DCC systems, how to choose which one?

Which system to use?

This is, of course, a subject of endless debate.

Every system has it's strengths and weaknesses.

I'll present some of my thoughts and lessons learned here.

Starting the decision process, do you really want DCC?

Let's make sure you are here for the right reason, i.e. for most people, DCC provides some unique, or much easier to use features:

• multiple train control on the same track
• advanced remote control features
• consisting of multiple locos, especially with "smarts" for lighting and direction
• Many choices in equipment, and interoperability, i.e. not "stuck" to all equipment by one supplier.
• even more advanced features like position location, automatic loco recognition, etc.

So to begin the decision process, you really need to think about how you want your layout to operate when it's "finished":

If you are sure you will use just a few locos, and limited use of advanced functions, maybe you should reconsider if you want DCC at all. Wireless battery operation has a lot going for it. No track power, no worrying about scrubbing the oxide from brass track (although you should go stainless for track power), relatively low cost if you have just a few locos, or run a trailing car with batteries and a receiver.

Do you really want them to run by remote control, or are they just circulating on loops, one loco to each loop.

OK, now my strong recommendation: if you are going to run DCC, do NOT get a limited or beginner system, entry system, or components you will stop using when your layout grows towards completion.

There are ways that systems can grow, but you do not want to buy a system that is "hamstrung" from the start.

(in my opinion, this leaves out LGB, CVP, MRC, Bachmann, etc.)

Also, think long and hard about systems that are not really 100% DCC. There are many systems out there that are half and half.

I would recommend you choose from NCE, Zimo, Digitrax, Lenz, ESU or Massoth. (not in any particular order). Furthermore, in the USA, you need to be sure you can get the support you want. For example Massoth is great in Europe and the UK, but honestly pretty miserable here (contact me for war stories on Massoth support in the US). Lenz likewise used to be quite well supported, but it is a shadow of it's former self. Finally in the USA, ESU is not strong on support.

These are the big boys, and these are the companies where the next great set of features will come from.

Alright, alright, which one?

A few things to "un learn" first! You need to go into the decision making process unadulterated.

Remember that your decoder selection is pretty much independent of the control system you use, so do not get caught up in the belief your control system must be the same manufacture as the decoders. This is the first mistake many people make.

There is a caveat to this, some decoders have features that are most easily used with the same brand control system, i.e. Zimo has a whole bunch of features built into the throttles that make the sound customization easier.

Some of the transponding (location) features are more fully integrated into the control system, although this is still in flux and not well standardized. So if this is really important, you need to research this. This can also be connected to the needs of a fully automated layout. Read up on it.

What was NOT important to me (Greg):

Why start here? Because a typical problem in this hobby is people unable to make a decision. The usual way a new person goes about this is to compile a HUGE list of features and try to figure out which is "best".

A better way to get to an answer is to ELIMINATE options that are unimportant, so you can make your decision on fewer criteria. Take the analogy of buying the "best" car. If you know you will often have more than 2 passengers, you can ELIMINATE consideration of 2-seaters. If you want to pull a heavy trailer, you can ELIMINATE consideration of a small, light vehicle.

I am not into transponding (location of trains) in an outdoor layout. On a large club layout indoors, with multiple blocks, etc, transponding is fun, but I walk alongside the trains, or run them where I can see them at all times, so there is no big benefit for me. It also adds quite a level of complication. If I ever want working signals, I'll put in simple block detection.

I am not wanting automated remote control built into the DCC system. In my opinion that is somewhat limited, I use JMRI software, it's free, and it supports most DCC systems. I can automate to a degree, but you really need some detection scheme to "know" where the trains are. There are plenty of options with the various systems, so I figure it's not going to be an issue if I want it in the future.

Many DCC systems also provide "streamlined" methods of programming unique feature in THEIR decoders. It does make it easier, but again, I want the freedom to choose the decoders I use, and all the systems listed above program easily. Also again, I can get JMRI software for free that supports almost every decoder from a computer, and can store their custom programming if desired.

This "exercise" helped me eliminate "false advantages" from various systems. Often the advantages touted of systems are unique features for only their hardware. If you REALLY have to have these unique features, then by all means, buy what you want, but you may be locking yourself into a smaller selection of equipment you can use unnecessarily.

Now that I have leveled the playing field, what is important to me were the things I would use most of the time:

• Easily selecting a locomotive
• Easily consisting multiple locomotives into one train (MU diesels, doubleheading steam, adding helpers)
• Ease of use to control turnouts
• Controlling accessories (mostly turnouts)
• Operating the sound systems' features.
• Reliable and extensible wireless control
• Throttles that are easy to see and operate
• History of updates for features or bugs or enhancements
• Good customer service record
• Good support on forums, i.e. user base

What about throttles?

Basically, using the system will be centered around the throttle, how easy it is to use, i.e. the user interface, so this is an important decision.

(Normally the system has a bus or wireless component that connects the throttles/radio base stations to the controller (which transmits the DCC commands), and they are pretty much equal, no big differences in terms of implementation.)

• From my long experience with software, I know that the fastest way to operate a system is a bunch of dedicated "buttons", (as opposed to multiple levels of menus). If you have ever been in a BMW car where there is a big knob, and you keep turning and pushing to "dig down" to the item you need, you have an example of too "deep" or complex of a menu structure.

• The display should be large, not just the size of the display, but the information on it. A large display with teeny characters is no good. It should be sunlight visible

• Extra"fluff" in the information screen is of no real utility to me. I don't really care to know the voltage or current at the expense of smaller characters on the speed or function status.

• Graphic symbols for functions are only helpful if they are customizable and help me operate things. One example is function buttons for sound. A picture of a bell is easy for anyone to recognize, remembering "function 9" is not.

For me, the NCE pro cab won hands down. It has a 2 line monochrome screen that is very visible in sunlight, and it is backlit.

Recently I have changed over a Zimo system, which has some better things and some worse things. I do miss the large screen and extremely sunlight readable screen of the NCE. The Zimo had other features that I needed over the NCE. I'm putting up with the "too busy" display on the Zimo. It is a touch screen though, that has some advantates.

An additional consideration was wireless operation. The NCE wireless controller works and looks exactly like the wired one, same for the Zimo.

Also, I liked the fact that it is a duplex system, i.e. when you press a button, you get confirmation that the command was received. If you send a command and do not know if it was received, then it's difficult to tell what the problem is. Many systems are not duplex. The Zimo is also duplex and better communications range and speed. In all the systems I have recommended the wireless operation is good, except the Digitrax, it's not real reliable for outdoor layouts.

Consisting:

This is a major feature of DCC, both for steam locos and diesels (or a mix!). Basically this allows a single control of multiple locomotives in one train, whether they are all together, or spread out along the train (distributed power). I run consists virtually all the time with diesels, so I need it convenient and easy, therefore the "support" for consisting in a DCC system is important.

First, let me explain consisting a bit, since there is a LOT of confusion on this, and people using terms incorrectly, and inconsistently. Also virtually EVERY manufacturer uses different or conflicting terms. My definitions focus on how the consisting is controlled, not marketing or trying to convince customers to go their way.

• Address consisting - all locos have the same address - the decoder and the system is unaware of any consisting, all functions go to all locos, called all kind of names by various manufacturers
• Command Station consisting - the command station is told which locos are where in the consist and the command station manages how they are controlled, functions, etc. it sends commands to EVERY loco in the consist. It's clear this uses more bandwidth and can be a problem on large layouts. - NCE calls it "old style" consisting, not helpful.
• Decoder-based consisting - all the consist "knowledge" can be in the decoder, the the important part is that the decoder is given a "consist address" via CV19, and the decoders react to that address. There are features to change whether decoders accept specific function commands, i.e. a decoder can be told to ignore a horn command, or a lighting command.
• Throttle-based consisting - really pretty much the same as #2, but now the "data base" resides in the throttle, not the command stations. Some versions allow sharing/exporting this information between throttles.

1. Address consisting - all locos have the same address - the decoder and the system is unaware of any consisting, all functions go to all locos, called all kind of names by various manufacturers, sometimes known as basic consisting long ago, when DCC was pretty new. Other terms have been used but they lead to confusion.

1. • The DCC system has no clue if there are one or more locomotives "listening" on this address.
   • Also the locomotives have no clue they are in a consist.
   • Any system can do this, since neither the DCC system nor the decoders in the locos "know" what is happening.
   • All locos have lighting on all the time, no matter where they are in the consist.
   • Pain in the butt since you cannot "make up" a consist on the mainline, since they are not independently controllable.

2. Command Station consisting - the command station is told which locos are where in the consist and the command station manages how they are controlled, functions, etc. it sends commands to EVERY loco in the consist. It's clear this uses more bandwidth and can be a problem on large layouts. - NCE calls it "old style" consisting, not helpful.

• • The locomotives have no clue they are part of a consist.
  • The Command Station does "all the work" to send commands to each individual loco.
  • The door is open for the Command Station to do intelligent things with lights, sounds and everything.
  • Downside is sending commands to EVERY loco every time can slow the system down. Way too much traffic on a club layout for example.
  • Consists with a large number of locos might tax the memory of the Command Station

3. Decoder-based consisting - all the consist "knowledge" can be in the decoder, the the important part is that the decoder is given a "consist address" via CV19, and the decoders react to that address. There are features to change whether decoders accept specific function commands, i.e. a decoder can be told to ignore a horn command, or a lighting command.Often called Advanced consisting even though the NMRA never uses the term. It has "advanced control" There are additional CV's in the loco decoders so they have an ADDITIONAL "consist address" and set other parameters on how the lights, bell, horn, etc. behave in a consist. Once you have set these parameters in all locomotives, then you can command the consist with the single consist address, which is a short address, and the DCC system does not need to know this is a consist. 

• • the consist address is set in each loco
  • locomotives CAN be set to ignore their "normal" address for safety
  • configuration of which loco acts on functions, like horn, bell, lights can be individually configure as prototype.
  • There is a specialized Consist Control command that provides additional control to the locos in the consist (not all decoders support this properly)

4. Throttle-based consisting - really pretty much the same as #2, but now the "data base" resides in the throttle, not the command stations. Some versions allow sharing/exporting this information between throttles.

• • Some additional features can be utilized, like sharing consist information with another throttle, usually by LCC
  • Otherwise not a great idea since if you don't have sharing of information between throttles then you can have conflicts
  • Needs a mechanism to hand over a consist to another throttle
  • In terms of a system, is going the wrong way, adding memory and function to each throttle, makes the overall system more expensive and complex.
  • Utilized by companies that want to sell their throttles into another manufacturers' system, but clearly cannot integrate it software wise

There are some advantages and disadvantages to each method, but for people that use consisting a lot, advanced consisting is the only thing that makes sense. Thus you want a system that supports advanced consisting if you will run multiple locos. This means that the DCC system should make it EASY to program locos to join and leave a consist, not spending a lot of time programming individual CV's for each loco.

Again, the NCE system is a great example of very good support. There are dedicated buttons on the throttle to create, add to, remove and delete a consist. It supports all 3 consisting modes.

After 5 years, my Zimo system finally is still getting consisting "fleshed out". They only support Universal consisting for now, but I have faith, in that they continuously improve the product, and they have a powerful platform.

Some detailed pro's and con's of various types of consisting:

• Basic / Address Consisting. The only pro's I can find is that any system can do this, since nothing in the system understands that the addresses are "shared" among a number of locos. On the con side: now you can no longer individually run locos, since they share the same address. Also when programming on the main, all the locos will get the same command. This is really dumb and actually goes against the fundamental advantages of remote control of individual locos.

• Universal / Old Style Consisting. Again, the only pro I can find is that any system can do this. Again, the locomotives in the consist have no clue they are in a consist. The con's are many, the biggest is that by having to send every command to every loco in your consist, then a consist of 4 locos will QUADRUPLE the number of commands sent in the system. This usually results in sluggish system response, missed commands, and general limitations of just how many locos and other things can run. Another big negative is that now the system needs to "remember" the consist components, so you start taxing the storage space on the system or throttle. Another negative is that if you don't have a really smart command station, only one throttle may "know about" the consist. It get's worse when you think about the lighting and sounds. Now the system needs to do something smart about which unit sounds the horn and bell. Which unit has the front or rear headlight when. Some systems do not do anything, and when you sound the horn it blows on all locos. Some systems do remember the direction of the consist, so the appropriate horn or bell goes and the right headlight comes on. Even though this CAN be done correctly, it's normally done poorly. The system limitations often rear their ugly heads soon, for example AirWire, often touted as DCC can only have ONE consist per throttle and a maximum of 4 locos in the consist. It does try to handle the lights and sounds correctly If someone wants to run your consist, they cannot just take it over, they need to either take your physical throttle, or they need to "build" the consist on thier throttle. You can easily see that Universal / Old Style consisting is really a crutch.

• Advanced Consisting. This is where DCC shows it's capabilities, and why many people use it. Basically there is a third address "available" in decoders (besides the short address and the long address). This is the consist address. You can easily switch between using the "normal" address for a loco and the consist address. When using the consist address, the system only sends one command to the consist address, so it has the economy of commands that Basic / Address consisting has. But each loco is still individually addressable. Also there are CV settings for EACH loco that tell it how to act when in a consist, so you can tailor the lighting, sounds, momentum, etc. These configurations allow you to have locomotives in the front, rear and middle of a train, and behave properly. Also, since you can still individually control the locos without reprogramming, you can "build" the consist by driving the locomotives to each other and then running the consist as a unit. It's also easy to add and remove locomotives from the consist, and in a similar way to cut on and off a helper loco.

The reality is that any modern system should allow and support advanced consisting.

Some final notes:

Go see one of these systems in action if you can, even if it is on an HO layout.

If you are deciding between Digitrax and NCE, or just want to see them in action, watch this video. (It's pretty accurate, some small mistakes, but overall gives you a good idea) (If you buy a Digitrax after watching this, don't blame me later)