JMRI - Hardware needed for Signal Switches

[kvp]

I can't find any info if the SE8C would operate from the DCC data stream on the railsync lines at all as it looks like a pure loconet accessory. This means it will only work with the turnouts if the command station emits the turnout change commands both towards the DCC and the loconet bus or if the software on a connected computer sends the right commands to both buses.

.

Btw: A slow motion turnout control source with constant DC output can be used to drive a Kato or Tomix turnout if a high enough capacity (around 1000 uF) polarity-less capacitor bank is inserted in series with one wire of the bipolar coil drive as it will act as a timed current limit and convert the DC signal into a spike of bipolar switching power required by the turnout coils.

[inobu]

SS,

 

I went over the setup again and the issue remains the same. You have your switch controls (DS51K1) on the DCC track and the lighting control (SE8C) on the Loconet network. The problem is when the command is sent to the DS51K1 on the track the SE8C will never know.

 

The SE8C has an association with the switch numbers and the corresponding lights/signals. A Zephyr or Digitrax command station will resolve the problem by issuing the command across the track and loconet. This bridging allows both components to react to the same command on two systems. Although the SE8C has 8 switch controls build in it does not support Kato's switches (bi-polar). 

 

Another solution is to change your switch controls to the DS64 which is expensive. By incorporating the DS64 into the scheme the switch controls will move onto the loconet side of the layout. The DS64 can listens for commands on the track and loconet and will allow you to use the SPROG. In this configuration the DS64 and SE8C are connected to each other without the commands being sent over the rail. When you look at the cost that is about $160 to $200. The same price of a Zephyr. Getting the Zephyr gives you added features. 

 

 

In any case I think a design document is in order. Typically I create the layout, generate a parts list and place it on a excel spread sheet. Afterwards I hit Modeltrainstuff and price it out. Now, I have a design doc I can make a good assessment on everything. The design doc will help you in verifying the design/components. 

 

Inobu  

Edited February 27, 2016 by inobu

[kvp]

Although the SE8C has 8 switch controls build in it does not support Kato's switches (bi-polar).

I think the outputs of the SE8C can be set up for bipolar DC turnouts with stall motors or end switches (like tortoises). Adding AC capable capacitors in series with the Kato or Tomix turnouts would work. The caps will convert any constant DC signal into pulsed DC. Around 1000 to 2200 uF would work at 12V DC. (just make sure the caps are AC capable!)

[surfingstephens]

Hey Guys I am back, so the process begins.  So I started out with:

    1. Sprog Controller (Switches and Train)

    2. Digitrax (SE8C Signaling)

    3. RR-CirKits (locobuffer-usb rev-n)

    4. JMRI

    5.  SSB-Gateway  (Did not order this, but it turns out I needed it)

I was able to easily add the Locobuffer as a second command station in JMRI.  Pretty cool how it handles multiple devices.  However I could not get the Locobuffer to "speak" to the SE8C signal board.   

 

After a lot of reading I figured out I needed two things, power to railsync and a termination point.    Turns out the old Locobuffer had a jumper you could put in place to have a termination, but this new version does not.  I sent an email to the RR-CirKits guy who was very helpful.   He explained I also needed a SSB-Gateway device.  This device puts the power to the railsync and also does the termination.  It only costs  $26.95.

 

http://www.rr-cirkits.com/description/index.html

 

At this point I found a document that showed in great detail how to wire up the BDL168  (Occupancy Dector).   The wiring was pretty involved to make it work with a non-Digitrax (sprog in this case) command center. 

 

http://jmri.sourceforge.net/install/LocoNet_for_non_Digitrax_layouts.pdf

 

At this point I paused, and thought, path of least resistance.   I could probably make this work, but why make it more difficult.    If I just purchased a Zephyr I would avoid a lot more challenges that I do not feel like spending time on.   So I bought the Zephyr.   I will discuss where I am with that on the next post with some pictures.

Edited March 1, 2016 by surfingstephens

[surfingstephens]

So off to almost the last hobby store for trains left in Houston, Texas on Sunday.  There are three decent ones left.    I purchased a Zephyr from Digitrax and connected it to the Se8c board successfully using the Locobuffer to my PC.  After being a bit lost for a while I got the signal heads to actually start working using JMRI. This entire signal head functionality is cool, but I am still trying to understand it as it is all new to me and a bit much to absorb.   JMRI appears to have many different ways to implement this and there is big learning curve.  

 

I should get my BDL168 board in the next few days, so I am going to first get the blocks set up for my oval and get that working and then circle back to the signals I guess.      I was thinking of just starting with a signal for each block to start and then see about signaling for the turnouts.    Here is the pictures of where I am now.  I threw in a portion of the real layout I have as well.   I have had bigger setups and many more trains and buildings, but I decided to scale back and keep the hobby to manageable size as I have other interests as well.   This layout is 18ft long.

 

I did run into a potential problem.  My hobby store guy is extremely knowledgeable and tells me the DS5K1 decoders(Digitrax) that I used in my switches are going to be a problem.   These decoders are soldered into the switches and draw their power off the track.   I thought this was pretty neat so I have all my switches set up this way.  No holes to be drilled and I can move them around easily.   Well, apparently this will cause my BDL168 to incorrectly believe there is a train on the track because of the power draw!    I hope my guy is wrong on this, but I bet he is right.   I am going to set up a diagram or copy one for my oval to get my blocks set up and get a more formal plan together and we will see if this is an issue.     If these decoders are a problem in the switches, not sure what I am going to do about that.  Lot of labor and money went into that.  

 

More to come.    I am going to just do some more reading on it and will update the post when I get the  BDL168 in the next few days and start laying out the blocks.

 

 

[inobu]

That's the way to go. You can be up in no time which you proved that. lol.  Now you will be able to control the train from a fixed location with simple controls. As you install components you will be able to test them from the Zephry and migrate the control to JMRI. If there is an issue you can roll back and use the Zephyr to troubleshoot it.

 

As for the DS51K.

 

He is kinda right. Because the BDL168 looks for current draw. the switch will register as a decoder on the track causing the BDL168 to flag the block as occupied. Depending on where you place the blocks and how you isolate them the switch may be a hassle. The additional wiring needed to provide power to the switch because it needs to be isolated from the block may cause you to do away with them.

 

Your best bet is to work on your yard first. It is the easiest means to visualize the functionality of the BDl168 and how you need to work around this issue.

 

Let me know when you are ready I have a training board that will help in understanding it.

 

Inobu

Edited March 1, 2016 by inobu

[kvp]

There are 3 locations where you could connect your turnout decoders:

-to the track: they will be seen as any other locomotive

-to a separate wire (from another booster): they will act like any dcc accessory

-to loconet: this allows bidirectional communications (but they may need external power)

 

Personally i would vote for the 3rd. Also that standalone loconet pullup is way too expensive for a diode, 2 transistors and resistors and a 12V adapter. (and it's only good for the cab bus, aka loconet T)

 

The bdl problem can be avoided if you wire all your turnout districts as separate undetected blocks. Looks ugly but only needs isolators and wires. The other option is to power the decoders from the dcc accessory power wire. (it's the main dcc power split before the block detectors)

[Martijn Meerts]

The general rule of thumb is that a turnout should never be part of a block, so using a DS5K1 should be no problem at all. You do of course need to add insulated joiners, and wire the turnouts.

[surfingstephens]

Thanks for the input all.  I like the thought that a turnout should not be part of a block. 

 

I might just isolate the switches/yard for detection, but still use some signaling just to show which way the switch is thrown rather than whether the track is clear. I am pretty surprised that Digitrax does not mention this anywhere that I have found to warn you of this issue ahead of time.  I would have reconsidered the switches at that point had I been aware. 

 

The guy at my hobby store(ex computer hardware guy) states that he might be able to work around my issue by doing some work on the BDL168 board.  Essentially dialing in the level of resistance that is the base line with the switches plugged in and then it would detect anything above that resistance. We also discussed using transponding as a possible work around as well.   Shockingly, he loaned me some brand new digitrax transponding detectors to try that out, but I am not certain any of my trains have transponding decoders, will need to review it.   For now, I will just isolate the switches from the plan on my oval track and see how to make that work.  I will also try leaving the switches in to test for myself how the bdl168 reacts to that set up. 

 

Inobu, tell me about your training board?

[kvp]

 

I am pretty surprised that Digitrax does not mention this anywhere that I have found to warn you of this issue ahead of time. I would have reconsidered the switches at that point had I been aware. 

I'm pretty sure digitrax does not suggest wiring the turnout decoders to the actual traction power for anything more than a playtime layout. The roadbed embedded decoders are actually meant for floor running, where block detection is rarely an issue. For real layouts, the loconet bases turnouts decoders are suggested, that can feed back position data. (if you hand throw one, it will show up on the layout control board)

 

 

The guy at my hobby store(ex computer hardware guy) states that he might be able to work around my issue by doing some work on the BDL168 board. Essentially dialing in the level of resistance that is the base line with the switches plugged in and then it would detect anything above that resistance.

Imho this is a bad idea, as this would still make them show up as phantom locomotives while moving the turnouts and the setting would essentially make cab cars invisible. (so if you are using emu/dmu sets, only the motor car will show, not the head/tail cars)

 

For real control situations, the turnouts can be either undetected, detected as single block or detected in smaller blocks for route control to know when the trains have fully cleared them. The latter is required for automatic route release systems.

[inobu]

I know you are not going to like this but your best bet is to pull the decoders out and hang the decoders/wires for the time being. This way you can swap the decoders with DS64 in the future. The reason I say this is you are going to have to run track power to the switches anyway. So it's a lot of work now matter how you look at it.

 

One thing I learned is to address and resolve the problem directly. It will save a lot of time. Usually we remedy the problem only to have generate another one.

 

It's good that you have a hobby guy that knows Digitrax. Most people shy away from them because it more complex. I have to admit it's somewhat convoluted but they have the best feature sets.

 

If you have Digitrax FX3 decoders then you have transponding. It is a Digitrax specific feature.

 

Inobu

[surfingstephens]

Ok, I am going to use the diagram that is in the pinned write up in this forum for advanced blocks.   See the top most sticky link (Chapter 3).    They have the exact layout that matches mine so why not go with that for a first try?   Also it does give the logic of why you do not want a switch to be part of a block.   If I go with that thinking then my switches will be isolated and I will not have an issue.   Hmm. I guess that means that if I isolate them and then power them directly to the power bus, not part of the power blocks(BDL168).  Seems like that will work fine for starters.  

 

At some point, I will have to circle back and deal with these switches.      I understand the pro's and the con's now.      My hobby guy(will remain nameless) apparently has had a few long involved conversations with digitrax on this problem and feels they are ignoring the problem and that they should publish at the very least a warning.    He does have a point, but I am not really interested in spending time on making the argument, so I will just accept it and deal with it at a later point when I decide I need to make switches a part of a power block.

 

Postman has arrived, going to see if my BDL168 is here....

 

​

[inobu]

Here is the Demo board.

It gives a better visual on the wiring of a layout.

 

The BLD186 has 4 Zones.

 

Zone A - Yellow wire

Zone B - Orange wire

Zone C - Blue wire

Zone D - Violet wire

 

The Zones are then converted into 4 sections and the  The sections are wired Gray,White,Green,Red. this is how you get 16 sections.

 

The transponding is achieved by the RX4. This configuration allows for Zone monitoring. This is called Upstream because the RX4 is wired before the Zone are divided into sections. In this configuration the reporting is done on a Zone bases. Meaning that you can only see where what zone the unit is in.

 

If you want to see further down (into the sections) you must place the RX4 Downstream on the output of the BDl168. This means that the section wires (Gray,White,Green,Red) will be ran through the RX4.

 

The SE8C would be to the Left of the BDL168.

 

Inobu

 

This wiring scheme is the way to go as you are able to reconfigure the system easily.

Edited March 2, 2016 by inobu

[Dani]

Hello,

 

Just to add another opinion:

 

About question 1: I completely agree with you about letting the software manage the lights. This will allow you to setup a more complex behavior of the lights, not only change them depending on the switch position. About what to use, it depends. If you want to continue using your DCC only digital system, then any DCC output module or decoder can be used. But I also suggest to change to a Loconet capable system. In this way the range of output modules, sound modules, sensors, .... available are almost infinite. You can choose any Loconet capable command station for that (Uhlenbrock, Digitrax, Digikeyjs, ... or even build your own). Depending on the country some area easy and cheaper to find than others. This will also allow you to use a more advanced software. Some are free like Rocrail, others must be purchased. May be this post can help you:

http://www.clubncaldes.com/2015/09/how-to-make-your-first-step-in-digital.html

 

About question 2: (what pre-built red/green light to buy) I personally bought dwarf signals in ebay and I'm very satisfied with the price and appearance. Each one is more or less 2$:

http://www.ebay.com/itm/5-pcs-N-gauge-LEDs-made-Dwarf-Signals-2-aspects-G-Y-R-/400233606033?hash=item5d2fc82b91

 

About question 3: I'm sorry I just used JMRI for decoder programming, so my knowledge is too basic to be able to help you.

 

Cheers,

Dani

Edited March 2, 2016 by Dani

[kvp]

Is there a more simple solution? Like having a single circuit with a DCC input, loconet connector and N block outputs, with occupancy and transponding reported over loconet on a 1:1 basis. (the BDL168 supports 4/16 or 8/8+8 transponding/occupancy modes, 8/8 or 16/16 would be the ideal without all these extra wires) The preferred way would be pluggable screw terminals for DCC in/out and standard 6 wire phone sockets for loconet. This would give a pure loconet solution with minimal wiring (essentially i'm talking about block detector and rx circuits on a single board). It would be ideal if it would be fully optically isolated and common rail / common ground free.

[surfingstephens]

Simple quesition, I hope.   There are 4 zones and within each zones there are 4 sections for a total of 16.  I was reading through the manual last night and it shows that per zone you have a common rail, but each zone(4) is completely electrically isolated.  Did I understand that correctly?

 

Second question is, if I am doing 5 blocks for occupancy detection(testing), then I would wire the first 4 blocks in zone 1 and the 5th one in zone two.  Correct?  

[inobu]

Question 1.

 

Yes, that's correct. The BDL is taking a single A rail input and creating 4  A Rail outputs. Those outputs are called sections. In general terms they are blocks but withing the context of your design and operation the term is interchangeable with section and zone. It depends on the managing device. 

 

To further clarify the BDL168 a smaller version of the BDL168 exists. The BD4.

 

This is the basic BDL device.

 

 

The BDL168 is four BD-4 on one board. Now you can see where the 16 comes from and the 8 in its name is from the 8 RX4 transponding inputs. Your statement is visualized in this image of the BD4.

The Block detection units will monitor the 4 outputs for current activity. Because they are separated into individual circuits the BDL can identify/label the activity and report it to Loconet where other devices will see the occupancy status.  

 

 

 

2nd Question

 

Yes, but there is a variable to your answer. You must decide how you want to organize your sections/blocks. Lets use Martjin example.

Although you can make the assignments as you stated (Utilizing Zone A completely and 1 section in Zone B ) you can better manage your BD assignments by laying out the blocks as follows

 

Block 1 and 2 as Zone A Section 1 and section 2

Block 5 as Zone B Section 1 

Block 4 as Zone C Section 1

Block 3 as Zone D Section 1

 

Further defining the zones as cardinal points or compass directions ..........Zone B is East........... Zone C is North.......Zone D is West .........Zone A is South.

 

Managing this type of nomenclature will allow you to expand. For example adding another switch or two in zone A will be as simple as wiring the new track to Section 3 and 4. 

 

If you refer to the Demo board image you can trace the wire bundles and see how the sections repeat in color to identify the sections and the color band represents the Zones.

 

It is all about documentation and labeling at this point.

 

Inobu  

 

As I stated the product line is vast and has depths. The problem is it makes documenting it difficult. So you must piece together information in order to understand the system.    

Edited March 3, 2016 by inobu

[surfingstephens]

Very helpful explanation.   It is all starting to make sense.   Ok, now I just have to find a slice of time to get the wiring done and configure it.  It might be a while, since I have two kids, job, electric rc boat racing, other obligations...life.   Hopefully, some time this weekend I will get enough time to do the wiring at least.   I will take some pics when I am done.   I will not be going for "neatness" as it is my test area. 

 

I am using Kato track, and I must say, they have thought of everything it seems.  I have joiners that are electrically done as well as non-electrical.   Makes it a breeze to do without a lot of soldering.  Very simple. 

 

I used to have a large HO set up with mountains and scenary and Peco track and switches  While I enjoyed it, the soldering over and over gets old after awhile.    I have spent some time visualizing how I would do the wiring above the table in a neat way since I am not going for aesthetics, but there will be so much wire I am pretty sure I would be happier going underneath the table so I don't have a bunch of clutter on top.  Going to ponder that some more down the road...

 

oh yes, I also like the thought process on naming conventions. 

Edited March 3, 2016 by surfingstephens

[inobu]

Yes, it takes time but neatness counts when it comes down to trouble shooting. Everything counts even orientation.

 

In terms of orientation................. I had the SE8C mounted upside down in comparison to the drawing/image. I placed the sign mast just as they had it in the instruction sheet. I went through the test sequence and it would never activate sw 257's lights. I was thinking that the board was bad and boxed it up ready to ship back to Digitrax. the next day I thought about it, opened it up, installed it back the same way and started testing again. 2 hours later I realized that the way I had it mounted was upside down. So I was plugged into DRV8 instead of DRV 1. Looking at the instruction without taking the orientation into consideration.   Afterwards I labeled the DRV's on the board. Wasted time and lesson learned. 

 

In any case I think you have the general idea and it will be easier now. Make sure you label everything an take your time. if you wire the connector wrong you can toast the board. I built demo boards for every configuration and made them interchangeable which makes testing easy. .

 

Inobu

Edited March 3, 2016 by inobu

[inobu]

I think I better add this.

 

This endeavor as a whole can be over whelming. The wiring can be challenging especially if you are doing it alone. Once you get past the electrical challenges the software/JMRI challenge makes its appearance.

 

JMRI is a killer in itself as it fools the crap out of you. Getting the system to control a few switches via the computer can build you up. The problem is reality is just around the corner out of sight. When you turn that corner (after seeing the switch throw magically) you have this huge JMRI dragon looking down at you. You are now facing the "JMRI dragon" with a Kato #6 turnout in one hand and a self printed certificate of achievement in the other. What was supposedly a relaxing hobby is now a battle for your sanity.

 

In side JMRI you will find Decoder Pro and Panel Pro with Panel Editor, Control Panel Editor, Layout Editor. As you wonder through the corridors of JMRI you suddenly get this overwhelmed feeling that darkens your world. As times progress your visions of grandeur crashes and burns...............

 

So in giving a heads up to the darkness that looms ahead I will also give this link that will lead you past the wilderness of JMRI despair.

 

http://www.quaker-valley.com/CTC/JMRI_Panel_Tutorial.html

 

This is a great tutorial that will help you setup a complete system. You will be able to piece together a small layout with the basic components to get you going further along then remotely throwing a switch or two back and forth.

 

Inobu

Edited March 5, 2016 by inobu

[surfingstephens]

Progress...

 

I finally got some free time to make some progress on this. Got the wiring done for my test track.    Lot's of soldering and etc.   One pictures shows the areas that are not monitored being wired.  These are the switch area as well as areas to give the train a chance to stop before hitting the next block.  I just got the wiring done, but have not turned it on yet, ran out of time.  The wiring gets crazy real quick.    

Edited March 21, 2016 by surfingstephens

[inobu]

That's it Stephen lol . You are getting to done.

 

Go to Home Depot and get you a sheet of the 24x 48 whiteboard material. I think its $10 and the color pens another $10. You seem to be an Ad Hoc type of guy. That way you can put your ideas on the board and do them.

 

Keep it going.

 

Inobu  

Edited March 21, 2016 by inobu

[surfingstephens]

Thanks, yes this little oval is my 'playground" to figure it all out, before I apply it to my layout.   When I do that set up you will see something much more organized.   I like the write up on the signaling you did!  Now that the wiring is complete I will be going through all of that and will start working on the jmri side of things soon.

[surfingstephens]

This picture doesn't look like much but it shows:

   1. Everything is working  (green on all boards)

   2. The train is still going around the track.

   3. With the testor supplied with the BDL168  I can see that it is correctly "seeing" each block as the train goes around the track.   Never thought it would be so exciting to see a red light come on indicating the train is in the block.  Chuckle.  

   4. I am able to manually change the lights for the SEC8 as well

   5. The switches (2 of them) are working fine as well.

 

Electrically, I appear to be in good shape.

 

Now, Blocks in JMRI.

     1. How to set up the 16 sensors for the BDL168 to show up in JMRI.

     2. How to graphically represent those blocks in Panel Pro for my oval.

 

After I get the blocks working correctly then I will go after the signaling...

 

Any hints on how to do the blocks at this stage are welcomed.  

[surfingstephens]

Hey!  Got it working with Blocks!   I found a really nice video that shows how to set up blocks which made sense to me.  Picture is worth a thousand words..

 

 

 

Very cool to see it work...