Preparing for DCC: Block wiring with Kato Unitrack

[JanW]

This may be a very dumb question with a very straightforward answer:  
I am planning my new lay-out to be DCC ready. So I want all block wiring to be already in place but I intend to start running the trains simply using the standard Kato power pack.
I now realise that I will need a lot of wiring connected to the tracks as I will have some 24 blocks and 10 turnouts that are not in any block. Each block has 3 occupancy detection sections. This brings the wiring to 82 pairs of rail connections, some even to a single track or turnout. 

Kato sells the terminal Uni-joiner. However that does not allow to wire a single track or turn-out, always at least the two tracks connected by the uni-joiner. Moreover, I would need 82 of them! 

I am worried to simply solder wires to the tracks because I am worried to damage the plastic trackbed and the tracks may come loose from its sleepers when the plastic clips melt when soldering. 

 

Is there a clever way to simply 'clip' a wire to the Kato Unitrack ? How do you guys solve this problem? 

Edited October 4, 2017 by JanW
added the lay-out plan

[Yavianice]

First, welcome to the forum Jan!

What about buying a few of the unijoiners and soldering the wires to that?

 

I -love- your layout plan! It looks awesome. I do have some small comments about it:

 

- I'm not sure how the loop on the right (with the connection between the shinkansen and commuter track) will work. It seems to me you'd have very steep slopes or very low clearance.

- I'm not sure how the automatic road crossing is supposed to work for the station? How will you connect the sidings with the sensor track? 

- Having the double crossover in the station area will result in some trains scraping along the platform, when combined with KATO platforms such as 23-107/23-108

- KATO signalling tracks (20-650 and 20-650-1), which I see you are using in one of your pictures, are not compatible for use with DCC. 

- Having a corner, and then an opposing switch right after one another, may result in many derailments. It is recommended you have a straight between a curve and a switch.

 

Edited October 4, 2017 by Yavianice

[JanW]

Hi Yavianice,
Many thanks for your valuable comments. You can tell I am novice here!  Let me try to react to your observations:

• Yes, I can buy a few unijoiners, however they will still power at least two tracks. If the detection section is to be short (like in the station to get an accurate train stop) you need to install two short track sections. I was hoping to avoid that. 
• The commuter loop along the edge of the table will run 50mm lower than the station level. It will rise in the purple section at the right hand side to station level. So the clearance between this purple track and the connection loop between the Shinkansen loop and the commuter loop will be around 50mm. The track support will need to be thin plywood at the intersection point or so to allow for enough clearance. 
• About the road crossing: I actually tested that. The lay-out is intended to be used with left hand driving trains. There will be a detector track in both tracks before the turn-out leading to the dark blue track. The road crossing allows to connect two detectors for each direction.  It turns out to be a busy road crossing! 
• Good point about the double crossover. I will test that and redesign.
• Indeed, the signalling tracks are a bit primitive. The one under the viaduct does not even fit there (too cramped). I indeed need to rethink this. Any suggestions? 
• "Having a corner, and then an opposing switch right after one another, may result in many derailments":  While test-driving, I had no issues with derailment switches in and out of the commuter station. Nevertheless I will try to redesign to include a straight section (e.g. at least 62 mm).  In general, my Nozomi Shinkansen seems a lot more vulnerable to derailments than the commuter trains. 

[kvp]

For short straight sections, you can use the short straight terminal tracks, as they are pre wired in the middle. Other than that, rolling your own feeders using unijoiners is the most simple way. If you are disabling power routing on any #4 turnouts and go with block wiring in analog too, then it's possible to use the two bridging screws to feed power into a turnout alone while having isolating joiners on every side. For #6, there are power routing bridge points on the circuit boards, that could also be used as feeding points.

[SubwayHypes]

WHat program did you use to make those designs!  Beautiful work!

 

I really need to do something similar for my new layout, instead of graph paper sketches and trial and error.

[JanW]

On 10/6/2017 at 10:12 AM, SubwayHypes said:

WHat program did you use to make those designs!  Beautiful work!

 

I really need to do something similar for my new layout, instead of graph paper sketches and trial and error.

I use RailModeller Pro. It is easy to use has a large database with tracks from various manufacturers and allows 'layers'. However it only seems to be available for OSX (Apple). See http://www.railmodeller.com/home-railmodeller.html

[cteno4]

I second rail modeler, it's a very nice program and has been pretty steadily improved.

 

jeff

[velotrain]

"I am worried to simply solder wires to the tracks because I am worried to damage the plastic trackbed"

 

kvp provides solutions for both single track and turnouts, but beyond that I wouldn't be so worried about the soldering.  I don't have tons of experience, but have done a fair amount of soldering of power leads to flextrack, which does have a plastic tie base, and I suspect is far less sturdy overall than the Kato track.  I think it's almost impossible for the rail to come loose from the sleepers, although carelessness could disfigure some of them.  If you do tackle any soldering, drill a hole between ties on the outside of the rail, send the wire up through with just the end of it stripped and bent in an L shape.  Tin the wire and the outside of the rail, and a brief touch of the iron should take care of it.  It is important to use a thin tip for this, to provide maximum control over just what the iron touches.  No doubt there are multiple sites and videos addressing this on the web.

 

A few notes on the trackplan.  I notice you have very few and quite short storage tracks, so I gather your main interest is watching trains run.  Unless you're willing to manually remove and re-rail entire consists when you want to watch different trains, you might reconsider this aspect of your plan.  Given what you've plotted out, there really aren't many options other than possibly enlarging the yard in the top-left corner - possibly having the access switch coming off the yellow track at the top.  Actually - I had initially taken the double-track loop as the Shinkansen line, but see that it's at a lower elevation, so am confused.  The yard should exit whatever track is at ground level.

 

Another approach would be to have a siding at the edge of the layout that leads to a removable cassette, or perhaps a rolling cart with multiple storage tracks that could be positioned next to the layout in the manner of a traverser.

 

You might re-evaluate the location of the tunnel portals at the bottom of the layout.  I think having the double track line emerge from the tunnel just where the single track crosses overhead would look rather strange when built.  Is there some reason for bringing the upper line away from the "wall ?" here?  It causes an awkward visual situation with the single track curve crossing over the double track line.  Eliminating the kink in the red-olive track would solve this.

 

Lastly - one very minor thing.  You have a siding next to what I take to be a canal basin.  However, prototype railways (virtually) never have the main line take the curved route from an industrial siding switch unless there is some dictating circumstance, and I don't see that here.  The dark green track could have a short zig-zag before the canal bridge.  If this is an industrial track of some sort, it's the only one on the layout, but you could fit additional ones inside the end loops.

 

 

Edited October 8, 2017 by velotrain
fix typos

[JohnPatrick]

Hi all - very informative thread! I have some basic but related questions and I have been going bonkers trying to find direct answers to them. I hope you can help.

 

I have moved from Marklin 3-rail to N scale 2-rail, using Kato Unitrack and Kato switches. I became fairly good at occupancy detection (after many mistakes 🙂) using Marklin 3-rail and S88s, but with 2-rail, I'm a little lost. I'm trying to figure out what products are easiest/most reliable for occupancy detection and that work well with the ECOS, whether I'm even setting up the blocks in the best ways, and how to provide power to switches once I have isolated one of the rails.

 

As background, I use an ECOS and Digitrax DS64/74s switch controllers, which thus far have been simple to use and work well with the ECOS without requiring any additional equipment or converters. I've been using iTrain for some years.

 

This is how I'm creating blocks: I isolate one rail at both ends of a block and within the block I provide power to both rails. Switches are endpoints of blocks but never within a block. I will likely use at least 2 occupancy detectors within each block, so for my layout about 16 blocks, 32-40 occupancy detectors. I am hoping that the power draw of a locomotive will be "enough" for reliable occupancy detection and that I won't need special axles on rolling stock or other techniques. I don't plan on using an occupancy detector on switches, I never needed this using Marklin 3-rail and iTrain.

 

I'm all in favor of using the simplest/easiest/most reliable products even if they are a little more expensive. So, at the risk of asking you questions that have been asked before, here they are:

 

1. What product, easy/simple/works w ECOS, do you recommend for occupancy detection? I suspect the EcosDetector Standard is the most simple device to use since it's the same brand equipment and connects directly. ESU products seem to have a good reputation. I've considered the less expensive Digitrax BDL168 but haven't figured out exactly how to make it work with the ECOS, e.g., do I need additional products like the ESU L.Net and how expensive will that get?

 

2. With 2 rail Kato Unitrack, does it matter which rail is used for isolating blocks? I am so confused - I've seen all sorts of posts saying different things, especially when it comes to proving power to switches. This is what I did: I chose one of the rails to isolate and made sure I always use that same rail. I don't know if I chose ground or not. Am I doing this correctly?

 

3. A switch will always have one rail isolated and thus, how do I provide power to that isolated rail? Unless I am all wrong in how to isolate blocks, I am thinking I'll need to solder a wire to one of the rails. With Marklin 3-rail this was much easier.

 

Thank you in advance for any responses!

 

Cheers, John

[chadbag]

Are you using Digitrax decoders and transponding or are you using Railcom or are you using some other system for detection?  

 

 

[JohnPatrick]

I'm not sure just yet and am a little lost here with some of the terminology in 2-rail. I have yet to purchase a decoder. So far I've gathered that, for detection, the simplest thing to do is use a decoder that senses the power draw of a locomotive, e.g., the EcosDetector, connecting it to the Ecos unit via the EcosLink bus. But, I don't know if this is considered a reliably simple way to do detection or if I'd be better off using something like the Digitrax BDL168 and somehow connecting it to the Ecos unit. Or the one of the digikeijs products.

 

For what it's worth, in 3-rail Marklin I used their s88 decoders, which connected to a Marklin CS3 or the Ecos unit via a bus. It's what I know how to do and I am probably leaning towards the EcosDetector because it more or less is the same approach as I used with Marklin and I don't yet know much about Railcom, loconet, etc. Except that the s88s detected the presence of a ground on the isolated rail, which I understand is not the way to go with 2-rail.

 

Am I making any sense here?

[chadbag]

I don't know much about this but I got it working on my sample layout using Railcom.  Not only does it detect but you get the ID of the train.  Railcom is a "standard" that is implemented by many of the decoder makers (but not all).  All the European ones and TCS and probably some other ones here.  The Digitrax transponding is in some ways the same thing but a Digitrax exclusive.  I don't know of any decoders other than Digitrax that use it,  Thats why I went to Railcom.  You probably can do simpler detection without the ID etc but the ID is useful to do things -- especially if using something like iTrain.  The normal current sensing type sensors are the simple ones without ID detecton / transponding.

 

I assume the ECOS unit supports Loconet or can support it?  While it's a Digitrax thing, its been licensed and adopted by a lot of folks so it makes a handy bus for connecting stuff.  I kind of assume like the S88 bus?  I know my DR5000 has both Loconet and S88 hookups on it.

 

BTW, Digikeijs has gone out of business.  The spiritual heir is  YaMoRC Digital -- http://yamorc.de.   It is not directly connected to Digikeijs but was founded by the engineer who did much of the original Digikeijs lineup before they parted ways.  In the US they are sold by Iron Planet Hobbies https://ironplanethobbies.com (and probably others).  They should have detectors for both Railcom (which is unfortunately delayed) and current sensing, which should be what your Märklin system used.  If you can find a Digikeijs Railcom detector unit that will work as well.  And there are several other companies that make them.  I suspect ESU has a RailCom detector for the ECOS AS WELL:  https://www.esu.eu/en/products/digital-control/ecosdetector/

 

I chose Railcom for my layouts as it is supported by a lot of manufacturers and lets you do interesting thing.  Because of that I've replaced my EM13 decoders (which are the KATO made by Digitrax decoder for a lot of the KATO stuff in a funny form factor) with the JNS forum EM13 compatible board that works with D&H decoders (and others can be made to work) so that I get Railcom.  ESU and Vimo also make EM13 form factor versions of their decoders with Railcom.  The one or two trains I have where I don't replace the DIgitrax decoder (some of my early conversions using the small wired decoders) I am add a RailCOm transmitter unit: https://www.esu.eu/en/products/former-products/railcomr-transmitter-unit/ or RailCOm capable function decoders to enable it.

 

YMMV and your needs may vary.  I did get a simple RailCom based system working before my house project took all my time and will be soon rebuilding it.  I'm slowly decveloping my own software for tghrotle and eventually train control and need to get it set up so I have a test bed 🙂

 

But I am not the expertn  @inobu may chime in as he is much more knowledgeable about the details.

 

[inobu]

Chad,

 

I just came to search something and the notification popped up stating "Just Now"......lol

 

This is what I have learned and as usual it goes against what most modelers think or say.

 

There is more to block detection than meets the eye. Block detection identifies what is in a block. In some cases

the locomotive only identify its self on entry and exit. You can enter a FIND command (Digitrax Throttle) and the decoder can be located.

 

Some devices like NCE's BD20 only detects current draw. This may suffice for some but not others. 

 

Current sensing device only know something is in the block but not who? Which leads to my contention.

 

If you want full control of your layout. You have to look at your layout block design from 2 aspects.

 

You need to know if it is occupied in order to use it. Then you want to know where within the block the loco or train is located.

Is it in the East entry, Mid or West exit depending on its direction of travel.

 

So you want current detection for block occupancy then Physical sensors (IR,Reed or Halls effect) to trigger one of three or more position indicators.

 

With the kind of mindset you can better design your layout for more advance automation.

 

For example as a train comes into the station the current detection using transponding or Railcom will identify the train. Once it enters

the stations block entry the trains ID can be announced. As the Physical sensor are tripped the control and location of the loco is fully controlled.

 

Being that you have the ECoS I would used railcom and stick with the Lok Decoders. I left Digitrax as their development is stagnant. I switched my clients to 

ESU products.

 

The ECoS is the best all around Command Station then comes the Z21. 

 

What I tell my clients "Always build to the future" because rework is costly and leads you to never ending concessions.

 

Hopeful this give you a different aspect to consider as there are many many direction one can end up in.

Feel free to contest what I'm saying as its an learning discussion.

 

Inobu

[Madsing]

Hi John,

Some additional info.

DCC has no +, - or ground. But it’s still important to identify the two wires. I don’t think there is an official denomination so I personally call the two wires A and B.

Make sure that A is never connected to B. So always connect A (and B) to the same rail. You are right, you just need to cut one rail to isolate blocks. I always cut the rails connected to the B wire. So, all B wires go through the occupancy detectors.

You are right. Leave the switches out of blocks, so power to the switches should not go through occupancy detectors. In addition, I make sure that there is always at least one piece of track outside a block on each side of the switch, and I use that piece of track to provide power to the switch. This is difficult, especially on small layouts, so there are exceptions…
I use Tomix tracks and I have modified the switches to remove their power routing function, thus making sure that power is always routed from and to all rails connected to the switches.

I hope this helps…

Marc

[chadbag]

6 hours ago, inobu said:

Being that you have the ECoS I would used railcom and stick with the Lok Decoders.

is there a practical reason to not use other Railcom decoders with the ECoS?  

[inobu]

Block Detection

 

This is the ECoS Detector wired up.

There are a total of 16 Detection Ports 4 are Railcom equipped. Railcom blocks 

have 2 Way communication. You can query the decoder and it responds.

 

You can also set the ECoS Detector to act as an Input device monitoring events on the layout.

Reporting activity to the ECoS.

 

 

Signal Pilot

 

The Signal Pilot is versatile as you can configure it to operate a number of

devices. It is drag and drop configuration.

 

 

You can configure it for 7 Segment Digital Display.

 

The Signal Pilot could throw Kato Switches.

 

 

It is far more advance than Digitrax.

 

Inobu

 

 

[JohnPatrick]

Oh wow - these photos are great, very helpfu!l! Thank you Inobu and all others, your posts have helped me a great deal. I've perused this forum and have found of wealth of other material, so I have a lot of weekend reading ahead of me - also learning more about Railcom capabilities in products. Inobu, you have the nicest wiring I've ever seen.

 

With iTrain I had good success with using two s88 feedbacks per block, the first to announce a train as it enters the block and the second to tell it to stop. However, I ran into occasional but disastrous problems when something else on the layout would occur that iTrain couldn't handle, and thus I like the idea of a 3rd detector in the block, e.g., optical, as has been suggested here.

 

I get tremendous enjoyment from this hobby because, to some extent, we are modeling the actual operation of a real-life railroad - and, railroad operations are much more complex than I initially realized.

 

Cheers, John

[chadbag]

On 2/6/2025 at 10:46 PM, chadbag said:

Quote

Being that you have the ECoS I would used railcom and stick with the Lok Decoders.

is there a practical reason to not use other Railcom decoders with the ECoS?  

 

@inobu -- Just in case you missed my question I am quoting it right here.  I didn't tag you on it.  

[inobu]

11 hours ago, chadbag said:

 

@inobu -- Just in case you missed my question I am quoting it right here.  I didn't tag you on it.  

Sorry, I didn't see it.

 

The ECoS manages decoders itself.  With Railcom +Plus the ECoS will automatically update your Roster.

There is a broadcast sequence where the ECoS will look for any new decoder placed on the track. If

the decoder has the registration flag High the ECoS will query its information and Update the Roster.

Why not stick with the Lok product line. Its made for the ECoS. 

 

Railcom has a lot of capabilities over Loconet and its transponding. True two way communication.

Some manufactures are utilizing the ability of the decoder to report Fuel levels. This will be a

serious feature when released. It will take modeling to the next level.

 

The German translation can be difficult but with the translation app it just take time to translate.

 

I made my own JR Signal group with the same program that manages the Lok Decoder. See what I mean, the

design map links all their product lines together.

 

 

Inobu

 

[chadbag]

@inobu thanks.   I’m not saying you shouldn’t use the ESU decoders but was wondering if there were specific advantages to them compared to other Railcom or Railcom+ decoders.   Just a learning question.  
 

It sounds like any Railcom+ decoder you might already have would also work with the auto roster add.  Ie no need to replace them if you already have them.  
 

thanks for the reply.  

[JohnPatrick]

Hello all,

 

I wrote earlier about how to setup occupancy detection using DCC and the issues I had in moving from S88 detectors using Marklin 3-rail. You helped me a great deal and I've now had some good success, using the Ecos 50200 and the Ecos Detector. I'm currently setting up iTrain to manage my test layout.

 

My question here is about how to extend the the Ecos Detector, if it is at all possible. It has 16 detection ports but I will need up to 32 (2 occupancy detection sections per block, 16 blocks). I purchased the Ecos Detector Extension, at first thinking that it would provide the additional detection ports but I misunderstood the documentation - it does provide occupancy detection but, unlike the Ecos Detector, it does not output sufficient power for locomotives to work properly within the detection section.

 

So, to get to 32 occupancy detection sections, must I purchase a 2nd Ecos Detector?

 

I'm wondering, though, if it's still possible to use the Extension such that locomotives will still have sufficient power within the occupancy detection sections. I haven't tried this yet, but, assuming that iTrain works with the Extension, could one create occupancy detection sections by isolating rail A, then within the section provide power to rail A, and then connect rail A to the Extension? I have searched the web but haven't come across any other documented method to use the Extension in this way. But, I don't want to try it just yet for fear of damaging it.

 

Does anyone have any opinions here? Must I purchase a 2nd Ecos Detector to get to 32 detection sections? Or is it possible to use the Ecos Detector Extension in such a way that it will be useful for providing the needed occupancy detection, e.g., such as using the method I have described in the preceding paragraph? This all assumes iTrain can work with the Extension, but I suspect it does.

 

Thank you in advance for any help. Cheers, John

[inobu]

John,

 

The Ecos Detector Extension plugs into the ECoS Detector. It provides a means for you to operate or control things like LED's or activating switches.

The Extension has nothing to do with track power. It just uses the ECoS Detector to access the ECoSlink bus so the ECoS 50210 can communicate with it.

It is used as a means to operate an external status board. It allows you to control devices outside of the DCC system.

 

You do need another ECos Detector as it is part of the track power distribution. The ECoS Detector

takes the ECoS 50210 track out put and breaks it into 16 sections. It monitors each sections or blocks for

occupancy and reports it to the 50210.

 

Inobu

[JohnPatrick]

Thank you so much for the reply. Yes, I suspected that I would have to purchase a 2nd Detector. I was hoping there was some way to still use the Extension. The Detector is expensive but it is easy to use and works well.

 

Cheers, John

[inobu]

John,

 

It may be a good idea to get the ECoSlink terminal.

 

On a large layout I place one on each side of the layout and run CAT5 cable between them.

Instead of placing everything at the 50210.

 

Map your layout out and you can see how these will make things easier.

 

Inobu