Hi,
I haven’t been in the hobby for quite a while but I’ve been keeping tabs. Has DCC progressed to the point where trains can be run automatically? I’m not talking about running in a loop but in a realistic environment such as a single track mainline with passing sidings and opposite direction traffic? (computer controlled dispatching). You may be asking why one would want such a capability but I’d think it’d be neat to have. Say you want to run a local and have through traffic (and dispatching) controlled by a computer on your home layout.
Thanks!
BDF
It's been there for a while actually. ANd not insanely expensive - block detection is down to $15 a block or less. Turnout controls, under $10 if you use Tortoises, the stationary decoders for those are the lowest cost because they need no fancy high power driver circuits. You'r eon your own for the software - you can buy a commercial program like RR & Co or use a free one like JMRI, but neither has a "click here to automate your layout" button.
Overall the whole process is probably easiest and cheapest with Digitrax, if only because of the wya Loconet works - you can;t 'cheat'. If you operate a turnout with your throttle, the computer will see it - you don;t need extra contacts and feedback on the turnouts. Block detection for 64 blocks is about $500, if you want signals as well, the SE8-C is $100 for each 32 signal heads. NCE Switch-8's are probably the cheapest Tortoise stationary decoders. Make resistor wheelsets, or buy them, and you're all set. Not cheap, but not megabucks, either.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
Take a look at RR-CirKits - that's where my $500 figure for 64 blocks of detection comes from - one of his Tower Controllers (plugs into Loconet), 8 of his BOD-8's, 8 block detectors, and 8 sets of his detector current transformers. 64 blocks worth of detection for $500, less if you don;t have 64 blocks. That's MSRP right off his site - Tony's actually sells all his stuff for a discount.
You don't need entry/exit/middle detection on blocks to automate your operation. You do need to make each switch 'plant' a seperate detection zone though. A passing siding would be 4 detection sections, the turnout on the left, the main, the siding, and the turnout on the right.
You can use non-sending switch decoders, too - at least with a Loconet system, because the software will pick up the commands right from Loconet. If you really MUST grab absolute contacts, then you might as well go right to added microswitches to the turnout in case the Tortoise linkage breaks - if lives were at stake you can't have enough safety equipment but this is a model, 'wasting' two inputs to signal turnout position isn't necessary. If you MUST have this, just *** a second Tower Controller for $115 to monitor 32 turnouts with total feedback (2 contacts per turnout) or 64 just monitoring a single contact. I was goign to build some LocoIOs but even getting the PIC processor for free it's still cheaper to just buy Dick's Tower Controller.
Transponding, also, is not necessary. If you use commercial software like RR & Co it tracks each train given a start position - I don;t have it but this is one of the advertised features of the software. Most of what would be needed actually IS in JMRI, they've done a ton with the PanelPro part as far as havign variables and so forth to be able to store train information. Depending on how complex you were getting you might need to go beyond what the user interface provides - at which point a commercial program would probably be easier. At $349 it would probably be the single most expensive component in the system.
This is where Loconet really shines - there are multiple third party vendors making components that sell for less than the Digitrax branded ones and accomplish the same thing - in some cases, like the RR-CirKits stuff, in mass quantities compared to the Digitrax offering. There's a lot more you can do then just ordering from the Digitrax catalog. I didn't say it was going to be cheap, but you can make it work, and work well, for a lot less than you think.
Aside from the costs, another hurdle is the complexity (I can barely understand anything in the past few posts). It seems really complex no matter what the cost. It's unfortunate that there are no "packages" that you can buy that are all inclusive including block detection, software, interface, etc.
Check out the Deming Sub by clicking on the pics:
check out this site.www.rnrdigital.com they have a system that will do what you are thinking of. they were looking for some people to test their system on their home layouts for free but i think they're not offering it anymore .
You should also check out www.cti-electronics.com
They use software to track the different locomotives running the layout. You also get to write your own code, so you can make it do exactly what you want, not like some others. All hardware is independent of the DCC system, so it can be used on most of them. The limit in that regard is not all DCC systems are supported, but most of the major ones are. You will have to look through the on-line information to know which systems are supported. Right off hand, NCE, Digitrax, and Lens is supported, and there are a few others.
Elmer.
The above is my opinion, from an active and experienced Model Railroader in N scale and HO since 1961.
(Modeling Freelance, Eastern US, HO scale, in 1962, with NCE DCC for locomotive control and a stand alone LocoNet for block detection and signals.) http://waynes-trains.com/ at home, and N scale at the Club.
davidmbedard You keep saying "for alot less than you think". Why not take my above example and price it out with the componets that you have in mind? I would be interesed as well. David B
You keep saying "for alot less than you think". Why not take my above example and price it out with the componets that you have in mind? I would be interesed as well.
David B
OK - although I actually would end up using mostly the same components because it's kind of silly to buy 64 blocks of detection when you only need 14.
All Digitrax:
1 BDL-168 $149.95
1 DS-64 $60
4 Tortoise $60
PR-3 computer interface 84.95
About $355, less software. You didn't include software either. Add $350 to both for RR & Co which has things liek TrainTracker so you don't need Transponding.
Or with RR-CirKits stuff (let's not get too nuts - there's Team Digital and CML as well)
1 Tower Controller $115
2 BOD-8 $75
2 coil sets $29
1 SMD-8 (drive 8 Tortoises) $29
4 Tortoises $60
1 Locobuffer $63 (MSRP)
$371 less software but you have leftover capacity to drive 4 more tortoises and FIVE free ports inthe tower controller for 40 more lines of I/O. I WAS goign to put in an NCE Switch-8 but they're twice the cost of the SMD-8 and since we were gettign the tower controller anyway.. Then there's the DIY option, I think it cost me about $30 each to build a Locobuffer from Hans Deloof and about the same for a LocoIO. You'd need two LocoIO's for total of 18 I/O ports you need (14 blocks plus 4 Torti). About $90 worth there, plus two BOD-8 and coil sets $90 plus $75 plus $29 plus $60 - $254, but this assumes you knwo enough about electronics to order the parts and assemble them on Hans' circuit boards, and wire up a connection between the BOD-8's and the LocoIO ports (they WILL work together - I verified this since this is probably what i will do on my own layout
And yes, what others said, it's not so much the cost, it's also not eay to implement all this. I haven't played with RR & Co but I seriously doubt for a fully automated system you can just drag stuff around on the screen and magically it works. And for the ultiamte in low cost, a bit of electronics knowledge helps. I guess I cheat - my degree is Electrical Engineering and I've been playing with wires and circuits since I was a kid. I actually ENJOY putting together a circuit board - which is why my own interface was a Hans Locobuffer, and I had one LocoIO I put together just to play with. I also built another LocoIO for someone here - Spacemouse I think. I can crank out a couple LocoIOs in a short evening - the hard part is waiting for the circuit boards to come from Belgium
Bringing this back to light as I am now getting into automation/computer control.
I have about a 300 square foot layout. I have 4 separate continuous lines running 2 trains on each line that I want to automate. I am not worried about automating staging etc but want to automate running of the trains. I want to have it so that if say train 1 on line one stops for whatever reason that the system sees this and stops train 2 so as to avoid collision etc.
So far I have JMRI downloaded and am looking http://www.rnrdigital.com/ as well. I have it that I need a BDL168, (maybe more than one) and some RX4 units. It looks like I have to setup trans-ponding etc and want to make sure I am starting correctly.
I feel fully confident but it's a lot to absorb at once and need to know I am taking the correct steps in the correct sequence.
To run the 8 trains you suggest you do not need transponding. Since you have downloaded JMRI you can enter your track diagram, divide it into appropriate blocks, use LocoNet Simulation and AutoDispatcher and run the trains on your screen to see how it would work. It really helped me understand what needed to be done to the physical layout.
If you have additional questions you might post a drawing of your layout and people here will gladly help you through the steps.
Dave
From Mt Pleasant, Utah, the home of the Hill Valley and Thistle Railroad where the Buffalo still roam and a Droid runs the trains
You definitely do not need transponding. You also limit yourself to Digitrax decoders if you use Transponding. Or at least adding the Digitrax TL-1 transponder to locos with other decoders - although some simply do not get along with Transponding.
A friend of mine has automated his trolley loop. 16 blocks connected to a BD-16. Yes, the old one, that supposedly does not work with supersonic decoders. Well, it works fine. He's up to 5 sets of trolley cars runnign now, with no Transponding, in fact most of the two car sets (they're little 4 wheel Birney cars - in N scale) have Lenz decoders wiht the USP module in the second car (a dummy). This is driven by RR&Co TrainCOntroller Bronze version. The software keeps track of where each car is via the changes in block occupancy. RoboThrottle and AutoDispatcher in JMRI can do similar things but I really fought with AutoDispatcher trying to make it work, Traincontroller was easier.
It's all quite neat, turn on the power, open the file, and hit the Go button, and the trollies take off, one by one, starting and stopping before they run into one another. In the meantime, we run trains ont he main using the throttles.
Oak Tree Systems is also doing a lot of work in this area.
www.oaktreesystems.com
Sheldon
TexasSP ... I have 4 separate continuous lines running 2 trains on each line that I want to automate. I am not worried about automating staging etc but want to automate running of the trains. I want to have it so that if say train 1 on line one stops for whatever reason that the system sees this and stops train 2 so as to avoid collision etc. ...
... I have 4 separate continuous lines running 2 trains on each line that I want to automate. I am not worried about automating staging etc but want to automate running of the trains. I want to have it so that if say train 1 on line one stops for whatever reason that the system sees this and stops train 2 so as to avoid collision etc.
...
Just how much automation are you wanting? If all you are wanting to do is to make sure following trains do not run into the rear of the train in front of them, you don't necessarily need computer control. You can use "breaking sections" triggered by detectors, setup so that the breaking sections follow the trains around the layout. Any time a train gets too close to the one in front of it, it enters a breaking section and will slow down at it's programmed deceleration rate, coming to a stop if necessary, until the train in front releases it at which point it would accelerate back up to speed using it's programmed acceleration rate.
There are two ways to implement a breaking section(actually there is a third way, but it currently is only supported by Lenz and Zimo decoders) - one applies DC to the track to use the "Brake on DC" feature that most decoders have, and the other uses a broadcast set speed zero packet generator. It is not recommend that you use DC and DCC on the same layout for fear of the possibility of short the two together; however, you in this situation you can generate the DC from the DCC by feeding it through a bridge rectifier alleviating that issue. The broadcast set speed zero packet generator just repeatedly sends a set set speed to zero packet to the broadcast address. I know of two sources for a set speed zero packet generator. One is made by DCC Bitswitch, and the other is by using a Digitrax DCS100(there is a setting to turn the program track outputs into a set speed zero packet generator). Either set speed zero generator requires the use of an additional booster for the set speed zero packets.
Whichever method is used for the break sections, you only need on break section generator for the whole layout. For each break section on the layout you have a relay that switches the section between the regular DCC bus and the break section bus.
I need more than breaking sections though. I am setting this up for my father to use to be simple and want a point and click solution for him to run trains.
He wants it where he clicks the file and the trains run. More setup for me, but easier for him.
The layout is ours, yet I do 99.999999% of the work. I have showed him countless times how to power everything up on the Digitrax and run the trains, I even have set speeds to run them at which work well. However running 8 locos at once is hard to keep up with.
I myself will use more control over it when running but in the end this isn't for me and I am not funding this part of the venture. Money is surely not the issue, just time and ease of implementation for me.
Thanks for the posts so far. I will play with JMRI so more and look more into RR and Co.
Transponding is not at all necessary to have an operating signaling systems/ Automatic control.
Here's What I have for my signalling system:
LocoBuffer USB: $58.45
RR Cirkits TC-64 Tower Controller (Loconet Interface) @ $108.95
RR Cirkits BOD-8 (Block Detector, 8blocks) @ 35.95
Current Transformer coils (for 8) @ $13.60
Digitrax DS-64 Stationary Decoders @ $50.95
Resistor Wheel sets (~$80.00 per 100 if you make them yourself)
All Prices above found on Tony's trains website.
You don't even need a signaling system to do layout automation such that trains pass each other. You just need a block detection system to know where your trains are.
If you add the above up it comes out to $563.70 for 64 detection blocks or $8.80 a block. The nice thing about the TC-64 is that if you don't use it completely for block detection you can also use it to run switch machines and signals with its other modules.
if you have 16 turnouts that is additional $204 for the stationary decoders.
One SE8c will control up to 32 signal heads.
The signals are what are probably going to cost you the most money unless you fabricate your own. The average prefabricated signal is about $20 for a single head and about $40 for a double head.
Its quite conceivable to signal/automate a medium sized layout for about $1000 dollars.
As far as code goes for automation I am using JMRI for my ABS setup. I am still working towards the CTC panel but I want to get ABS up first. The Panel Pro module of JMRI is quite powerful and has been able to manage all the different quirks I have on my layout.
Colorado Front Range Railroad: http://www.coloradofrontrangerr.com/
Here is a rough drawing of my layout. 20.5 feet by 13.5 feet.
There are two lines running parallel to each other around the perimeter which are colored dark blue and aqua, then two lines the run the interior colored red and light blue. I didn't include the rail yard, staging, sidings, etc because for this they won't be automated and I didn't have the time. Anywhere you see lines crossing each other there is a difference in elevation and not a cross track section. The areas shaded are walk ways in the interior of the layout.
So after some more searching, calling, discussing, reading, etc I have come down to two solutions:
1. DCC Bitswitch-all electrical/mechanical no computer needed. Uses solid state relays, magnets, and reed switches. It seems very robust and I talked to the owner at length. Start it is fairly simple though not as simple as straight computer control. Installation and implementation seems very simple though. I also got a lot out of my conversation with the owner Fred.
2. CTI-Electronics-computer automated using their software, IR block detection, directional travel detection, pricing is good and startup should be simple by opening the system and letting it run. Looks to be a little more intensive to set up and get programmed although not rocket science by any means. Owner was difficult to get on the phone (said he had been out of town etc so that seems reasonable) but when on phone friendly and helpful although not to the same degree as Fred at DCC Bitswitch. I do not have anything bad to say about CTI owner TIm though as he was friendly and helpful.
So with these two in mind does anyone have any experience/opinions on either? I am leaning slightly CTI for my fathers sake of simplicity in running trains but if it was up to me alone I would probably go with DCC Bitswitch. Neither one at this point has presented me with a deal killer or made it a slam dunk so this is why I am back asking opinions.
Figured I would give this one more shot to see if anyone has opinions on either system.
I have no experience with Bit Switch, although I have seen their video, and the system looks as though it may do what you want.I have had experience with CTI on a DC layout that I automated once, but not to the extent that you want to do. It will be just as difficult as using any of the other systems mentioned.At the stage that I am at now, if I was going to automate my present layout (which I have been thinking about doing) I would use the RR&Co program because I already have block detection and signals installed.I think that you have to look at what your ultimate goal is going to be when the layout is finished. If you ultimately want signals, I would go with the Digitrax detection and signaling system and RR&Co program. If you never ever want signals, you can go with something else, but don't expect to be able to come back later and install the signals later without some problems. They should be planned for from the get go, even if you don't install them right away.If your father is having trouble operating your Digitrax system, maybe you should switch to NCE which is more user friendly. And by the way, I don't think anyone can operate 8 trains all at the same time on the same layout. That is not what DCC was designed to do. It was designed to allow 8 people to each operate one locomotive, all on the same track without using a bunch of special wiring and toggle switches.
gandydancer19If your father is having trouble operating your Digitrax system, maybe you should switch to NCE which is more user friendly.
That's your opinion. I, for one, find Digitrax to be more user friendly and ergonomic than NCE.
gandydancer19 And by the way, I don't think anyone can operate 8 trains all at the same time on the same layout. That is not what DCC was designed to do. It was designed to allow 8 people to each operate one locomotive, all on the same track without using a bunch of special wiring and toggle switches.
I do agree with you on that point. But I also think that's what his father is having trouble with, not the Digitrax DCC system itself. In fact, since you can simultaneously and independently control two "trains" with DT4xx throttles, it arguably could be easier with a Digitrax system.
gandydancer19...And by the way, I don't think anyone can operate 8 trains all at the same time on the same layout...
Well, I've never done eight, but I have done seven! It was on a layout with three separate loops. On a layout the size of the one TexasSP showed with four separate loops, I could do eight.
to BDF:
It's very strange that in this entire thread, nobody has mentioned Bruce Chubb's C/MRI, which may be exactly what you want - IF you have some electronics skills and beginning programming knowledge. In short, you could build a system where YOU sit as dispatcher in front of a computer screen, with CRT representation of signals, blocks, etc. Your father could run a/some train(s) as engineer.
C/MRI does take some electronics know-how and some assembly. The programming is entry-level Visual Basic or equivalent, and many routines are available already.
Before all the know-it-alls-with-no-actual-experience chime in on how terrible C/MRI is, you might want to check out the email list for yourself:
http://groups.yahoo.com/group/CMRI_Users/
C/MRI runs very well with DCC, can easily run JMRI as a visual interface, etc.
By the way, if all your father wants to do is point-and-click-run-trains, you might get an expensive system built (they are ALL expensive...) only to find he has no real interest remaining after the first session of watching trains go round-and-round. If point-and-click is really wanted, you might try one of the popular TRAINZ or MS train simulator programs - seriously, no offense intended.
Dick Chaffer
He can do the same with Digitrax and third party bits that interface with DIgitrax. C/MRI is a great choice with other DCC systems that don't have a signal bus that can accomodate block detection and signal drivers.
C/MRI is great, if a bit expensive per bit (unless you source the parts and build allt eh modules yourself). Programmign is a wash compared to other systems since they all can use JMRI, although I don't know if RR&CO works with C/MRI - RR&CO, while not cheap, is MUCH easier to set up basic looping trains under automated control.