Passing sidings, signals and cornfield meets

ecco

New member
I have a large Route (not my original design) with a lot of single track and a bunch of passing sidings. As far as I can determine everything is signaled correctly. However, I still end up with trains stuck at both ends and in the middle of passing sidings.

I set up a test bed.
QTySC8x.jpg
[/URL][/IMG]
Four passing sidings. Identical locos in the corners. TM's also in the corners. Driver commands for two locos to Drive To TMs clockwise and two locos counterclockwise.

The AI plays nicely (40 minutes) and the locos wait in the passing sidings until the approaching loco has entered from the opposite end.

But this doesn't really test the "real world" conditions of my big layout. The monkey wrench I threw in was a speedster Driving Via the trackmarks.

Within five minutes I had this...
Leu1ucx.jpg
[/URL][/IMG]

Two locos in the passing siding section. One incoming loco on either end blocking their exit.

Each passing siding has a Signal USA1 02 L at each entrance and Signal USA 04-3A's at the exits.


My question is this. Can something like this (random multi train variable meets) be controlled with just passing sidings and signals, or do I need to go to something else?

Thanks
 
Place a signal at the end of the loop before it becomes single. Don't place signal before a junction. Remember a train looks in the next block between signals to see if it is clear, it does not see a train in the block beyond the next signal.
 
Just as a suggestion.

passingloop.jpg


No approach signals to the loops - a consists right of way or path on the single track sections takes them to the signal at the end of the loop.
Use triggers and the Set Junction rule (or ITs/EITs) to lock and unlock the junctions at the ends of the loops.
 
If you're building your own route, my suggestion is: Don't use the extremely basic signals that have been around since 2004... A couple of signal systems have been written with their own logic controls that provide better (albeit not always perfect) protection against this sort of thing. The old-skool signal logic I think only looked a block ahead. If I'm not mistaken, both JointedRail and the RRS Supersignals looks further ahead to try to determine if it needs to hold up, but like I said, I don't think any of them are completely perfect.

Another option, using the EITs, might be to make sure the appropriate paths cover the distance between these sidings so that an opposing train cannot access the single track until the approaching train has cleared the entire single track section.


You said this wasn't your original route, so it might not make sense for you to go through and try to replace them all, so the only option might to be keep an eye on things and try to redirect problems before they occur.
 
Hi Eco.

To avoid any cornfield meets, the only solution is to authorize a train to enter a single line track only if you are sure there is one of the next passing place track which is currently unoccupied.

The easiest way for me to do this is to use EITs to secure access to the single line track between two successive passing places, and using two TCBs to protect the passing place tracks. The exit signal for the paths needs to be included in the TCB, so that you can use the Check exit signal TCB option in the EIT path to activate the path only if the single line track is available and the following TCB is unoccupied. Doing so your train will enter the path only if it is sure that it can access the next passing place track without any cornfield meets.

Hope this helps.
Regards.

Pierre.
 
Just as a suggestion.

passingloop.jpg


No approach signals to the loops - a consists right of way or path on the single track sections takes them to the signal at the end of the loop.
Use triggers and the Set Junction rule (or ITs/EITs) to lock and unlock the junctions at the ends of the loops.

As noted above, I did try eliminating the 02's - unsuccessfully.

I realize I can handle this with EITs. But I've got a lot of track and a lot of passing sidings.

That being said, I may have to bite the bullet and install EITs when sidings become problamatic - a few at a time.
 
If I'm not mistaken, both JointedRail and the RRS Supersignals looks further ahead to try to determine if it needs to hold up, but like I said, I don't think any of them are completely perfect.
Thanks for the suggestion.

I tried the JR Searchlight series and got identical results.
 
I moved the starting positions of some of the locos and it consistently hung up at the same siding in about 10 minutes
Ctclark1 suggested I try better signals.
I manually replaced all the Signal USA's with JR's Searchlights. It hung up in the same place in the same amount of time.
I manually re-replaced all the JR's Searchlights with Signal USA's. It ran for over 45 minutes before I stopped it!

The only thing difference between this time and the previous runs was a slight difference in the location of the signals since they were deleted and replaced manually, one at a time.

I'm running TANE 90945. I CDP ported it to Beta 94217. I started both within seconds of each other.
94217 hung up at ten minutes; 90945 ran over 1/2 hour before I stopped it. I tried it again with the same exact results, 94217 hung up, 90945 is running as I write this.

This is all really confusing.

Anyway, I'm going to go with EIT's controlling the single tracks - a long and slow replacement process.

Thanks again to all for your responses.
 
Hi ecco

I built a test route identical to yours and set a session up with EITs and TCBs as described by Pierre above. At first the session still hung at one particular loop and I couldn't find any reason for it as all paths etc were correct.

However I then adjusted the default directions of all junctions in Surveyor so that they were set from the single line section into the correct leg of the loop. For some reason this has cured the problem.

I even put 6 locos on i.e. 3 in each direction and it still worked perfectly, running for almost an hour without problem before I stopped it. As I was using Drive via Trackmark it may be possible that it was trying to set its own path and screwing up the path setting so really I need to run the session again using Autopilot instead of Drive via Trackmark but I haven't time at the moment. I'll have another go at it tomorrow.

EITs and Mission Codes are the way to go with this type of problem and once you get used to using them an EIT can be set up in a matter of a few minutes. Of course this will vary depending on the number of paths needed in an EIT - I have one EIT with over 70 paths.

Regards

Brian
 
All levers must have default setting into the loop. Use direction markers on the loop tracks. A down train will not change the next loop junction unless it is told to use the correct track, it will assume the up track is the way to go.
 
Hi ecco

I built a test route identical to yours and set a session up with EITs and TCBs as described by Pierre above. At first the session still hung at one particular loop and I couldn't find any reason for it as all paths etc were correct.

However I then adjusted the default directions of all junctions in Surveyor so that they were set from the single line section into the correct leg of the loop. For some reason this has cured the problem.

I even put 6 locos on i.e. 3 in each direction and it still worked perfectly, running for almost an hour without problem before I stopped it. As I was using Drive via Trackmark it may be possible that it was trying to set its own path and screwing up the path setting so really I need to run the session again using Autopilot instead of Drive via Trackmark but I haven't time at the moment. I'll have another go at it tomorrow.

EITs and Mission Codes are the way to go with this type of problem and once you get used to using them an EIT can be set up in a matter of a few minutes. Of course this will vary depending on the number of paths needed in an EIT - I have one EIT with over 70 paths.

Regards

Brian

I have no problem setting up EITs. However, I have not been able to wrap my head around TCBs. Is there any chance you could send me a copy of the CDPs?
 
I've been following this thread with great interest.

When the Interlocking Towers and then the Enhanced Interlocking Towers came out I experimented with them. Extensively. And from time to time have returned to see if I can solve the problem of AI and Player trains on a single track line with passing loops - the problem of the "Mexican Standoff" or the "Cornfield Meet".

This video highlights the problem:


The yellow locomotive is controlled by the Player, the grey by AI.

In the first part of the video (no Signal Link) the AI train takes control of the switch. The Player cannot access the passing loop. In the second part of the video (1:05 onward) Signal Link is configured and the Player retains control of the switch.

Try as I might I cannot achieve the latter using Interlocking Towers. To those who have experimented even more than me, can it be done?

Phil
 
Hi Phil

The key to a problem such as this is that a train entering the single line section needs to have control of the exit junction before it is allowed to enter the section. An Interlocking Tower would need to have two paths set up - one for the AI train which included both junctions and one for the player train in the opposite direction which also includes both junctions. Whichever train arrives first at the trigger point (about 1 Km from the junction) will set its path and the IT will prevent the other path from being set until the first train has cleared its path. The set up that you show may be complicated by the proximity of the portals to the first junction as that distance appears to be less than 1 Km in which case the path may need to be set to manual instead of automatic and then set by a driver command. The path for the AI train can be left on automatic.

If you prefer the path setting to be left as late as possible for the AI then set both paths to manual and have the AI train drive through a trackmark followed by a driver command to set the path.

It may be necessary to use mission codes for the path selection for the player train but as I have only used ITs with AI trains I have no experience with using player trains and AI together but perhaps someone who has can advise further.

There is another way of controlling single line sections with AI which involves using variables and conditional driver commands which I have had considerable success with, even being able to allow trains traveling in the same direction to follow one another into a long section with a number of signals on it but it is more complicated than using ITs.

Regards

Brian
 
I belive I read somewhere on one of the trainz sites, to put a signal before the junction, to control which way to send the engine, and another at the end where it rejoins the mainline, that is if it does. I have a lot of the routes set up this way and it works for the most part. If a lot of trains are running close to each other they can and do get jammed up waiting for the signal to turn.

IS the above a good way or not to have the signals set up? Thanks.
 
A signal before a loop junction will allow a train into the loop which may have to wait for a train from the other direction. While it waits the train behind could approach the junction signal and block the line for any train trying to leave the opposite loop.
 
Back
Top