over a common serial bus (e.g. rs-485) and a USB rs-485 adapter

C/MRI-ish board are available commercially, but arduinos can be used as well.

once you have computer control available, including turnout control, i believe JMRI supports control of the turnouts as well as both manual and automatic control of signals. in other words CTC

it seems that layouts with functional signal systems either have someone with a technical background to help or the owner has a lot of patience to buy, install and configure commercial products.
 

·
Registered
Joined
·
318 Posts
You talk about block detection and computer controlled software. It can be done with optical sensors and micro controllers like an arduino board or similar. I havn't done it yet, but I'm looking into it. Ken Stapleton had a block detection unit that looks like it will do lights, I'm taking to him about driving a servo for a gate or a sound board.

There are also some boards being sold on eBay both using block detection or IR sensors, no computer required. I'm not sure how well any of them work, but might be worth a try!

Sent from my Moto Z (2) using Tapatalk
 

·
Registered
Joined
·
1,060 Posts
you normally don't clear the signal until the block is entirely cleared. optical detectors are good for spotting cars

i built a system using pairs of optical detectors at each block boundary so that you can detect when a train is entering or exiting a block.

simple (half) block detector using arduino
 

·
Registered
Joined
·
318 Posts
That's what I was getting at. You don't really need a block detection if your using optical sensors. A sensor on either side of the crossing, at whatever distance you want, will turn them on and off. Block detection is more reliable for sensing locomotives, but not cars.

This is theoretical on my party, since I havn't used these devices yet. But I would think a week designed board with optical sensors is all you need.

Sent from my Moto Z (2) using Tapatalk
 

·
Registered
Joined
·
1,060 Posts
it's not theoretical for me because i've done what you're suggesting and understand it's limitations

if you choose to use optical sensors, you need a pair at the boundaries and more complicated software logic that makes assumptions on when the block is clear

current detectors require cars with resistors, is more reliable and easier to implement
 

·
Registered
Joined
·
1,371 Posts
That's what I was getting at. You don't really need a block detection if your using optical sensors. A sensor on either side of the crossing, at whatever distance you want, will turn them on and off. Block detection is more reliable for sensing locomotives, but not cars.

This is theoretical on my party, since I havn't used these devices yet. But I would think a week designed board with optical sensors is all you need.
vette-kid - I think you are talking about grade crossing signals, I think everyone else in this thread is talking about ABS/CTC/Interlocking train "traffic control" signals. Vastly different scale of complexity.

Although even a grade crossing can be more complicated than that too - you can do a simple approximation with a simple on/off trigger circuit using a pair of optical or IR (infrared) detectors, but this doesn't account for real world situations like:
  • crossing signals trigger well in advance of an approaching train, but stop as soon as the train is clear the other side (so need at least 4 detection points)
  • if a train triggers the approach signal, but then backs up to clear it (because they're switching something down the line) the crossing protection should stop
  • if a train triggers the approach circuit, but stops and doesn't actually reach the crossing, the signals usually "time out" and stop. Signals start again once the train actually hits the crossing "island" circuit (moving very slowly and protecting the crossing per operating rules). Of course if the train stops ON or across the crossing, the signals should keep going indefinitely until the train actually clears
  • a short train like a single car RDC or "doodlebug" passenger, or engines moving "light" or "van hop" will be short enough to be entirely between the sensors and not "seen"
A "trigger on with one sensor/trigger off with second" or "trigger on and simulate with a timer/controller" circuit setup will simulate a crossing to the 80-90% point for a mainline crossing not near anything, but if any "work" events happen anywhere near by (including passing sidings or passenger stations where trains stop for a period of time) it's not perfect. We've been discussing a solution for actually lighting up crossing signals at a grade crossing on the club layout (which is located smack in the middle of a passing siding so features basically all of the scenarios above, plus having to account for BOTH tracks) and the current suggested design involves a long detected electrical block and IR sensors on either side of the crossing on the main track, plus IR sensors on each side of the crossing on the siding and a bunch of wiring/controller logic based on sensor inputs. Of course, being a non-signalled line, this is pretty much our simplest case... doing crossing detection on other parts of the layout that feature ABS/CTC signalling with detected track blocks for that causes additional complications for overlaying grade crossing blocks on top of that. We're figuring out that part...

Also, block detection is definitely the most reliable, but to detect cars you need to have wheelsets with resistors on all rolling stock that doesn't already have power pickup (e.g. lighted passenger cars). Depending on the size of your layout and rolling stock fleet, this is a whole project in itself. Detection can be approximated with optical detectors at block boundaries, but this will fail to "see" standing equipment between the block ends.

Detecting blocks and turnout positions for ABS signalling is another topic, and then laying on route control for interlockings/CTC is more advanced on top of that and pretty much universally takes computer control.
 

·
Registered
Joined
·
1,371 Posts
it seems that layouts with functional signal systems either have someone with a technical background to help or the owner has a lot of patience to buy, install and configure commercial products.
Some technically ability to do the wiring and installation, etc. and also some RR knowledge of how the signals work and which "pretty colours" are displayed in which situations, and where and how to located and configure the signal arrangements...

(I've seen a lot of people on forums argue that signals have "the top head displays occupancy for the main track and the bottom head displays occupancy for the siding".... and while you certainly could design your own system under this logic, it's actually a gross over-simplification and actually not at all accurate in most cases when it comes down to it of how real world signal indications are read. Also that whole "different railroads have different rule books and sets of signal aspects/indications (coloured light combinations and what they mean)" thing.)
 

·
Registered
Joined
·
26 Posts
Discussion Starter #14
I want to install a operating signal system on my ho layout and wondered if any of you folks have had any experience with a particular brand good or bad.Thanks Bill
Thanks for all your replies, It would be very complicated time consuming to redo my layout to go with blocks as ATS so I've decided to go with Z-Stuff their signals have built in sensors and also carry stand alone sensors so I can install them before the crossings,etc. it won't be prototypical but it's the best way I can think of to go at this time. Thanks again for all your advice. Bill Oberst
 

·
Registered
Joined
·
318 Posts
vette-kid - I think you are talking about grade crossing signals, I think everyone else in this thread is talking about ABS/CTC/Interlocking train "traffic control" signals. Vastly different scale of complexity.

Although even a grade crossing can be more complicated than that too - you can do a simple approximation with a simple on/off trigger circuit using a pair of optical or IR (infrared) detectors, but this doesn't account for real world situations like:
  • crossing signals trigger well in advance of an approaching train, but stop as soon as the train is clear the other side (so need at least 4 detection points)
  • if a train triggers the approach signal, but then backs up to clear it (because they're switching something down the line) the crossing protection should stop
  • if a train triggers the approach circuit, but stops and doesn't actually reach the crossing, the signals usually "time out" and stop. Signals start again once the train actually hits the crossing "island" circuit (moving very slowly and protecting the crossing per operating rules). Of course if the train stops ON or across the crossing, the signals should keep going indefinitely until the train actually clears
  • a short train like a single car RDC or "doodlebug" passenger, or engines moving "light" or "van hop" will be short enough to be entirely between the sensors and not "seen"
A "trigger on with one sensor/trigger off with second" or "trigger on and simulate with a timer/controller" circuit setup will simulate a crossing to the 80-90% point for a mainline crossing not near anything, but if any "work" events happen anywhere near by (including passing sidings or passenger stations where trains stop for a period of time) it's not perfect. We've been discussing a solution for actually lighting up crossing signals at a grade crossing on the club layout (which is located smack in the middle of a passing siding so features basically all of the scenarios above, plus having to account for BOTH tracks) and the current suggested design involves a long detected electrical block and IR sensors on either side of the crossing on the main track, plus IR sensors on each side of the crossing on the siding and a bunch of wiring/controller logic based on sensor inputs. Of course, being a non-signalled line, this is pretty much our simplest case... doing crossing detection on other parts of the layout that feature ABS/CTC signalling with detected track blocks for that causes additional complications for overlaying grade crossing blocks on top of that. We're figuring out that part...

Also, block detection is definitely the most reliable, but to detect cars you need to have wheelsets with resistors on all rolling stock that doesn't already have power pickup (e.g. lighted passenger cars). Depending on the size of your layout and rolling stock fleet, this is a whole project in itself. Detection can be approximated with optical detectors at block boundaries, but this will fail to "see" standing equipment between the block ends.

Detecting blocks and turnout positions for ABS signalling is another topic, and then laying on route control for interlockings/CTC is more advanced on top of that and pretty much universally takes computer control.
You are correct,I was taking about grade crossings. My bad. Although the boards in talking about seem to be made primarily for train traffic signals, at least a few. I don't really know much about that, so I have bothered with it.

Sent from my Moto Z (2) using Tapatalk
 

·
Registered
Joined
·
6,031 Posts
Has anybody used the Atlas signaling system? I want to try it and wondered what users have to say about it.
 

·
Registered
Joined
·
250 Posts
I designed and built my signal system using 8 Azatrax TS2's and signals from N.J International. Spent over 1k and almost 2 weeks installing it. After all that work, there's nothing more satisfying than watching prototypically correct block signals function. In the future I may expand it to include turnout function for the sidings, but I never really do switching anyway :LOL:
 

·
Registered
Joined
·
6,031 Posts
A thousand bucks for signals? That's a lot of groceries. I hope I can do mine for less than that. Glad to hear that it works well for you though.
 

·
Registered
Joined
·
3,240 Posts
Be glad you aren't hanging catenary.
 

·
Registered
Joined
·
1,060 Posts
partially signalling your layout may be enough to provide the feel, especially in the more visible parts.

in other cases, a signal may serve a useful purpose, when a tunnel may be occupied or a turnout is not thrown and hard to see.

even with fully signaled layouts, the signal may not be easy to see by an operator in the aisle. large layouts may have signals shown on the fascia when necessary.

when considering the need to know the signal in the next block, along with various turnout positions (that may affect speed signals), signalling logic can be complicated both on the prototype (below) as well as on a model layout.

however, if you already have computerized control of layout turnouts along with block detection (e.g. CTC), signalling becomes less complicated.

545583
 
1 - 20 of 24 Posts
Top