DIY DCC anyone?

ModelRailroadForums.com is a free Model Railroad Discussion Forum and photo gallery. We cover all scales and sizes of model railroads. Online since 2002, it's one of the oldest and largest model railroad forums on the web. Whether you're a master model railroader or just getting started, you'll find something of interest here.


D

DougM

Member
Just wondering if anyone here has tried any of the Do It Yourself DCC projects, especially the command station ones.

Rocrail and OpenDCC look particularly intriguing but I was wondering if anyone has tried any of the DIY DCC stuff.

OpenDCC (In german but look for the english button at the top right)
http://www.opendcc.de/
http://opendcc.sourceforge.net/

RocRail
http://www.rocrail.net/

MiniDCC© - A Digital Command Control Do-It-Yourself Project!
http://www.minidcc.com/

DCC-MB
http://web.syr.edu/~mobrandt/dcc-mb/dccmbhom.htm

MERG
http://www.merg.org.uk/resources/dcc.htm#MERGDCCComponents
 
Last edited by a moderator:
Hi!

I recently built a dozen of the MERG decoders and for a command station/handhelds, I tried the NanoX-s88 and 4xMinimaus handhelds at:

http://www.tinet.org/~fmco/home_en.htm

It works well, but since starting I obtained a Prodigy Advance command station as part of a large purchase. The DIY is not as simple to use as the Prodigy Advance, but building handhelds for the DIY is much cheaper and I have 2 kids...

I would like to reprogram the DIY handhelds to have the simplicity of the Prodigy and add some features that just aren't there but make sense. All this keeping in mind it is for 2 young children. (3 if you count me) But that takes time...

I have also bought the parts to build the interface to the PC and detectors as well. Perhaps in the coming months I will get this done.

If nothing else, it is an interesting site with interesting projects. It has taught me a lot, but it is not for everyone. I would imagine for most, just buying the decoders and command station is cheaper, faster, and easier. But I needed a dozen decoders, 4 or more handhelds, a command station, detectors, decoders for switching, and an interface to a PC. So I went with DIY.

Let us know how you make out! I'd be interested in hearing your results,

Cheers and Good Luck,
Keith
 
I have just completed the MiniDCC and a self designed booster using the L298. It seems to work as specified. At least it will control my locos and Bachman turnouts.

I have also built the MERG Acc5 accessory decoder with the output modified for Tortoise machines. It works OK except with the MiniDCC, after I leave the Route display (where you can also switch individual turnouts), the turnouts that are active (thrown), get reset back to normal.

I am working on a decoder based on the Parallax Propeller MCU. If I can get that to work, it could control 16 or more functions. I am not that familiar with the DCC protocol so its takeing me some time to learn.

Since about the only brand of HO available here is Bachman (they have 2 factory stores here), the only choice I have for a more complicated layout is to do it myself.

Thsi is my first visit to this forum. I hope it is more active than the other one I have seen. At least this one loads much faster. The other one can take 10 minutes to reload a page.

Donald
 


Donald, I think you are the first member we have that lives in China. I take my hat off to all of you guys with enough skill to build your own DCC systems. I have enough trouble just following the manuals on my Digitrax Zephyr. :)
 
UP2CSX said:
Donald, I think you are the first member we have that lives in China. I take my hat off to all of you guys with enough skill to build your own DCC systems. I have enough trouble just following the manuals on my Digitrax Zephyr. :)
Click to expand...

Hi Jim,

After being away from the hobby for some 25 years, I take my hat of to the many who have created such fantastic layouts. In trying to create our layout, I have been amazed and intimidated be the fantastic work and ideas out there. I find the art of modelling far more challenging than the technology. With technology, I can always open up a manual or google a page. But modelling is such an endless array of possibilities.

keith
 
UP2CSX said:
Donald, I think you are the first member we have that lives in China. I take my hat off to all of you guys with enough skill to build your own DCC systems. I have enough trouble just following the manuals on my Digitrax Zephyr. :)
Click to expand...

Thanks Jim
I am impressed with the artists here that do such realistic layouts. I think the electronics is easy compared to the detail other people can put into theirs. Everyone has their different skill sets. For me, building a circuit board is a lot easier than building a realistic mountain but I've been doing that since I was 12.
I love it here in Shanghai. Not as impressed with the other parts of the country. Here is like Manhattan except the streets are safe and the people are nicer. So much construction going on.

My wife has given me the dining room for my 6' x 12' layout. That would never happen in the US.

The forum is the only way I have to correspond about the hobby in English. I do not speak Chinese but my 3 yr. old daughter does. She likes the trains also.

Donald
 
I finally have the code working on the Parallax Propeller MCU working as a decoder. Figuring out the code for Locos wasn't too hard but the accessories took some time. Apparently, not everyone uses the same technique. The turnout operation on the Bachmann was completely different from the MiniDCC. I can now operate up to 24 Tortoise machines. Since I only have about 20 turnouts on my layout, that should be enough. It will also come in handy to "read" other commands later. It can actually do a lot more but I haven't figured out what else to do with it.

If anyone is interest in this, you can PM me.

Donald
 
I have built a few kits over the years. The last decoders that I built was in the late 90s when a friend asked me to build 15 NCE decoders. The kit came with all the resisters, caps, diodes, transisters and chips in separate bags. I just needed a pin point solder iron and a magnifying glass. Prior to that I had built a tethered throttle with momentum and had modified some of the Aristo Craft radio throttles to work with a sound system.
They all come with adequate instructions. Ya just need to know some of the basics of electronics and how to solder.
 
After getting the turnout decoder working, I have started (restarted) on the other projects. While operating my layout the other day, I had a problem with 2 locos entering the 2 reverse loops at the same time. Obviously, having both loops on the same reverser has potential problems. Now that I actually have 3 loops, that may cause more problems.

So, I started working on my boosters again, this time adding reversing capability to all branches except the first one.

My booster circuit is actually very simple. Based on the L298 dual H-bridge modules with a 74LS04 hex inverter feeding the second H-bridge input. I have added a 74LS32 XOR gate to the input to be able to reverse the signal with just 1 input and no relays. The next problem was to generate the reverse signal which could be the same as a "short" signal but with some timing issues. You need to keep from comntinually reversing if there is an actual short so there needs to be some delays. After giving this some thought and looking at the parts count for the opamp/comparator solution, I decided to try the microprocessor approach. The PIC12F683 is a small (8 pin) micro with 1 A/D channel. I figured if I just measure the current through the H-bridge, I could use it as both the short circuit cutout and the reversing signal. This worked out very well. On the first "short" indication, I reverse the signal. If I still have a short, I disable the DC and retry every 1 second for a short time to see if the short still exists. After the short has been cleared for 4-5 seconds, I reactivate the reverse ability.

So, now armed with the actual current measurment capability, I wondered what else I could do with it. I can also use it to indicate block usage if there is any current used.

Then I wondered about using it for ACK detection. Not so good. Too little current change for reliable detection. Back to the comparator solution for that.

That got me to wondering about how to go about adding the decoder feedback to the booster circuit and I got very confused after reading the NMRA RP-9.3.1. Does anyone understand it or use it? How do I design this into a booster. I am assuming this is only needed on the programming track like the simple ACK detector but that may not be the case. If so, I need to rethink my individual block boosters.

All help or discussion is welcome.

Donald
 


Thanks, I had not seen that page and it has lots of ideas. I noticed that it was by Robert Paisley. He was kind enough to email some of the code for his MiniDcc which I had built. I use it to test my booster circuits with. I did not realize that he was still actively designing circuits. I will have to ask him for suggestions as he seemed very friendly and a nice guy.

After doing some searching on the web regarding RP-9.3.1, I see some disagreeent about it in that it may cause incompatibility for existing systems

My goal is to construct a system that will allow networking of throttles (or possibly PC)so that some automation is possible. I still have lots to learn about the intricacies of the DCC protocol but that is what retirement is for.

Donald
 
Wow Donald...some impressive stuff.

I have not gone too far into design at this point. Just this week, I finished all five (instead of 4) of the Minimaus throttles (xpressnet) and the PC interface for the already finished NanoX-s88. The only customization made was to add an emergency loco stop (not emegency power off) to the handheld. I will get back to recoding the handhelds later. For now, the kids are having a great time on their own, and no more fighting for control. They each get their own loco. With the PC or extra controls, I can monitor and take over things when they are about to go bad.
I have yet to build the accessory/turnout decoders, but I have etched the boards and have the parts sorted into bags...soon i hope.

The MERG loco decoders have worked out great. I love the low speed performance. Tuning is a science...but when done, wow. Now I want to add forward and reverse trim and perhaps a couple other bits...

Thanks for the updates.
And Thanks to Doug for the link. I will have a closer look later.

Keith
 
Keith,
Thanks for the info. It has given me a lot of search sources. I had not seen the Minimaus before. Sometimes I cannot get to some websites from here such as the rocrail above. I have not decided which network to use but was leaning toward Loconet as it was well documented. I have now found xpressnet and SRCP which also look interesting. My German is not good enough to make sense out of the S88 stuff though. The schematics I can understand but the dialog goes right over my head. I see I have a lot of studying to do.
 
Thanks Doug.
I have been trying out Rocrail and JMRI. Although I have not spent enough time to choose, I like creating lots of individual throttles in JMRI so as to control many locos simultaneously. DCC is amazing.

Donald, I haven't read alot on s88, but from what I can gather, it is a simple shift register for input (to things like JMRI or Rocrail). You chain together boards stretching across your layout for monitoring things like occupancy sensors, turnout position indicators, whatever. Easy and cheap to build, although I wouldn't be surprised if you get erroneous results now and again based on the design.
Xpressnet is interesting. I have been reading through the document from Lenz describing the command sets on the bus. That is how I found the emergency stop for the current loco and added it to the Minimaus. What I really like is chaining up as many throttles as I can make. And the PC interface works great giving me all the throttles I can fit on a screen and more...

I planned to start my layout soon, and then implement and test occupancy detection and remote dcc turnouts and signals. But since my current layout was an excercise in learning all the do's and don'ts of track laying, yard switching, heights, radii, etc. I should probably figure out the rest on it too.

Not enough hours in the day...
 
Hi Again,

I had a chance tonight to try something I have been curious about. I have a couple of old Axim PDA's with network capability and wanted to try them out as wireless controllers with touch displays. So I fired up Rocrail, setup an ad-hoc network connection from the PDA to PC, and with the java client installed on the PDA, started controlling a couple loco's on the layout.

The response of the PDA was a bit slow (but I think that is because it is an older/slower PDA?) but aside from that, very cool.

Rocrail also does web, etc.

One thing I forgot to mention about the NanoX-s88 controller is it is limited to 16 simultaneous loco's. I don't know if I would ever reach 16, but it is a hard limit. If I try to start a 17th, I get no response until another is stopped.

I can't see myself turning over control to the computer, but the thought of automating some trains while I control a few is very intriguing. It is so far from where I started as a kid with a 'wall' of switches for blocks, 2 locos and 2 throttles.
 
Snowman, I am very impressed. I have been able to download, install and at least look at the JMRI. It looks good if I can just get a controller that will network. I still have not been able to download the Rocrail. I have gotten as much as 38% before it times out.
I do understand the s88 now. Fairly easy to implement, but, like you say, may be prone to errors.
I may try to build the NanoX-s88 as I have the PIC16F628. I just have to get the MAX485 chips.
I have asked for the source to the OpenDCC controller to try and port it to the Parallax Propeller chip. Xpressnet looks very promising. That may be the way to go. It seems that you have been very successful with it. PM me with details when you get a chance.

Donald
 
Snowman, you're way ahead of me on most of that. I do know a little about PDA's though. The smartphone market has killed off the stand alone PDA merket. The Dell Axim is very slow while running Java. The best bet would be a Plam TX, the last of the Palm PDA's. It has a fast processor and 256 Mb's of RAM and will run a form of Java that should work for you. It also has wifi so it is easy to netowrk into a wireless system. They are available on the used market for about $150 but they are really holding up in value because some folks (like me) prefer a phone I just use to make calls and a real pocket computer to do computer things. The Windows Mobile smartphones have much faster processors but the battery life is terrible. You'd be lucky to get an hour out of most of them using it as a controller. A PDA would make a very cool control device but the long term future for any of them in terms of parts and support is dismal at best. :(
 
Hi Jim,

I am not sure about being ahead, as all I am doing is putting together things others have designed/programmed/etc. I'm just a 'dcc kit basher' for the most part.

I have been itching to try out the PDA since I started putting this stuff together. I know it is not really practical, but the idea of a touch screen has been really appealing. Having endless 'switches' and 'knobs' all in the palm of your hand as you walk around and without having to run dozens and dozens of wires...

But I do like using a regular throttle, digital or otherwise. I like knowing when I (or the kids) drop it, I can pick it up and carry on. Not so with an expensive PDA. And of course the batteries in a PDA always strategically die 1 second before disaster...

Every day I am a bit more amazed at what the community has created and made available. Just when I think I've nailed down all the links, I find more. The whole computer/electronic side to the modelling appears endless.

But now that I have a better feel for the electronics, I am ready to start a real layout with provisions for all the details/wiring for turnouts, signalling, detection, etc, etc. Kind of like finishing your basement and knowing what wires to stick in the walls for your speakers, cable, network, phone, etc.

On a positive note regarding the PDA's. Networking and bluetooth should be a given as time goes on. Java shouldn't be going anywhere soon. Battery life and CPU power generally improve over time. And it's not like you need a bunch. I have a palm kicking around. Maybe I will try it in the coming days. Thanks for the suggestion.

I wish I had a better imagination so I could better exploit all of this stuff. But at least my kids see the potential in all of this. They seem to take what I have showed them and found better/different ways to combine and apply it.

I will let you know how the palm works out.

thanks
keith
 


With great thanks to Snowman for his assistance and the link to Paco's Website, I have managed to build more units including the NanoX-s88 and sucessfully getting it connected to my PC. That site has some amazing stuff.

I have a couple of questions, though. I have 2 reverse loops whose turnouts are presently controlled via photocell connected to flip-flops. That makes them change automactically when the train is in the loop. They currently are NOT connected to the DCC decoder as I do not need to manually control them. Should I connect them to the decoder anyway? I am using Tortoise machines so getting the sensor position to Rocrail is not a problem. I just don't know if there is a need to manually control them.

Secondly, I currently pair some of the turnouts together such as on the double crossovers and on passing sidings so that they operate simultaneously. Do I need to seperate them for use with Rocrail or other PC control?

Donald
 
You must log in or register to reply here.




Affiliate Disclosure: We may receive a commision from some of the links and ads shown on this website (Learn More Here)

Back
Top