Forgot to add two more sceanrios that Ed can explain for us using the "top signal main, bottom signal siding" hypothesis.
Beginning of two main track territory. Approaching from single track. You get a diverging approach. How does the bottom signal work if BOTH tracks are main tracks?
Junction between two subdivisions. Approaching from single track. You get a diverging approach. How does the bottom signal work if BOTH tracks are main tracks?
By the way, I know the answers to these questions, and all the above situations are valid. I want to see how the "top main, bottom siding" folks explain it.
Dave H. Painted side goes up. My website : wnbranch.com
dehusman 7j43k So. Tell me. What are the circumstances that would generate a yellow over yellow for the plain ole siding described above. Remember. YOU'RE the guy who is asserting it can happen. Not me. So........ Very simple. Lets use the 1980 BN signal rules in the CCOR. Yellow over Yellow. Rule 501A. Proceed prepared to stop at the second signal. Its probably territory with a 60-79 mph max speed. Both siding switches lined for main. Next signal does not display not stop. Signal after that displays stop. Both blocks to the stop are not occupied. That wasn't that hard was it?
7j43k So. Tell me. What are the circumstances that would generate a yellow over yellow for the plain ole siding described above. Remember. YOU'RE the guy who is asserting it can happen. Not me. So........
Very simple.
Lets use the 1980 BN signal rules in the CCOR.
Yellow over Yellow. Rule 501A. Proceed prepared to stop at the second signal.
Its probably territory with a 60-79 mph max speed. Both siding switches lined for main. Next signal does not display not stop. Signal after that displays stop. Both blocks to the stop are not occupied.
That wasn't that hard was it?
Interesting. for "my" siding, one could do the same thing with flashing yellow over red. Followed by yellow at the siding leaving-signal. Did BN not have flashing yellow in 1980?
Now I have a couple questions for you. You assert that the top signal indicates the condition of the main track and the bottom signal indicates the condition of the siding. On single main track, no siding for 10 miles, there is a two headed signal that displays yellow over yellow. If the bottom signal pertains to the siding, how does it display yellow over yellow if there is no siding?
Now I have a couple questions for you. You assert that the top signal indicates the condition of the main track and the bottom signal indicates the condition of the siding.
On single main track, no siding for 10 miles, there is a two headed signal that displays yellow over yellow. If the bottom signal pertains to the siding, how does it display yellow over yellow if there is no siding?
The discussion is only about a passing siding. So this doesn't apply. Sounds like it's speed signaling, though.
At a location there is a siding on both sides of the main track (east siding and west siding). The signal approaching one end displays diverging approach, red over yellow. To which siding does the signal apply and how does it apply when one siding switch is lined for the main and one is lined for the siding?
It applies to the route through the interlocking. I am presuming there are three signals at the other end of the interlocking, right?
At the end of a siding, where the track goes from a main and a siding to a single main, the LEAVING signals have two heads. Once again, if the lower signal head pertains to the siding, why is there a lower head when beyond the switch there is no siding, only single main?
Because there is, at this location, a need to give the engineer more information than a single head can. And, therefore, this is not the "simple siding" that I am referring to.
Ed
7j43kSo. Tell me. What are the circumstances that would generate a yellow over yellow for the plain ole siding described above. Remember. YOU'RE the guy who is asserting it can happen. Not me. So........
cv_acr It's been demonstrated that yellow over yellow is a real thing (and what it means, and when it would be used).
It's been demonstrated that yellow over yellow is a real thing (and what it means, and when it would be used).
And yet no one can demonstrate how to get a yellow over yellow on the entering signal to a siding with route-signaling.
You keep doubling down on each head being set independently based on main track or siding. So if each head of a signal has it's own independent input, based on your rule what conditions would drive a yellow over yellow?
You keep doubling down on each head being set independently based on main track or siding.
So if each head of a signal has it's own independent input, based on your rule what conditions would drive a yellow over yellow?
Based on my "rule", it can't happen. So, again, show me how it happens in the above circumstance.
Obviously everything is done with a number of different physical inputs, logic relays and drivers. But it's more involved than just setting heads separately based on track conditions. My point all along is just that there's more to it than "top light main, bottom light siding" and that doesn't account for the design of special approach signals, restricting signals, etc.
Obviously everything is done with a number of different physical inputs, logic relays and drivers. But it's more involved than just setting heads separately based on track conditions.
My point all along is just that there's more to it than "top light main, bottom light siding" and that doesn't account for the design of special approach signals, restricting signals, etc.
I am not talking about special approach signals. Nor restricting signals. Nor etc.
I am talking about a plain ole siding. Route signaled. All by itself. Out in nowhere. About as simple as it can be. With interlocking for each switch.
OF COURSE, if you make things more complex, there will have to be more complex signaling. With additional physical inputs. But that isn't what we're talking about. At least, I'm not. And haven't been.
So. Tell me. What are the circumstances that would generate a yellow over yellow for the plain ole siding described above.
Remember. YOU'RE the guy who is asserting it can happen. Not me. So........
I believe I no longer know what you're arguing.
In past posts in this thread you've stated:
1. The top head refers to the main track
2. The bottom head refers to the siding
3. That a yellow over yellow was likely "not possible" (since clearly you can't be lined into the main track and siding at the same time)
Eastern speed signalling is also a bit more involved than western route signalling, but really quite similar when it comes down to it, just with additional indications and logic to differentiate between allowable speeds to take through diverging routes, where the western route signalling just says "take diverging and know what speed you're supposed to take at that point". Otherwise the logic is all similar except you'll get slightly different versions of signal indications for "Slow Clear" vs. "Medium Clear" etc.
Chris van der Heide
My Algoma Central Railway Modeling Blog
cv_acr If the rules and tables are so irrelevant to the design of the system, how does your straightforward "physical reality" explain how a Yellow over Yellow aproach signal is generated in your system and what meaning to infer from it?
If the rules and tables are so irrelevant to the design of the system, how does your straightforward "physical reality" explain how a Yellow over Yellow aproach signal is generated in your system and what meaning to infer from it?
I didn't say the rules and tables were especially irrelevant to the design of the system. If they WERE irrelevant to the design, they would very likely also be irrelevant to the operation. And then, what point are they?
My straightforward "physical reality" explains yellow over yellow as being based on physical imputs. Based in reality. How do YOU think a yellow over yellow would be generated?
7j43k The table of signal indications isn't "how to intepret the information based on how the signal is showing the status of the different tracks", it's "this is the signal indication to use for this situation". I know that. Those are rules. I am talking about physical reality.
The table of signal indications isn't "how to intepret the information based on how the signal is showing the status of the different tracks", it's "this is the signal indication to use for this situation".
I know that. Those are rules. I am talking about physical reality.
cv_acr 7j43k For a route-signaled entrance to a passing siding, each head reveals the condition of its route. And does not reveal the condition of the other route (because it has its own head that does that). That's a physical description of operation. The point is that's not how it works, it only looks that way on a basic level if you only look at one signal example in isolation but it's not that simple. Your question isn't quite being ignored, we're trying to point out the premise is wrong.
7j43k For a route-signaled entrance to a passing siding, each head reveals the condition of its route. And does not reveal the condition of the other route (because it has its own head that does that). That's a physical description of operation.
For a route-signaled entrance to a passing siding, each head reveals the condition of its route. And does not reveal the condition of the other route (because it has its own head that does that).
That's a physical description of operation.
The point is that's not how it works, it only looks that way on a basic level if you only look at one signal example in isolation but it's not that simple. Your question isn't quite being ignored, we're trying to point out the premise is wrong.
Actually, some of them ARE being ignored. And I do not believe I have asserted the signal I am using as an example operates all by it's lonesome. I do assert that the ABS system will cause more restrictive indications, as necessary.
In a simple example of a signal at a siding where one track goes to two, and only looking at that one signal, and ignoring the rest of the signal system which is all tied together, it might sort of look that way, but it's actually more complicated than that.
See my comments about ABS, above.
You're attacking the problem from the wrong direction.
cv_acr ATLANTIC CENTRAL But yes, you are correct about how the individual signal head in a group is controlled, as it applies to interlocking signals and or CTC signals - not so much with ABS or other systems, which sometimes exist within CTC signals. They are talking about the meaning, you are talking about how they work. As I commented, many people do not know how they work, model or prototype. But no, that isn't how they work. The heads do not independently indicate the status of each track. The signal system evaluates all the conditions, determines what "meaning" needs to be displayed, and displays an appropriate aspect. As a whole. Across all heads of the signal.
ATLANTIC CENTRAL But yes, you are correct about how the individual signal head in a group is controlled, as it applies to interlocking signals and or CTC signals - not so much with ABS or other systems, which sometimes exist within CTC signals. They are talking about the meaning, you are talking about how they work. As I commented, many people do not know how they work, model or prototype.
But yes, you are correct about how the individual signal head in a group is controlled, as it applies to interlocking signals and or CTC signals - not so much with ABS or other systems, which sometimes exist within CTC signals.
They are talking about the meaning, you are talking about how they work. As I commented, many people do not know how they work, model or prototype.
But no, that isn't how they work.
The heads do not independently indicate the status of each track.
The signal system evaluates all the conditions, determines what "meaning" needs to be displayed, and displays an appropriate aspect. As a whole. Across all heads of the signal.
I know, each head responds to a logic chain, but those logic chains do "follow" the possible routes. Having designed and built my own relay based signal system, I have some experiance here, not to mention many years of industrial control experiance before computers and PLC's......
The only difference between what you are saying and what Ed is saying is that when the routes "interact" both can be in the same logic chain for a given signal head.
On that point you are correct.
But on a basic level, he does seem to understand where the information comes from for each signal head.
Sheldon
cv_acr 7j43k For a route-signaled entrance to a passing siding, each head reveals the condition of its route. And does not reveal the condition of the other route (because it has its own head that does that). That's a physical description of operation. The point is that's not how it works, it only looks that way on a basic level if you only look at one signal example in isolation but it's not that simple. Your question isn't quite being ignored, we're trying to point out the premise is wrong. The table of signal indications isn't "how to intepret the information based on how the signal is showing the status of the different tracks", it's "this is the signal indication to use for this situation". In a simple example of a signal at a siding where one track goes to two, and only looking at that one signal, and ignoring the rest of the signal system which is all tied together, it might sort of look that way, but it's actually more complicated than that. You're attacking the problem from the wrong direction.
Ed is not challenging the idea that you "read" the whole indication.
Fact is from a wiring standpoint, now that I understand his question, he is mostly correct.
In any given interlocking senerio, there is a logic chain for each signal head. That logic chain follows the condition of the respective route, in total.
So using my example in my first post, the upper head would look at dispatcher route selection, position of the crossover, position of the diverging route and detection along the route selected.
Possible routes not selected would default red. If the crossover is thrown to diverge, both top heads will defualt red, the main is not available.
The second head on the right will go green or yellow based on speed restrictions.
The third head would stay red, diverging route not selected.
Routes with detection would default red.
Available routes selected would go green or yellow based on speed restrictions.
7j43k I can't help but note that my questions are being ignored, directly or indirectly. Rather, I am being accused of ignorance. For a route-signaled entrance to a passing siding, each head reveals the condition of its route. And does not reveal the condition of the other route (because it has its own head that does that). That's a physical description of operation. IF you can show that statement to be untrue, partially or completely, I'd be very interested. Telling me that there are official rules about HOW to interpret that information is unnecessary. I've know that for a long time. Ed
I can't help but note that my questions are being ignored, directly or indirectly.
Rather, I am being accused of ignorance.
IF you can show that statement to be untrue, partially or completely, I'd be very interested.
Telling me that there are official rules about HOW to interpret that information is unnecessary. I've know that for a long time.
Ed,
I don't think they understand what you are saying. Honestly it took me a while to understand how you worded things.
PS: Sheldon. I agree that they don't understand what I am getting at. But I think my questions are very clear. So it's a puzzle how they're missing the point. Which I tried to make just now. Again.
cv_acr Let's try something else. Imagine a new completely made-up set of signal aspects where instead of extra signal heads, we introduce extra colours. So our sample indications/aspects are let's say: Clear - Green Approach (stop at next signal) - Yellow Stop - Red Diverging Clear (reduce speed through turnout(s)) - Purple Diverging Approach (reduce speed through turnout(s) and stop at next signal) - Blue In this system, if lined straight through you'll get a Green (instead of Green over Red) because the indication should be "Clear". If lined into the siding, you'll get a Blue (instead of Red over Yellow) because the indication should be "Diverging Approach". Apart from my made-up signal colours, this is exactly the same concept as how any other actual North American signal system works! You determine the meaning (indication) required, then display some combination of lights (aspect) to show that indication.
Let's try something else.
Imagine a new completely made-up set of signal aspects where instead of extra signal heads, we introduce extra colours.
So our sample indications/aspects are let's say:
Clear - Green
Approach (stop at next signal) - Yellow
Stop - Red
Diverging Clear (reduce speed through turnout(s)) - Purple
Diverging Approach (reduce speed through turnout(s) and stop at next signal) - Blue
In this system, if lined straight through you'll get a Green (instead of Green over Red) because the indication should be "Clear". If lined into the siding, you'll get a Blue (instead of Red over Yellow) because the indication should be "Diverging Approach".
Apart from my made-up signal colours, this is exactly the same concept as how any other actual North American signal system works! You determine the meaning (indication) required, then display some combination of lights (aspect) to show that indication.
I don't think you understand Ed's question/position on this.
Not completely sure I do either, but maybe I do?
But again, as Dave has pointed out, aspects, and signal location differ depending on CTC vs ABS, etc.
And, the lights may look the same, but rule books dictate different meanings/speed restrictions/etc, on different railroads, and in different eras.
I model the 1950's, I know way more about signals from the 1930's to the 50's than what I know about current practice - no flashing yellows in 1950.....etc.
7j43k Thus the aspects of each head reflect the condition (occupancy) of ONLY its respective track (siding or main) and all other tracks "outwards", including switch position.
Thus the aspects of each head reflect the condition (occupancy) of ONLY its respective track (siding or main) and all other tracks "outwards", including switch position.
Once again NO that's not how it works.
7j43k You seem to think otherwise.
You seem to think otherwise.
That's the first correct thing you've said.
A few more thoughts.
I have likely not followed this thread well enough to have really addressed any of the pending questions (or misunderstandings) in my post above.
That is kinda the point.
Without setting up a full set of critria, any conversation about signaling includes too many assumptions to get correct or thoughtful replies.
7j43k cv_acr 7j43k For a route-signaled two-headed signal at the entrance to a siding, how likely will the switch be set for the main if the signal is green over red? And how likely will it be set for diverging if the signal is red over yellow? 100% likely in both cases because Green over Red is the indication for "Clear" and Red over Yellow is the indication for "Diverging Approach". It is 100% unlikely that a Green over Red signal will be displayed if the switch is routed into the siding because that *must* display an indication with "Diverging" in the name. Likewise, the straight route will not display a "Diverging" indication because it's not diverging... None of this is because "one light is for the siding and one is for the main". So then, how is the aspect of the lower head generated? Is it especially related to the condition of the diverging track? Say, accessibility through the switch? Or occupancy? Does the condition of the through track generate information for this head? And how is the aspect of the upper head generated? Is it especially related to the condition of the through track? Say, accessibility through the switch? Or occupancy? Does the condition of the diverging track generate information for this head? Ed
cv_acr 7j43k For a route-signaled two-headed signal at the entrance to a siding, how likely will the switch be set for the main if the signal is green over red? And how likely will it be set for diverging if the signal is red over yellow? 100% likely in both cases because Green over Red is the indication for "Clear" and Red over Yellow is the indication for "Diverging Approach". It is 100% unlikely that a Green over Red signal will be displayed if the switch is routed into the siding because that *must* display an indication with "Diverging" in the name. Likewise, the straight route will not display a "Diverging" indication because it's not diverging... None of this is because "one light is for the siding and one is for the main".
7j43k For a route-signaled two-headed signal at the entrance to a siding, how likely will the switch be set for the main if the signal is green over red? And how likely will it be set for diverging if the signal is red over yellow?
For a route-signaled two-headed signal at the entrance to a siding, how likely will the switch be set for the main if the signal is green over red? And how likely will it be set for diverging if the signal is red over yellow?
100% likely in both cases because Green over Red is the indication for "Clear" and Red over Yellow is the indication for "Diverging Approach".
It is 100% unlikely that a Green over Red signal will be displayed if the switch is routed into the siding because that *must* display an indication with "Diverging" in the name. Likewise, the straight route will not display a "Diverging" indication because it's not diverging...
None of this is because "one light is for the siding and one is for the main".
So then, how is the aspect of the lower head generated? Is it especially related to the condition of the diverging track? Say, accessibility through the switch? Or occupancy? Does the condition of the through track generate information for this head?
And how is the aspect of the upper head generated? Is it especially related to the condition of the through track? Say, accessibility through the switch? Or occupancy? Does the condition of the diverging track generate information for this head?
How they are generated is not really the issue. The issue is that they must be read as one indication.
If you want to understand how they are generated, go back into some old MR issues and find Bruce Cubb's original relay based signal system or Ed Ravenscroft's MZL control and look at the relay circuits they used.
They both used circuits nearly identical to the prototype, slighly simplified, and adapted to our different situation of detection.
The problem with trying to answer your question, is that they are generated differently depending on the exact type of system (CTC, ABS, etc).
Today it seems modelers are not interested in the nitty gritty of signals, they just want some processor to spit out the right indication after looking at the inputs.
Fact is 80 years before computers or PLC's, relay logic did this job just fine - BUT you had to actually understand the whole thing.
As others have commented, prototype signaling is very complex, varries from region to region, railroad to railroad, and era to era.
Our models only need something that looks plausable.
I chose to just have interlocking signals, and their approach signals. And I am doing it with relays. It is amazingly simple once you break it down to its basic elements.
Example - a crossover moving right to left on double track with a diverging route to the left after the crossover - from the direction just described. Three heads on a signal bridge for the right track, two for the left.
Crossover set straight thru - no trains in next blocks - diverging route not selected - right track will be G/R/R, left track needs only two heads G/R
Train in next block sets top G to R for the effected tracks, train in block after next sets top to Y (I don't use this indication, in our model world, the way my layout is designed, we don't need you slowing down for a train half way around the layout, and you will see an approach signal that will be Y half way thru the next block if that is the case anyway).
Crossover is selected, diverging route not selected, no train in next block - right track will be R/Y (or G for a high speed crossover) /R, left track will be R/R (absolute stop)
Crossover and diverging route both selected, no train in diverging route block - right track will be R/Y/Y, left track still R/R.
In this application, CTC signals convey speed and route, and protect against occupancy ahead, but the rule book will define those speeds and stop rules.
So how are those aspects generated? By a relay logic chain thru the turnout positions, authority from the dispatcher (cab assignment of next block in my case), and detection status of next block.
These logic chains exist in both directions.
Signaling is just too complex for all these generalizations........
dehusman 7j43k So then, how is the aspect of the lower head generated? They are NOT generated independently. First off, we are talking about CTC or an interlocking since red over yellow or red over green aren't used outside those territories. That means a dispatcher is involved. The dispatcher lines the switches and requests a route. Based on the positions of the switches, the occupancy of the tracks on the selected route, the conditions of the next switches, the limits of the route selected, the system selects the most favorable indication for that set of conditions.
7j43k So then, how is the aspect of the lower head generated?
They are NOT generated independently.
First off, we are talking about CTC or an interlocking since red over yellow or red over green aren't used outside those territories. That means a dispatcher is involved. The dispatcher lines the switches and requests a route. Based on the positions of the switches, the occupancy of the tracks on the selected route, the conditions of the next switches, the limits of the route selected, the system selects the most favorable indication for that set of conditions.
(in the case of a route signaled passing siding)
I maintain that the aspect displayed on the lower head is not influenced by the occupancy of the main that is between switches. And the same holds for that main, with respect to the siding.
You seem to think otherwise. If you do not, then we are in agreement.
dehusmanIt doesn't figure out the top signal then figure out the bottom signal, it figures out the situation and displays ONE indication, that uses all the heads on the signal.
You need two heads to convey the indication. That's because there (at least under NORAC) are over a dozen possible signal aspects that need to be displayed in the field. You can't do that with one signal head with 3 colors of bulb. You need at least two signal heads to get a combination of lights to convey track condition/occupany.
It's been fun. But it isn't much fun anymore. Signing off for now.
The opinions expressed here represent my own and not those of my employer, any other railroad, company, or person.t fun any
7j43kSo then, how is the aspect of the lower head generated?
It doesn't figure out the top signal then figure out the bottom signal, it figures out the situation and displays ONE indication, that uses all the heads on the signal.
A signal indication is based on any and all of the following conditions:
- whether the dispatcher has actually cleared a route through the signal (CTC)
- whether an opposing route has been set up (CTC or APB)
- whether or not the switch is normal or diverging (if there is a switch at the location of the signal)
- whether the block beyond the signal is occupied or not (the one that the switch is routed into)
- the indication of the next signal down the line (*not* the following block(s)...)
Take *ALL* of this information into account, and figure out what indication needs to be displayed. ("Stop", "Clear", "Approach", "Diverging Clear", etc.) Then display that indication, based on the chart of signal aspects in your rulebook.
The Signal aspect will be Green over Red (for example) because that's the aspect in the rulebook for "Clear", not because the top and bottom heads independently reflect the status of different tracks.
7j43k So then, how is the aspect of the lower head generated? Is it especially related to the condition of the diverging track? Say, accessibility through the switch? Or occupancy? Does the condition of the through track generate information for this head? And how is the aspect of the upper head generated? Is it especially related to the condition of the through track? Say, accessibility through the switch? Or occupancy? Does the condition of the diverging track generate information for this head? Ed
Have you not listened to the 50 times we've said "no, that's not how things work"?
Gents, the aspects and indications of signals vary from railroad to railroad.
Here in the east (former Conrail lines), and perhaps down south as well, signal indications convey SPEED.
West of the Mississippi, I understand that signal indications can convey ROUTING as well.
Signals at interlockings will normally display "Stop Signal" UNLESS the dispatcher or operator has cleared them for a movement.
On the other hand, automatic block signals will usually display their LEAST RESTRICTIVE aspect (depending on current of traffic in CTC territory) until a train passes.
An interlocking signal displayed for a movement into yards will almost always be "restricting" (yellow on the bottom), because there aren't any track circuits in such places -- hence, a more favorable signal CAN'T be displayed.
On an interlocking signal (and also on the distant signals that govern the approach to interlockings) with 2 or 3 units, the speed is indicated by the location of the green or yellow aspect.
Green on top = clear
Green in the middle = medium or limited speed (flashing is limited)
Green on the bottom = slow clear (on a 3-unit signal).
It's the placement of the green or yellow that tells you what to do.
The "red" lights are just there to show the unit is working.
In either case the signal could simply be "Stop" if this is CTC track and the dispatcher has not set the signals to allow a movement through (or if the movement has been cleared in the opposite direction, even if that movement is still miles away and the blocks are clear...)
7j43k I still look forward to hearing about how the yellow over yellow indication comes about. Ed
I still look forward to hearing about how the yellow over yellow indication comes about.
Try this on for size:
(Only eastbound signals are shown for clarity)
Copy and past link if image doesn't work: http://vanderheide.ca/files/route_signals.png
A train is travelling west to east on double track track, and is lined up by the dispatcher to be routed through crossovers around a stopped train, or MOW work block, or just because...
A block away from the crossover, which is already lined and cleared by the CTC dispatcher, the train sees an "Approach Diverging/Approach Medium" signal showing it it needs to slow to take the diverging route at the next signal.
(Bonus signal, just for fun) At the crossovers, the signal "Diverging Approach Diverging" because the train will be again taking a diverging route at the next set of crossovers which are at the next set of signals down, no intermediates between here. (All other conflicting routes not lined into the interlocking show "Stop").
At the final set of crossovers the signals show "Diverging Clear" as it will be straight sailing from here on at least for the next block or so.
If the train were to simply go straight at the last set of crossovers, the overall signal progression would be "Approach Diverging" (Yellow/Yellow), "Diverging Clear" (Red/Green/Red or Red/Red/Green on this signal), "Clear" (Green/Red) instead of "Approach Diverging", "Diverging Approach Diverging", "Diverging Clear"...
If the signals at the last set of crossovers aren't cleared by the dispatcher yet, they'd be at stop, so the signal progression would be "Approach Diverging" (Yellow/Yellow), "Diverging Aproach" (Red/Yellow/Red or Red/Red/Yellow), "Stop" (Red/Red). The last signal would remain at stop until the dispatcher clears it, giving either "Clear" (Green/Red) or "Diverging Clear" (Red/Green) depending on whether the route is lined through the crossover or not.
Indications based on this chart of BNSF signals: http://utu199.progressthroughunity.org/documents/Signals
And then there's this:
I am well aware that an engineer is supposed to do what the signals tell him to do, according to the rules of the railroad.
I am talking about how those indications arise.
In the 1968 Uniform Code of Operating rules, the yellow over yellow is rule 282, advance approach, proceed reducing to 50 mph before reaching the next signal.
In the 1980 Consolidated Code of Rules, BN signal rules, the yellow over yellow is rule 501A, advance approach, proceed prepared to stop at the second signal.
In the 2003 UP General Code of Operating Rules, the yellow over yellow is a version of rule 9.2.5, Approach diverging, proceed prepared to advance on diverging route at next signal at prescribed speed through the turnout. Not to be confused with Diverging Approach, red over yellow which is proceed on diverging route at prescribed speed through the turnout, immediately reducing speed to 30 mph (passenger trains 45 mph).
In all of these cases the signal means begin to slow down and be prepared to do something at the next signal. Its a signal used to give additional stopping/slowing distance.