I am new to DCC but I am handy with electronics and programing. I would like to create a DCC system without spending alot of $$$. I assume most DIY people buy the decoders and then create the control station via building one or using a pc. What are the do’s and dont’s to this systems, ideas, compatability issues. Thanks
At this point it’s probably a lot less expensive to use off-the-shelf commercial systems than to do it yourself.
Put it this way: is it cheaper to buy a video card for your computer, or to buy a prototyping board and design and assemble your own video card using off-the-shelf components? It used to be cheaper to build your own, about 30 years ago when you were building, say, a serial port controller for an Altair 8088. Similarly, electronics hobbyists used to design their own command/control system, based on an Apple II or TRS-80 (I recall one series called the CTC) but they weren’t easy or cheap.
Unless your idea of “handy with electronics and programming” includes designing your own microprocessors, it’s probably no great advantage to build a DCC system from scratch. There are lots of ways to expand, program and otherwise more fully utilize off-the-shelf DCC systems than to try to homebrew 'em. Space is also a consideration–microminiature components and custom chip sets are part of why you can fit a DCC decoder into an N scale locomotive. Homebrewed circuit boards take up more space.
See my post over in the Layouts section to the fellow who wants to build a route control system for turnouts. While DIY DCC is entirely possible (See MERG for some examples), it won;t be any cheaper than a decent commecrial system. What might be better is to start with a commercial system and DIY some of the accessories and add-ons. Again I will use myself as the example here. I have a Digitrax system, but a lot of my control will be via Hans Deloof’s LocoIO board which is a 16 bit universal I/O board. I can put one together for about $20, maybe less if I an using for a specific purpose and elminiate the phone jacks used to make up each group of 4 ports and use my old computer power supplies, eliminating the rectifier and voltage regulator circuitry on each one. Likewise, Hans also has a nice 3 amp booster design which would be fairly inexensive to build. I’ve only stayed away so far because it uses a different PIC processor than the LocoIO/LocoBuffer circuits and to program that (he provides the code) I would need to buy, borrow, or build a programmer for that type of chip.
And Hans isn’t the only one. There are many interesting DIY projects built around Digitrax. John Plocher tackes the LocoIO approach to the next step and has designed a controller for Tortoise switch motors that provides control, actual position feedback, and block detection of 4 turnout sections. And a 16 port block detector that should be definitely lower cost than the BDL-168.
MERG has some DIY accessories that use the track signal, and thus work on any system, but I haven’t seen much DIY material for Lenz or NCE.
–Randy
I think once you do the research you’ll decide you don’t want to do this, but…
For info on the DCC specifications go here:
http://www.nmra.org/standards/consist.html#standards
There are several relevant DCC documents.
The compatibility issue only applies to the signal to the rails, and now the decoders interpret it, but not necessarily what they do (specifically thinking of decoders that have unique CVs, sound, etc.), but the standards should tell you what you need. Throttles and command stations do not have to be compatible with each other between vendors. The Digitrax and CVP websites have some pretty good info on how they do things. Also, you can look at the JMRI site for lots of stuff. There is also a gizmo called SPROG http://bbmgroup.home.comcast.net/sprog/
that can interface with DecoderPro and run a layout, at least to a degree.
Look at all that, and see what you really want to do!
WOW! Lots of information to absorb. I have seen some decoders on ebay sell from .99 cents to $14. Do I have to be weary about stuff like that and what would be a safe system or brand to stick with so that I can deal with compatability and expansion issues. What is the issue with an amp booster. If the track voltage is constant why cant you just run a regulated step down transformer? You can get a 3amp transformer at Radio Shack for $10. What are you using a switching power supply(p/c) for?
P.S. Is there a collision detection system available with DCC?
Since you are willing to buy decoders, what you have proposed is feasible (depending on your engineering and programming skills and patience). It has been done. The NMRA specs for DCC can be found at http://www.nmra.org/standards/consist.html. Note that the specs only address the command station to locomotive signals and communications. Anything on the user side of the command station (throttle network, throttle programming) is totally up to the developer, and there are a number of commercial propritary standards used here. Both sides of the command station must operate in a noisy environment that abounds with non-twisted parallel conductors; hence data rates tend to be low. Also short circuits on the track are common, and must be dealt with without ruining equipment.
You have to design, build, and debug a command station that merges the command signal with the power signal, and applies it to track that has an impedance varying anywhere from a few ohms to infinite.
You have to design, build, and debug a throttle to generate the commands, and a data link between the throttle and command station.
http://jmri.sourceforge.net/ has a lot of ideas that might help you in your quest. There is information about system components that work with Digitrax’s Loconet at http://embeddedloconet.sourceforge.net/
Hope this helps
Fred W
Yep.
Mark I Eyeball. Has the advantage of being 100% compatible with DC operations, too.
Digitrax uses collision detection/avoidence for their Loconet cab bus.
Jim
The track voltage is constant but it’s not DC. It’s a square wave signal that actually carries the DCC commands as well as provides the operating power. And when it comes to high current supplies, a switchign power supply is lighter and cheaper and more efficient than just using a massive transformer and a regulator. Good example - my first computer I built froma kit. It uses a standard regulated power supply, it puts out I think 3 amps at 5 volts and maybe 1 amp at 12 volts. I can take the switching power supplies from my present computers, that put out somehwere around 45 amps at 5 volts not to mention a ton at 12 volts and 3.3 volts, and three of them together are lighter than that one 3 amp supply.
The NMRA DCC standard applies to the track signal. Any sysem putting out NMRA DCC can control any NMRA DCC decoder. This compatibility does NOT extend to the throttle side. You can’t use a Lenz throttle on a Digitrax system.
The problem with the big 10 amp boosters is that 10 amps at 15 volts or so is plenty to arc weld things, like wheels to track. This is why they make circuit breakers to place between the booster and the track, and you break your layout up into sections called power districts. You have the overall 10 amps to drive your layout, but each section is individually limited to a safe current level - if any given section will only see 2-3 trains at a time, there’s no reason
If you really want to waste your time/money, here is a link for DIY DCC stuff:
http://www.cmlelectronics.co.uk/dcc/makingown.htm
I once saw an old 286 CPU converted to a DCC command station - All the guy could do was control a single engine and he had all kinds of extra stuff hooked up to it…
For about $150 on up, you can get some really nice starter systems from NCE(PowerCab) and Digitrax(Zephyr). I doubt you could build anything with as much function for the price. If you want a functional ‘kit’, check out CVP:
Jim Bernier
If you want to DIY, what do you think of a non-DCC system? I think it would be cool to design a system based on Bluetooth or Wireless-B. Decoders might be tricky because of their size, but a wireless PDA could be used for the control. I actually did a chip-based brushless DC motor control for a jetliner - VERY COOL! Of course, it could lift an automobile 
Just an out-of-the-box thought…
Jim
Jim22- Yea, I think if I had the smarts to build my own, I try and do something along the lines of a TV style remote control. I’m pretty amazed at the range of TV and car alarm devices these days.
thanks Randy , i haven’t seen John’s site before . looks like a very interesting project .
the CVP EasyDCC system was originally done as a series of diy articles by Keith Gutierrez in the feb to july 1997 issues of Model Railroader (years before that he also wrote a series of articles on his CTC-16 system , a pre-dcc command control setup) . i think all of the parts are still available from CVP , it should be possible to order just the hard to get stuff if you think you can source parts at a better price . i also think it’s pretty amazing that Keith was willing to share all the details about how to build his systems with the rest of us
If you ask me, I’d tell you to purchase the stuff already made. I also went thru the build it yourself phase and most of the systems I built didn’t work too well or work at all. After sending a lot of my hard earned money to places like Mouser, Digi-key, and Radio Shack (not to mention the hours upon hours spent soldering the components together) I found that they didn’t work, I realized the pre-made stuff is the only way to go. All it takes is a small piece of solder to get across a circuit or one connection soldered in the wrong place and the entire thing won’t work. It taught me a valuble lesson. I now just buy the stuff someone makes for a living. The only do-it-yourself circuit I use now is a simple one that converts 12 volts to 5 or 1.5 volts. Everything else is manufactured. It saves a lot of money in the long run to purchase electronic components that just need a few wires soldered in place and is ready to go.
CWCLARK, you actually built that CTC-16 system? Crazy man.
It seems it took a lot more dedication to be a “serious” modeller back then. Wow.
I remember hearing about that system in MR, but wondered how many people actually built that stuff.
Like you say, it only takes one little error.
Like when I was a kid, a friend of mine and I would sit there for 3 hours trying to type some “game” program into my commodore 64 from a book. In the end, there would be some small error that was actually wrong in the book. We’d never get it to work.
That wasn’t quite as much a DIY series as a “how do I assemble these kits” series, much the same as the Railcommand articles a few years before. The last real DIY command control series was the CTC-16e, and the CTC-16 before it. As far as I know, the only thing you can buy from CVP that isn’t assembled is the Booster 10, and the kit is only like $20 less than the assembled version. I don’t believe the command station is available as a kit anymore.
–Randy
That’s out there too - there’s a DIY project for Loconet that has the option of using a TV remote for a throttle. It even supports selecting new locos without plugging in.
http://embeddedloconet.sourceforge.net/
–Randy
You might also be iinterested in this one: http://loconetovertcp.sourceforge.net/ which allows access to the Digitrax Loconet from any computer on a TCP/IP network. You could theoretically control and monitor your layout from anywhere on the internet, but the real purpose is to have a single interface to the layout and have multple client apps connect into it, more along the lien sof, you have an internet conenciton, you can use a web browser to view web pages, an ftp program to transfer files, or an email program to read mail.
–Randy
Ok, I agree with all of you! DIY is a pain sometimes. I am an engineer so it’s par for the course. So, in an effort to keep costs down, ie buying used items, does anyone have good ideas for what is the best bang for your buck. Is using a p/c as a control station a good idea?