Block Detection and Occupancy Questions

[ianly]

I’m in the process of constructing a model railway in a dedicated 12’ by 10’ room. Locos and rolling stock are a mix of passenger and freight that has recently or currently runs on UK mainline and branch lines. The track plan, albeit compressed, is based on a real location, and I would prefer not to alter it too much.

 

My intention is that the layout be computer controlled with iTrain being the software that I’m considering at present. The command station is a Z21 (Black), turnouts are DCCconcepts IP digital with Yamorc (current sensing) occupancy detection.

 

I hope to commence track laying shortly, but before doing so, would like to be sure that I’m heading in the right direction when it comes to splitting the layout into blocks. To this end, I’ve attached a track plan showing my current thinking. The track plan has been annotated to show train directions, storage area, station area, etc. I’ve also coloured the tracks to represent the blocks. Not ever having done anything like this before, it’s my best guess based on the information I’ve gathered here in the forum and YouTube.

 

 

The ‘Hidden Storage’ area is split into 3 groups of 4 tracks, with the inner 4 holding trains going anti-clockwise, the middle 4 holding bi-directional trains, and the outer 4 tracks holding trains going clockwise. All of the tracks in this area can hold my longest trains. Likewise, all of the tracks coloured red, green, purple, yellow, orange and brown are long enough to hold the trains that I intend to run on them. The pink coloured track is borderline. Unfortunately, the blue coloured tracks are not long enough to hold my longest trains.

 

None of the turnouts or short lengths of track between them are included in blocks.

 

Operationally, some passenger trains will stop at the mainline station platforms while others and freight trains will pass through. Shorter trains (2 car DMUs) can only exit the branch line using the turnouts on the right side of the track plan. However, trains can enter the branch line using the turnouts on either the left or right side of the track plan.

 

Now, a few questions.

 

1.     Is what I’ve drawn workable or what changes to the block arrangements would be needed to make it so?

2.     Can the issues that make the plan unworkable be solved in software rather than altering the block arrangements?

3.     Would it be beneficial to provide 2 sensors for each of the storage and station blocks?

 

 

[Ray H]

I'd have thought that you might want to have some kind of detection in the area of the fiddle yard points otherwise you won't know for certain that a train has moved clear of the pointwork it has just traversed. Likewise, would you not want something similar to confirm that an arriving train hasn't stopped foul of the point at the far end of the siding it is using.

 

How many trains do you plan to have on the layout at any one time? Don't forget that if you want to reverse a train in the fiddle yard you'll need both a reversing siding clear in which to reverse the train as well as, presumably, the siding from which the train started and the siding where the train will end up having reversed. You could do this with just using the reversing sidings but that would limit the scope for reversing more than two other trains. Don't forget that one of the reversing sidings will probably be used for the branch train.

 

Will you get enough enjoyment out of what looks like two plain tracks (and bidirectional branch) through the station?

 

The crossover that links the branch to the anti clockwise circuit can presumably only be used by trains travelling in an anti-clockwise direction. Is that your intention? Is the crossover really necessary if not?

[RAF96]

The plan suggests to me that operations would centre upon a relay of different trains running the loops from the fiddle yard with the occasional branch line activity. Good for rotating stock and giving show front customers some interest, but operationally the whole thing would be a manual chore, but could be automated while the operator drinks tea, coffee, wine, beer, etc.

[RobinofLoxley]

The plan looks workable. If you put an oversize train on a block section it depends if you put resistors on the coach axles whether a control system knows that it's fouling the points. Or if you create conditions that don't allow that - the total length of the train is known compared to the block so it's possible to disallow oversize trains into certain blocks.

 

You only have one current detector per block. Accuracy improves for stopping etc when you create shorter blocks.

 

But I'm with @RAF96 on the layout it's very boring 

[ITG]

I don’t disagree with above comments about interest value of the suggested layout, but as I’m also at the start of using iTrain, I’d suggest adding feedbacks to the fans of turnouts at each end of the storage yard. Otherwise, there are long undetected sections, which may impact on the accuracy of trains stopping in the storage roads. Additionally, it would make sense to have at least two feedbacks in each of the storage roads.

Ian

[RFS]

10 hours ago, RobinofLoxley said:

The plan looks workable. If you put an oversize train on a block section it depends if you put resistors on the coach axles whether a control system knows that it's fouling the points. Or if you create conditions that don't allow that - the total length of the train is known compared to the block so it's possible to disallow oversize trains into certain blocks.

 

You only have one current detector per block. Accuracy improves for stopping etc when you create shorter blocks.

 

But I'm with @RAF96 on the layout it's very boring 

 

3 hours ago, ITG said:

I don’t disagree with above comments about interest value of the suggested layout, but as I’m also at the start of using iTrain, I’d suggest adding feedbacks to the fans of turnouts at each end of the storage yard. Otherwise, there are long undetected sections, which may impact on the accuracy of trains stopping in the storage roads. Additionally, it would make sense to have at least two feedbacks in each of the storage roads.

Ian

 

Automation software such as iTrain and Traincontroller are sophisticated software products which are well capable of dealing with issues such as you describe. I have used TC for many years now and have never needed more than one detector per block. TC, like iTrain, uses dead reckoning to work out where in a block a train is and it also knows, via defined stop markers, where a train is required to stop, or when to release the preceding block if the train is not stopping in that block. It will also know from the train length whether the train fits in the block, and if not, will not release the previous block (including any turnouts in between) until the train moves clear. 

 

As far as feedbacks on turnouts and resistor wheelsets are concerned, these are not needed for normal operation. They are however useful for providing protection against malfunctions such as coupling breaks, or for trains unexpectedly stopping in the wrong location. How far you want to go in this regard is up to you. I have resistor wheelsets on the last vehicle of every train, and I have provided detection for the point fans in the storage yards as these are mostly out of sight. I have not bothered with detection on turnouts in the scenic side as it would be very complicated to do so. 

Edited July 23, 2023 by RFS

[ianly]

Thanks for the information regarding the inclusion of feedbacks for the turnouts associated with the storage lanes. All turnouts will have their own pair of feeds from the track bus in any case. However, I'm not clear on whether each needs to have its own independent detector or that they can be grouped with a single detector as denoted in attached diagram of the turnout fan. That is, each of the turnouts in the blue group can share a single detector, etc.

 

 

 

 

 

[ITG]

Just now, ianly said:

Thanks for the information regarding the inclusion of feedbacks for the turnouts associated with the storage lanes. All turnouts will have their own pair of feeds from the track bus in any case. However, I'm not clear on whether each needs to have its own independent detector or that they can be grouped with a single detector as denoted in attached diagram of the turnout fan. That is, each of the turnouts in the blue group can share a single detector, etc.

 

 

 

 

 

Following guidance on the iTrain tutorial videos, I’m planning on feedbacks on a fan of turnouts, all  connected to the same sensor. In contrast with sensors within a block, no section length is required. The current sensor thus indicates when a train has left the previous feedback (block). But I understand you can use a feedback in a single turnout, but it would be somewhat uneconomical to do so for, say 3 turnouts which are all part of the same fan. My own approach for your scenario would lead me to think of one feedback on the three blues, one on the three yellows (facing same way) and one on the three greens.

Again, I myself am still learning, but looking at the extract of your track plan, I think you could include the lengths of track leading into these fans within each overall feedback. But not the uppermost blue/green/yellow turnouts.

Ian

[RobinofLoxley]

Maybe off topic, firstly just one point about the existing plan. On the RHS the branch line would have access to the clockwise line there via a single slip and turnout.

 

But mainly, I said that the layout plan lacked interest. Even if you are very focussed on automation and your idea is to have trains running on a sequence, thats all the plan can do and its going to get very dull after not too long.

 

Although you say that you have a location in mind, the track plan itself is generic, I am assuming that the sense of place comes from cues about buildings and stock. So some alterations are feasible to the plan.

 

Just in case you hadnt considered it, I have attached an Anyrail doodle where I have simply run a single track down an incline, which reaches -18cm after one circuit at a steady -1.8% which is not overly ambitious. After it gets there you can consider a yard at that level, giving more space at operating level to develop.

 

Theres lots of unknowns here, for all I know you might be building on top of cabinets that make this impossible, but the thing is anything can be considered at the planning stage, later on alterations get harder.

[ianly]

15 hours ago, RobinofLoxley said:

 

Just in case you hadnt considered it, I have attached an Anyrail doodle where I have simply run a single track down an incline, which reaches -18cm after one circuit at a steady -1.8% which is not overly ambitious. After it gets there you can consider a yard at that level, giving more space at operating level to develop.

 

 

I appreciate the time you've taken to put above together, and yes, I had considered using a gradient to a lower storage level but decided against. That being said, at some point in the future, I may use the 12' by 2' area directly above the storage lines for an end-to-end layout. This would be placed high enough not to foul or prevent easy access to the trains in the storage area below. The two layouts would not be linked.

 

At the risk of going even further off topic; the track plan that I included in my original post is a simplified diagrammatic version of one that had been worked up a few years back in the Layout & Track Design forum. Unfortunately, it along with others were lost when the forum moved. Therefore, I’ve included a copy below with my thinking of how blocks would be configured. The earlier plan is operationally very similar to above and is a loose representation of Castle Cary and the line to Froome at Blatchbridge. The latter only being included to give access/egress to a hidden DMU storage area.  More recently, I reworked the plan of the main storage area to provide dedicated bi-directional lines rather than having trains criss-crossing the storage lines. My thinking being that running the DMUs into one of the bi-directional storage lines would simplify operation, and eliminate the need for the turn-out combination shown on right side of the original plan. That may be a simplification too far though.

 

[5BarVT]

Morning,

Lots of interlinked thoughts going on here!  I’ll start with block detection comments and then move onto layout ‘issues’ but with an automation flavour.

1. Multiple sensors per block - it depends on how the block length compares to train length.  If you have a spare couple of feet there is no problem as you set the stopping point one foot back giving you one foot tolerance on stopping short or overrunning.  If you only have 3” spare, then that is probably a tighter tolerance than automation programmes can reliably achieve under absolutely any circumstance.

This can be affected by the stock you run - I have a Bachmann warship with less than perfect pickups that suffers from overrun much more than my fleet of Heljan locos.  Solution will be to sort the pickups!

 

2. You don’t want sensor sections within a block to be too short as it upsets the automation programme, though there should be ways to manage this in the software (there is in Traincontroller and I suspect that iTrain is equally good).  If you make provision for separate 12” section at the running out end of the block (thus both ends of bidi blocks) then that will give much better tolerance on stopping accuracy.

By make provision, I mean put the insulations in the track.  You can start off with one long section with both bits wired back to the same detector, but if you then find that stopping accuracy isn’t good enough you can add the separate detector without ripping up track to fit insulations.

 

3.  Lost vehicles: because of my big railway background I have put train detection through points, but it doesn’t do anything in the automation programme. (Waste of money? At this stage yes!). If you are running fixed rakes and the couplings are reliable then you don’t need detection through the points.  If couplings aren’t reliable, then one option is to put the effort into the couplings.

If you do want to use train detection to find lost vehicles then you need to extend the blue/yellow/green point sections further into the red/green/purple/orange block sections.  (I know one of the greens probably has a posh name but I’m male and I only do 4 bit colours!).

 

4 More lost vehicle: this is a bit technical.  Where you have crossovers, you can’t prove clearance properly between the two inside ends (yellow/blue and yellow/posh green joints).  You can do fancy stuff with conditions involving point positions if you really feel the need, but it might be easier to sort coupling reliability. (Choices, decisions!).

 

5. DMUs.  If you’re running short DMUs, consider splitting the bi-di block sections into two to get more in.

 

Layout questions

 

Not sure if your layout software give curve radius - the purple and orange blocks in the fiddle yard (especially the centre lines) look to be a very tight radius.  Removing the straight at the other end of the point might ease things a bit.

 

Comments above about ‘boring’:  for a manually operated layout I might well agree, but I think that with automation you will have fun programming sequences etc or just letting it run and see what comes out next to give sufficient interest.  Build the FY, put the other track in, get the automation going and see.  Don’t go to town on scenery etc until you find out the answer to the interest question.  There is room to put in a more complicated station layout later if you find the need.

 

Station limitations: I was going to suggest reversing the crossover at the branch end of the station so that you could loop trains while others pass on the main line, but if you’re copying Castle Cary, that changes the feel.

 

I would consider retaining the two branch sidings so that you have the option of one down followed by a different one back.  Automation will be good at managing that for you.

 

Hope these comments help, have fun, and you’ll find automation remarkably addictive (I did!).

Paul.

[Mike Buckner]

I'm intrigued by which make of track you plan on using, in particular, the turnouts.

[ianly]

6 hours ago, Mike Buckner said:

I'm intrigued by which make of track you plan on using, in particular, the turnouts.

 

Peco Code 75 Streamline

[ianly]

 

Thanks for your comments and suggestions, Paul. Much appreciated.

 

The answer to your question regarding the curve radius is below.

 

The software is Railmodeller Pro (Mac only). Unfortunately, freehand or point-to-point curves such as those linking the turnouts in the storage yard are not displayed by the software.  The two that you mention in the central lines are between 20 and 21 inch. The rest are approximately 24 inch. As you suggest, I can improve the two that are tight by reducing the length of the straights at each end to one inch, but I can't omit them altogether as the throw bars of adjacent turnouts clash. Inserting these short lengths of track is the only way I could think of to avoid adjacent turnouts clashing. Prior to the forum move and loss of images and photos I was often possible to find helpful examples in other layouts. Now?

 

 

 

 

 

[5BarVT]

2 hours ago, ianly said:

As you suggest, I can improve the two that are tight by reducing the length of the straights at each end to one inch,

I’d forgotten that.  I did the same (albeit with small radius straight) on my last layout.  I think 1” in each leg might assist as both pairs of sidings would be starting at a similar position.

Paul.