Signal Complexity

This forum is dedicated to Riding Scale Railroading with propulsion using other than steam (Hydraulics, diesel engines, gas engines, electric motors, hybrid etc.)

Moderator: Harold_V

User avatar
Erskine Tramway
Posts: 390
Joined: Wed Sep 10, 2014 4:13 pm
Location: South Dakota
Contact:

Re: Signal Complexity

Post by Erskine Tramway »

That's how it happens. The guy on double track didn't stop for the red. I worried about stuff like that, particularly after the Company started using the 'cruise control', aka 'Trip Optimizer'. I never used it, I wasn't about to let a computer run my train. The cruise control isn't PTC, it just knows what the speed limits are, not what the signals say. It's real easy, especially running into the sun, for the Conductor to go to sleep, and the Engineer to be just awake enough to hit the alertor when it goes off. That's when signals get missed, and stuff like this happens.

Mike
Former Locomotive Engineer and Designer, Sandley Light Railway Equipment Works, Inc. and Riverside & Great Northern Railway 1962-77
BN RR Locomotive Engineer 1977-2014, Retired
ccvstmr
Posts: 2230
Joined: Sat Jan 04, 2003 10:37 am
Location: New Lenox, IL

Re: Signal Complexity

Post by ccvstmr »

What many hobbyists don't realize...automatic block wayside signals (and even cab signals) do NOT indicate the presence of another train! Signals are used to convey TRAIN SPEED...as defined by the RR operating rules. Where: Red=stop. Yellow=approach at less than full speed. Green=proceed at limited speed (as defined by the RR and operating rules). As an example...the standard 3 light signal can indicate another train or obstruction 2 blocks in advance. This is indicated by a yellow signal. When the signal is red...there MAY be another train in the next block. Either way...do no proceed beyond the signal...unless permission is granted.

Another example...there are no trains present (usually) when a turnout is lined for a diverging route...or a crossover lined to take the train from one track to an adjacent track. The signals provide an indication of one or more aspects that define what speed the train can safely proceed thru the turnout or crossover.

Manual block systems as used in the hobby are more like a double 3-way switch circuit used for stairways. There is one circuit for each direction of travel. Diodes or other electronics is usually not needed. When one circuit is latched...there is only one manual switch that can defeat the latch and open the circuit for an opposing train or train the follows. One of the shortcomings of this...if a train crew forgets to release the block...traffic that follows has to wait some time...before proceeding. They will not be happy!

Turnout point position detection...if you can derive a system where the detection switches do not hamper movement of the switch points...go for it. Mother Nature usually will have something to say about the reliability of such an arrangement. In particular when 2 switches should be used to detect either point position. To say the points are line straight and/or the points are lined not straight...is not good enough.

Electric switch machines can provide an ASSUMED point position when using motor aux. contacts (internal or external to the motor)...but only if all interconnecting hardware is intact. This works...but it's not failsafe. Here again...any debris stuck in the points may not result in a true position indication.

No doubt, there's a lot to consider when developing a signal system for block occupancy or turnout point position detection. Best word of advice...use the KISS principle. Carl B.
Life is like a sewer...what you get out of it depends on what you put into it!
I don't walk on water...I just learned where some of the stepping stones are!
I love mankind...it's some of the people I can't stand!
User avatar
ChuckHackett-844
Posts: 232
Joined: Wed May 03, 2017 3:54 pm
Location: Tampa, Florida

Re: Signal Complexity

Post by ChuckHackett-844 »

johnpenn74 wrote: Sat Oct 27, 2018 9:04 pm Can I elicit some opinions on the acceptable level of complexity of signals on a live steam railroad? I would like to hear any horror stories about making things to complicated or perhaps success stories where in application of "Prototypical" heads achieved the desired result without confounding the regulars and visitors.
....
My inclination at this time is to use two Face Route Signal (Not speed signals) and even throw in an occasional Lunar for good measure.

JP
This is a subject that I’m sure you will get lots of opinions on.

For the last 10 years I have been designing signal systems for ride-on railroads with components installed for 8 years and I have started selling system components to others interested in automatic signals. For more info see http://MiniRailSolutions.com

I do not come to this endeavor with a background in full-size railroad signals. I am an Electrical Engineer, Software Developer and Live Steam Locomotive owner. I come at railroad signals from the standpoint of implementing reliable automatic signals that support full-time, bi-directional, go-anywhere-any-time (lots of route choice turnouts) running. The more route variation the better.

Just as building a Challenger from strictly scaled down drawings will not result in a locomotive that will run on typical 7.5” gauge track (way too tight), in my experience, strictly scaling down full-size railroad signal practice will not work for our hobby. For one thing, on a full-size railroad, engineers attend hours of classes studying the signals for a particular road – and, if they violate the signals they get fired which we can’t very well do at a large 7.5” gauge meet.

It amazes me that guys who wouldn’t dream of going through a red traffic light will treat a red 7.x” railroad signal as a “suggestion”.

How you implement signals depends on the railroad and who is running on the railroad. If it’s a railroad usually run by the same folks and very few visitors and these folks know and “care about” full-size practice, you can adhere closely to full-size practice. If, on the other hand, you have (or want to promote) a lot of visiting engineers who are not familiar with the “rules of the road” you need to keep it as simple as possible:
  • Green -> Go
  • Yellow -> Go
  • Anything else (Red/dark) -> Stop – Do Not Pass
… period

You can add in "Distant Approach" (50% flashing yellow) but all 90% of them want to hear is "treat it the same as solid Yellow".

You start throwing in things like Permissive Signals, Green-over-Red, Lunar, etc. and you will have mass confusion on your hands at a meet.

The next big decision is Automatic or Manual. Manual systems are usually (not always ...) cheaper but they can cause big problems during a meet where there are a lot of visiting engineers (forgot to capture, forgot to release, etc.). For an example of this and why I started developing signal system components, see: "The Need For Automatic Signals For Ride-On Railroads" http://www.minirailsolutions.com/the-ne ... railroads/

Manual systems typically cannot handle cases of more complex intersections (multiple track intersections, etc.), prioritize Eastbound over Westbound or provide any traffic flow management to prevent deadlocks, etc. For an example of this, see: http://www.minirailsolutions.com/improv ... ing-a-wye/

Automatic APB systems: I hear lots of folks say that they want to implement APB signals "like the full-size" guys. Most people do not realize that APB signals do not prevent two trains from entering the same block (passing a head-end signal) at the same time - there are situations where a train might have to back up. See: "Why ABS/APB Signals are not Sufficient" http://www.minirailsolutions.com/why-ab ... ufficient/

Ok, with that being said, let me give some guidelines that I advise people to adhere to in terms of signals and their placement:
  • All signals are absolute – if it’s RED, you stop and stay
  • All signals are located at the block transition – i.e.: no “Displaced” signals if at all possible. Engineers are told to pull up to ‘But Not Past’ a given signal. If the system is automatic, passing the signal would cause a red signal that would never go green because he is already in the block, causing it to be "occupied".
  • All signals are located so that, if the locomotive is not past the signal, other trains may pass (i.e.: signal is located at the “Clearance Point”)
  • All signals are located on the right-hand side if track clearances allow.
  • If located at a convergence of two tracks the signals should be carefully aimed so as to minimize engineer confusion about which signal applies to him.
  • Avoid using signals directly over other signals on the same mast - i.e.: prototype practice of 'main route' over 'diverging route'. For multiple-head (route choice) signals it’s better to have the center, higher signal be for the ‘straight’ route and another, lower head on the same mast but off to the side in the direction the turnout will take him. This is an example of a triple-head route selection (my system, not my signal heads):
    Image
    In this case, the engineer has a choice of three routes, Left, Center, and Right (selected by route selector in the foreground - BTW: the system prevents throwing a switch under a train).
  • All passing sidings should be right-hand-running unless you are prepared to implement the signals and logic to properly handle bi-directional passing siding tracks. Right-hand-running reduces the number of signals required from 8 to 2 and reduces the complexity presented to a visiting engineer.
I think automatic signals add realism, are safer, are a joy to run with, and enable more trains going more directions (i.e.: "interest") on a given track layout while deconflicting train traffic.
Regards,

Chuck Hackett, UP Northern 844, Mich-Cal Shay #2
Owner, MiniRail Solutions, LLC, RR Signal Systems (http://www.MiniRailSolutions.com)
"By the work, One knows the workman"
User avatar
ChuckHackett-844
Posts: 232
Joined: Wed May 03, 2017 3:54 pm
Location: Tampa, Florida

Re: Signal Complexity

Post by ChuckHackett-844 »

Another advantage of a good automatic signal system is that it should "do the right thing" even when engineers "do the wrong thing" - like back out of a siding into the block they just exited or pass a Stop (Red) head-end signal, etc.

This adds a level of protection for others when someone (accidentally .. or on purposed) violates the rules - not a perfect situation but better than nothing when .. oops, I "thought" I hit the capture button!
Regards,

Chuck Hackett, UP Northern 844, Mich-Cal Shay #2
Owner, MiniRail Solutions, LLC, RR Signal Systems (http://www.MiniRailSolutions.com)
"By the work, One knows the workman"
User avatar
ChuckHackett-844
Posts: 232
Joined: Wed May 03, 2017 3:54 pm
Location: Tampa, Florida

Re: Signal Complexity

Post by ChuckHackett-844 »

ccvstmr wrote: Mon Oct 29, 2018 8:13 pm What many hobbyists don't realize...automatic block wayside signals (and even cab signals) do NOT indicate the presence of another train! Signals are used to convey TRAIN SPEED...as defined by the RR operating rules.
....
Carl B.
I'm afraid that 99% in the hobby think the exact opposite (which is your point). They equate the signals to traffic lights, Red = Stop, Green = Go, Yellow = speed up to beat the red! :D

... and, if you are expecting visiting engineers, you are not going to change that and the signal system must take that into account.

ccvstmr wrote: Mon Oct 29, 2018 8:13 pm ....
Manual block systems as used in the hobby are more like a double 3-way switch circuit used for stairways. There is one circuit for each direction of travel. Diodes or other electronics is usually not needed. When one circuit is latched...there is only one manual switch that can defeat the latch and open the circuit for an opposing train or train the follows. One of the shortcomings of this...if a train crew forgets to release the block...traffic that follows has to wait some time...before proceeding. They will not be happy!
....
Carl B.
Exactly why I promote Automatic over manual ... especially if you have/want a lot of visiting engineers.

Another advantage of some automatic signal systems is that it is possible to have one system mode for "the local guys" which presents signal aspects as close to prototypical as possible, and a second, simpler mode for when there are a lot of visiting engineers. This simpler version might consist of only Red/Yellow/Green or possibly even just Red/Green, it should be the railroad owner's/operator's choice - the system should not dictate the approach to be used.

ccvstmr wrote: Mon Oct 29, 2018 8:13 pm ....
Turnout point position detection...if you can derive a system where the detection switches do not hamper movement of the switch points...go for it. Mother Nature usually will have something to say about the reliability of such an arrangement. In particular when 2 switches should be used to detect either point position. To say the points are line straight and/or the points are lined not straight...is not good enough.

Electric switch machines can provide an ASSUMED point position when using motor aux. contacts (internal or external to the motor)...but only if all interconnecting hardware is intact. This works...but it's not failsafe. Here again...any debris stuck in the points may not result in a true position indication.
....
Carl B.
This is exactly why I am currently testing a turnout point detector that has the following requirements:
  • Must be independent of any point actuation method (usable on both manual and switch motor actuated turnouts).
  • Must detect “actual” point position … NOT “desired”/"commanded" point position (i.e.: must be physically attached to the points after any springs, etc.). This allows it to detect rocks, or other debris that is preventing the point from fully closing.
  • Must repeatably guarantee that the point is within 1/32” (or less) of the stock rail.
  • Must be field adjustable.
  • Must have absolute minimal “hysteresis” (i.e.: minimal movement from ON to OFF and back to ON) but be free from “chatter/hunting” (i.e.: ON/OFF/ON/OFF … due to vibration, etc.).
  • Must be reasonably waterproof.
  • Must not be susceptible to dirt, sand, etc. interfering with its operation.
  • Must be adaptable to any type of turnout construction.
  • Must require minimal force so as to not require the point actuator to use heavier springs, etc.
My solution consists of a single moving part. The device is bolted to the outside of the rail (one for each point) as close to the end of the point as reasonable (closer to the end allows more accuracy). A sensing rod extends from the sensor body through a hole in the stock rail. When the point closes it contacts the rod an pushes it towards the sensor resulting in the detection of the point closure. The sensor is accurate to 0.010".

The switch motors we were using contained micro-switches that were supposed to provide this information by they were unreliable and indicated the "commanded" point position, not the "actual" point position - the difference between a fun run or being on the ground ...

ccvstmr wrote: Mon Oct 29, 2018 8:13 pm ....
No doubt, there's a lot to consider when developing a signal system for block occupancy or turnout point position detection. Best word of advice...use the KISS principle. Carl B.
People have no idea what it takes to reliably make those Red and Green lights do what they are supposed to do! :D
Regards,

Chuck Hackett, UP Northern 844, Mich-Cal Shay #2
Owner, MiniRail Solutions, LLC, RR Signal Systems (http://www.MiniRailSolutions.com)
"By the work, One knows the workman"
rkcarguy
Posts: 1730
Joined: Tue Aug 22, 2017 10:33 am
Location: Wa State

Re: Signal Complexity

Post by rkcarguy »

I like your system Chuck, I think it would work great for a larger multi-user private layouts or clubs. Multi-directional done any other way would be a myriad of relays diodes and wiring amounting to a huge rats nest of mess and still subject to limitations as one circuit somehow has to have priority over the other.
My home RR is shaping up to be about 4 blocks, one passing siding, one spur, and a loco/rolling stock shed with 2-3 tracks inside, so I don't think I can justify nor need anything beyond a simple relay/diode matrix. I could see one guest or potentially kids/family operating a second train in the future, that's about it.
I am interested in your turnout indicator switch, and if it would work with my groovy track style 3/8x1 flat bar steel rails?
With my first turnout nearly done, it would be good to fit this soon while its up on the build bench.
User avatar
ChuckHackett-844
Posts: 232
Joined: Wed May 03, 2017 3:54 pm
Location: Tampa, Florida

Re: Signal Complexity

Post by ChuckHackett-844 »

rkcarguy wrote: Tue Oct 30, 2018 6:58 pm I like your system Chuck, I think it would work great for a larger multi-user private layouts or clubs. Multi-directional done any other way would be a myriad of relays diodes and wiring amounting to a huge rats nest of mess and still subject to limitations as one circuit somehow has to have priority over the other.
My home RR is shaping up to be about 4 blocks, one passing siding, one spur, and a loco/rolling stock shed with 2-3 tracks inside, so I don't think I can justify nor need anything beyond a simple relay/diode matrix. I could see one guest or potentially kids/family operating a second train in the future, that's about it.
....
Understand fully. As I point out on my website - All railroads have different needs.

Keep in mind that, if you start getting a significant number of relays in the system the cost can get up there. If so, contact me and I might be able to help. My stuff also supports "Button Blocks" (Capture/Release) for those that do not want to bond joints, etc. for fully automatic. Even though it's designed for large railroads the controller can also operate localized interlocks on a stand-alone basis.

Whatever the case good luck with your railroad and, above all, have fun with it!
rkcarguy wrote: Tue Oct 30, 2018 6:58 pm ....
I am interested in your turnout indicator switch, and if it would work with my groovy track style 3/8x1 flat bar steel rails?
With my first turnout nearly done, it would be good to fit this soon while its up on the build bench.
No problem with "groovy track". The switch point detector should work fine for it.

Next week I will hopefully be installing two test units at our railroad and one or two at another railroad. If those tests go as I expect I should be able to offer them for sale in maybe 30 days.

What kind of signal output are you looking for? I.e.: What would you be driving with it?
Regards,

Chuck Hackett, UP Northern 844, Mich-Cal Shay #2
Owner, MiniRail Solutions, LLC, RR Signal Systems (http://www.MiniRailSolutions.com)
"By the work, One knows the workman"
rkcarguy
Posts: 1730
Joined: Tue Aug 22, 2017 10:33 am
Location: Wa State

Re: Signal Complexity

Post by rkcarguy »

The relays I have are overkill, I bought a 5 pack of those 40 amp automotive ones with the plugs/pigtails it was ~$13. Using two of them and one diode per signal, I'm able to set up a signal with R/Y/G aspects. As for the detection circuit, I'm leaning towards these little relay gizmo's on amazon that have an adjustable on-off delay so I can filter out intermittent wheel contact and allow the train time to exit the block before clear is shown. I haven't purchased them yet because there are several different ones and I want to play with a few in wet and dry conditions before choosing. I think a 6 volt, 1 amp trigger circuit is going to be the magic number such that wet tracks won't trip the system, a train closing the circuit will reliably trip it, and yet it's a safe power level to have across the rails.

My signals are pretty power hungry, I'm using the super bright 80ma 10mm LED's. They are all resistorized to take 12 volts and it took 4ea 1/2watt resistors in parallel(at 4x the calculated value) on each LED to do it.

Plan is pretty simple, signals will face away from the somewhat centered passing siding on the RR, uphill traffic takes the main and will not stop. Downhill traffic takes the siding(default), and a R/Y dwarf signal governs the sidings exit back onto the main. The "firewood customer" spur (might be laid together with the storage shed tracks) will also have a dwarf governing entry onto the main. At the turnouts on the main line, two headed searchlight signals will display track condition over dark if the turnout is aligned for the main, and red over a lit yellow IF the turnout is thrown to diverging *and* IF the diverging route is clear. This was a rather unique/rare way portions of SP's routes were signaled through Oregon.
User avatar
ChuckHackett-844
Posts: 232
Joined: Wed May 03, 2017 3:54 pm
Location: Tampa, Florida

Re: Signal Complexity

Post by ChuckHackett-844 »

rkcarguy wrote: Tue Oct 30, 2018 11:09 pm The relays I have are overkill, .... I think a 6 volt, 1 amp trigger circuit is going to be the magic number such that wet tracks won't trip the system, a train closing the circuit will reliably trip it, and yet it's a safe power level to have across the rails.
....
Be aware that, depending on soil, ballast, tie and rail materials, large currents on the track circuit can cause accelerated corrosion. I use a track current of about 0.000025 A (0.025 ma) per foot. It sounds like ("1 amp trigger circuit") you are way above that. This is less of an issue in dry locations or situations where the signal system is not on for extended periods of time. It can be a very real problem in wooded locations that tend to be damp for extended periods of time.

The trick in a relay system is finding a relay that will reliably drop out after the passage of a train when the track is wet. Most relays have too much hysteresis** to work reliably. Another issue is the variation of relay characteristics in inexpensive relays. This variation can cause the system to stop working when a relay is replaced. I have helped people get their relay systems to work reliably. It is usually addressed on a case-by-case basis but it can be done and there are many successful relay systems in use.

(**difference between the coil current required for the relay to pull in .vs. the drop in coil current required to cause the relay to drop out)

In difficult situations, many people end up adding solid-state components to a relay (or replace it entirely) to achieve a much sharper, more reliable, detection characteristic. Once you do this you have to deal with lightning issues which I have overcome after about 7 years of learning how to protect solid-state devices installed outdoors (See: Living With Lightning http://www.minirailsolutions.com/living-with-lightning/).
rkcarguy wrote: Tue Oct 30, 2018 11:09 pm ....
My signals are pretty power hungry, I'm using the super bright 80ma 10mm LED's. They are all resistorized to take 12 volts and it took 4ea 1/2watt resistors in parallel(at 4x the calculated value) on each LED to do it.
....
I use 20 ma, 10 mm LEDs. The PWM LED driver I use can be set to supply 20 ma with no resistors to waste power. One installed system consists of 19 controllers controlling 60 signal heads with over 100 inputs (track circuits, switch sensors, etc.). The entire system only draws about 1A from a single 24v supply.

I assume power consumption is not an issue for you given that you have relatively few controlled blocks.
Regards,

Chuck Hackett, UP Northern 844, Mich-Cal Shay #2
Owner, MiniRail Solutions, LLC, RR Signal Systems (http://www.MiniRailSolutions.com)
"By the work, One knows the workman"
Berkman
Posts: 679
Joined: Wed Aug 16, 2017 7:55 pm

Re: Signal Complexity

Post by Berkman »

How does a automatic signal system handle multiple train sections within a block?
User avatar
Erskine Tramway
Posts: 390
Joined: Wed Sep 10, 2014 4:13 pm
Location: South Dakota
Contact:

Re: Signal Complexity

Post by Erskine Tramway »

Berkman wrote: Wed Oct 31, 2018 8:47 am How does a automatic signal system handle multiple train sections within a block?
If I understand you question correctly, the first train 'drops' the entering signal to Stop. Any following trains then 'Stop and Proceed' at the Red signal into the block behind it. On the 'Big Railroad', you Proceed at Restricted Speed past a Stop indication. That's defined as 'proceeding prepared to stop within half the range of vision, short of train, engine or other obstruction, looking out for switches not properly lined, not exceeding 20MPH (on the BN). That's why the Trainmasters usually set up their Red Flag 'signal tests' around blind curves or other obstructions, to see if you are actually travelling at Restricted Speed :shock: The trick was to stop as far back as you could, so they had to walk a ways to the motors to tell you you'd passed :lol:

You can put as many trains in a block as will fit, if you have to, as long as the following trains comply with the conditions in the Rule Book. The signal system doesn't know how many trains are in there.

Addition...A Centralized Traffic Control system computer screen will show all the trains that are on a given piece of track, but it doesn't know exactly where they are. All the CTC Control Operator (Dispatcher) knows is which side of 'center' (between the controlled signals at the ends) they are.

Mike
Last edited by Erskine Tramway on Wed Oct 31, 2018 10:40 am, edited 1 time in total.
Former Locomotive Engineer and Designer, Sandley Light Railway Equipment Works, Inc. and Riverside & Great Northern Railway 1962-77
BN RR Locomotive Engineer 1977-2014, Retired
rkcarguy
Posts: 1730
Joined: Tue Aug 22, 2017 10:33 am
Location: Wa State

Re: Signal Complexity

Post by rkcarguy »

My setup is in the woods, but I'm using non-treated lumber ties(will be soaked with a linseed oil and tar blend) and steel flat bar rails.
It's pretty wet here, so I'm planning on having the tracks up on a layer of 5/8" clear gravel so the ties aren't sitting in any water or mud.
This is going to be a 12" railroad, so my rails are further apart. I have a 6amp 12 volt power brick I'm planning to use to power the system.
I'm planning on the train grounding the detection relay, and yes the signal system is only going to be active when I'm on the RR.
I have 2, maybe 3 locations, where I will be able to get 110V power trackside.
Post Reply