I am looking for advice/suggestions/recommendations on DCC control software. I originally posted this elsewhere on the forum, but have asked for that post to be deleted as I think I got the wrong end of the stick about what the 'Computer' forum was for!! I am well underway with the construction of a large N Gauge layout, which has been built for DCC operation, with track detection and pointwork operated by Servo control via a control board (that can also respond to DCC commands). Signals (and their control boards) have yet to be added. I do have one sound loco, but most are non-sound. I use DecoderPro for decoder programming. The layout will comprise a 14-road classification yard, separate holding (through) sidings for long coal and intermodal services, plus a locomotive facility. Three core lines feed out of hidden storage sidings (primarily UP and BNSF operated) with a separate CSX line crossing on the flat, which also serves a small intermodal terminal. I do have friends over for occasionally, but primarily it would be one-man operation, hence the need for some automation. I am starting to look at DCC control for some of the operations (ie through trains, or trains running into the intermodal/coal holding roads, changing of locos in the hidden storage sidings etc) but wanting to retain manual control for the yard switching and locomotive movements (the latter in most, but not all cases). Some departures from the hidden storage sidings need to be to a timetable (based on a fast clock), as are departures from the intermodal/coal roads, while departures from the classification road would be triggered by a 'ready to depart' button on the dispatch panel - which ideally would invoke a random timer before actually departing when there was a clear route. The storage sidings are hidden below the layout, so there are five connections from the hidden sidings (which are all inter-connected on a hidden loop) onto the layout. Along with numerous loco storage sidings. The aim is that the computer can swap the locos on every arrival, so the same train does not appear back on the layout with the same power at the helm. I had originally looked at TrainController for this, which basically I can do everything I want it to do, but does need the Gold option for some of the operations, which is certainly still a consideration, but expensive. I have given thought to writing my own software (I come from an IT programming background and considered using Java based on the JMRI library) but would rather concentrate on building and running the layout. Given the above, I wondered if any of the modellers on here experienced in computerised DCC control could offer some thoughts/ideas? I know there is iTrain, but I am not convinced it has the flexibility to achieve the realistic operation and all of the above. Can anyone offer any thoughts/suggestions on suitable software or options? Richie
I'd take a look at DCC-EX and see it if might be an option for you. It can do much more than I'm capable of understanding. Here is an example.... You can find a lot more info on their site here... https://dcc-ex.com/#gsc.tab=0 Check out some of the other catagories down the main menu under EX-CommandStation Add-Ons like this one ... https://dcc-ex.com/ex-rail/index.html#gsc.tab=0 Sumner
Have you looked at JMRI (Java Model Railroad Interface)? It is indispensable for programming decoders, but can also manage running an automated layout/trains.
Yep, I'm totally biased, but I second Sumner's motion to look at DCC-EX. You will get features that will cost you a lot of money from commercial vendors if you're ok to do some DIY. If you wish, it can integrate with JMRI also, which is another free option for the software.
I would also caution against using DCC (or at least the DCC track bus) to control turnouts (track switches). The reason is that one of the most common occurrences that takes down the track DCC bus is a short from a train running against a switch thrown the other way. Once that happens, you don't have an operating DCC bus to throw the switch and clear the short, thus restoring DCC. You can reliably use a separate booster exclusively for switch control and other accessories like signals, etc., so long as that booster's controlling (input) DCC bus is not susceptible to track shorts. A common approach is to run accessories and track booster(s) directly from the command station's DCC bus, and tracks only from the track booster(s). But there are better options for controlling stationary accessories, especially occupancy detection, etc. Depending on your command station, Loconet, LCC, or other choices providing full, two-way communications are more flexible and reliable. These are also supported by JMRI for various levels of automation, in addition to programming (configuring) DCC decoders in locomotive, etc.
Chaps, Thanks for the feedback, sorry for the delayed response. I had read the responses, but just realised I never replied! DCC-EX is defiantly one to take a look at I think, and yes @BigJake I follow your thought about using DCC to control turnouts, they will be on a separate bus I think. Richie
If you are considering true layout automation, then unfortunately, you're probably looking at a Loconet based system or one of the European systems as our friends across the pond are far more dedicated to automation than US modelers are. The ESU ECoS which has proprietary accessories and Loconet support is one. Your best bet is to choose your automation tool and then see which systems that tool support. JMRI and DCC++EX do offer some automation features, but they are nothing compared to the standalone options like iTrain (amongst others)
Thanks for that YoHo. I suspect something like iTrain or TrainController will be the option really. I just thought I'd throw it out and see if anyone had any other options. I am not convinced with iTrain yet however, although I keep reviewing YouTube videos on its operation. We'll see where we go. But thanks to everyone for their feedback and thoughts, much appreciated. I am not really dedicated to automation, id much prefer a group of five or six people running the layout, but where I am there are few other US railroad modellers so 98% of the time its likely to just be me, and I want to ensure I can create the impression of a busy railroad when I am running 'one man operation'! More to ponder! Richie
Also a fan of DCC-EX. It can be expanded to up to 8 output channels, DCC or DC. More than 4 does need a bigger processor board and a bit of wiring since the Arduino shield footprint runs out of pins. EXRAIL scripting allows full automation of a layout from within the command station itself. You'll still need a computer to write the automation scripts, but it can operate entirely without one. Routes, signals, even reserved block traffic control all included in the CS. Adding a wifi shield to the stack allows control from EngineDriver and similar phone apps to give you a walk-around throttle option.
The one thing not mentioned here thus far that I can see, use quality Oxygen free wire for your connecting wires and your buss wires. And you will have better overall performance from your locomotives.
Spend your money where you want, but O-free wire has typically 1% (max 2%) less resistance than regular, solid copper wire of the same gauge and length. By the time you add up all the resistance, from a DCC CS/booster's power supply to the locomotive's motor winding, it's a LOT less than 1-2% of the total circuit resistance. Oxygen-free wire is mostly used for it's mechanical/thermal properties in high-power-generation equipment (not transmission lines!), and in loudspeaker wires for audiophiles with money to burn, and apparently excellent hearing. Your mileage may vary Closed course, professional driver Objects in mirror are closer than they appear Do not look into the laser with your other eye Tax, title & license not included No other warranties are expressed or implied.
As an Electrical Engineer, the DCC Signal does not obey the rules of AC or DC electricity. In this case, Oxygen in the wire causes a reactance that also over distance causes the DCC signal to degrade. It also dissipates energy in the signal. If you were to do a Discrete Fourier circuit model of the buss wire along the length, the incremental losses due to the Oxygen in the copper start increasing with distance. The incremental reactance of the wire causes higher order modes that that grow to a point where they reach an energy point that causes an extinction function that wipes out the signal. You can also look at this as the total dissipation of the DCC signal. This is why they say the maximum length of a buss is 300 feet. (From Digitrax Documentation) Oxygen free copper extends this distance another 30%, and cleaner power is delivered to the locomotives.
Because the DCC signal is neither AC or DC. It is a square wave produced by the electronic equivalent of flipping the track connector over every 56 or 112 uS. Yes, those are the nominal half-bit lengths of DCC-EX, tied to a hardware timer in the processor board. 56 uS for a half 1, 112 uS for a half 0. Pulse stretching not supported on that system. Not really buying it on the oxygen-free cable, likely cheaper to just go up a size for a similar resistance decrease as long as the wire used is low capacitance. Line capacitance becomes a problem with distance, rounding off the edges of the square wave and causing ringing with the line inductance. A bigger problem for signal quality at distance would be back reflections and connector losses.
It is not the size of the cable that matters it is the purity of the copper that matters. Digital waveforms have been around a long time and the mathematics to calculate them has been too. Fourier analysis is how to properly analyze the DCC signal and all of the parasitic nature of the composition of the wire. Just increasing the size of the wire does nothing to improve the DCC information packets if it is not a better composition of copper. The mathematics prove otherwise. I have doctorates in Electrical Engineering, Theoretical Physics and Analytical Mathematics. I do know how to analyze signals and also the wire used to relay those signals. The reflections are because the end of the line is not terminated properly. You would have to calculate the intrinsic impedance of the circuit and decide what the proper termination would be.
Also as a (retired) electrical engineer with 33 years of high speed (yes, including GHz) digital and analog circuit design experience, the effects of additional reactance caused by oxygen in the copper matrix are negligible below high GHz frequencies. Above that, RF power transmission is done with waveguides and/or antennas (e.g. radar/comms, ), not wires. That's one reason modern radars put the emitter circuitry on the radar antenna's flat surface (it also allows accurately steering the radar beam electronically, without moving the mechanical antenna surface except for gross movements, and without heavy, expensive, flexible waveguides.) Trust me, if there were significant GHz harmonics on your DCC layout, you would receive a loud knock on your door (if not a boot through it!) from the FCC. The drivers in the booster are just not that fast, if only for very practical and/or FCC compliance reasons. Have you measured the rise time on the DCC output at the booster, let alone put a GHz-capable spectrum analyzer on it? The DCC bus driver's frequency cutoff, including harmonics, is well below GHz, for very good reason (cost, FCC compliance.) Those tiny ripples at the tale ends of the digital pulses are actually the telltale mark of the absence of such high-frequency content in an otherwise perfectly square wave with vertical pulse edges. Additionally, consumer electronics manufacturers intentionally do not use RF-capable drivers for these applications, for both cost and FCC compliance reasons. But, in the end, we all get to spend our coin where we want... Buy whatever makes you happy. /sarcasm_on/ And don't forget the speaker wires from the encoder to the speaker; O-free wire makes it sound so much better too! /sarcasm_off/
I’d spend my time worrying about good solder joints, tight connections, and keeping the track clean before I worried about the patina inside the wires.
For those of us who are neither EE, or theoretical or applied physicists, or PhD level mathematicians, (but just plain old code weenies doing software who took one EE class to graduate), I assume any of our wire runs in home sized layouts (even a larger one), where the power busses don't approach anywhere near 300', we can not worry about any of this? I would assume my 25' or 50' max power busses aren't enough to trip any of the above stuff. (Especially on N scale where the power is less)
