Talking Electronics DIY DCC

[PatB]

Has anyone had a play with this cheap and simple, albeit ultra-basic, DCC controller? I recognise that it's not even close to a serious system, but it looks like it might be an interesting exercise to build and would be useful for, eg, an Inglenook or Timesaver by providing the running advantages of DCC (full track voltage, ability to use a decoder tuned to a particular loco if you can get someone with a proper system) to do the programming.

[RedgateModels]

You know, that would make a great controller for a "kiddies" layout at model railway exhibitions ....

[Robin2]

I don't see any link to the program code for the PIC 12F629 so I guess you have to buy it ready programmed from the guy who created the design. That means you are completely stuck if it does not work, and you can't add any enhancements.

 

You can build DCC encoders and decoders with an Arduino - and have as much or as little capability as you want.

 

...R

[imt]

 

You can build DCC encoders and decoders with an Arduino - and have as much or as little capability as you want.

 

...R

 

Thanks - there's something I didn't know and I'd always wanted to try to program something myself (no NOT JMRI!) and this is great news.  Any details on how I lay my hands on the necessary?

[RedgateModels]

Just spoken (via email) to the designer/seller and there is one major limitation, it has to have two fixed DCC addresses programmed into the PIC. Still makes it workable for two engines in steam on a kiddies exhibition layout and I guess if you had 03 as one of the addresses would be ok to test a new installation etc ....

[melmerby]

I can't see why two 1N4004s in parallel would be a good choice of diodes for the bridge.

You can buy 3A diodes for less than twice the price for 1x 1N4004!

 

Keith

Edited August 15, 2018 by melmerby

[Robin2]

Thanks - there's something I didn't know and I'd always wanted to try to program something myself (no NOT JMRI!) and this is great news.  Any details on how I lay my hands on the necessary?

I think the best thing to do is Google "Arduino DCC encoder" or "Arduino DCC decoder".

 

I wrote a simple decoder program a few years back but I'm not sure I can find it now.

 

If you want to make an encoder I reckon that all you need apart from the Arduino is a power supply and a h-bridge module or chip. I am no longer using DCC myself but the Infineon TLE5206 h-bridge chip that I have been using for a club DC project should be appropriate. It has a max of 5 amps and has short-circuit protection. Of course you also need some means to give control inputs - maybe a keypad or a PC linked to the Arduino.

 

...R

[Robin2]

I can't see why two 1N4004s in parallel would be a good choice of diodes for the bridge.

You can buy 3A diodes for less than twice the price for 1x 1N4004!

I hadn't noticed that. It seems so much simpler just to buy a full-wave rectifier as a single unit - all the diodes in one package. I have 6amp 600v devices that cost about 30p each

 

...R

[imt]

I think the best thing to do is Google "Arduino DCC encoder" or "Arduino DCC decoder".

 

I wrote a simple decoder program a few years back but I'm not sure I can find it now.

 

If you want to make an encoder I reckon that all you need apart from the Arduino is a power supply and a h-bridge module or chip. I am no longer using DCC myself but the Infineon TLE5206 h-bridge chip that I have been using for a club DC project should be appropriate. It has a max of 5 amps and has short-circuit protection. Of course you also need some means to give control inputs - maybe a keypad or a PC linked to the Arduino.

 

...R

 

Thank you for your time.  I have already built a control system using NCE Mini-panels and a "one button route setting panel" for my layout.  It was fun and works a treat.  I was wanting to look at taking it further - but that needs a DCC encoding system - either on a chip or in a program - and a method of injecting the output onto the accessory bus.  Your info will help me dig further.

[RedgateModels]

OK, got a kit on the way, thought for 16 quid delivered from Oz it's worth a punt for a little project. Not sure what for yet, but thoughts are initially for either a stand alone test track or simple controller for the kids layout at next year's Mansfield Show .....

[Ian Morgan]

Also search for DCC++ for an open source Arduino based command station.

[RedgateModels]

OK, got a kit on the way, thought for 16 quid delivered from Oz it's worth a punt for a little project. Not sure what for yet, but thoughts are initially for either a stand alone test track or simple controller for the kids layout at next year's Mansfield Show .....

 

Quick correction, the price was in USD not AUD, so that makes it about 22 quid

[Ian Morgan]

I quite like the idea of the TalkingElectronics controller finding the DCC addresses of two locomotives. That is quite neat, and saves the need for any sort of GUI, keypad or display.

[RedgateModels]

I agree, but I don't think it works like that any more, here's what the designer/seller told me today

 

 

 

The "find" section has not been programmed into the chip and the parts are not fitted. 

You can change any of your locos to any address via your controller. 

 

You order the kit with two addresses pre programmed in, I picked 1 and 3

[slngshotx]

I'm sure folks will have seen this, it's a DCC++ controller that controls a DCC++ basestation running on an Arduino, think the whole thing is about $40 to build

 

http://trainelectronics.com/DCC_Arduino/DCC++/Throttle/

 

If you go to the main web page http://trainelectronics.com/miscellaneous_projects.htm he has loads of projects, both DCC and DC. Probably any of them would be a good intro to electronics, and any arduino ones will probably give you a start on coding.

[Robin2]

I'm sure folks will have seen this, it's a DCC++ controller that controls a DCC++ basestation running on an Arduino, think the whole thing is about $40 to build

I figure the parts (excluding the case) would be about £20 if bought from a UK supplier.

 

But if I was building something equivalent I would use my laptop as the input device and save the cost of the screen and keypad. Or add on WiFi or Bluetooth and control it from a smart phone

 

...R

[PatB]

OK, got a kit on the way, thought for 16 quid delivered from Oz it's worth a punt for a little project. Not sure what for yet, but thoughts are initially for either a stand alone test track or simple controller for the kids layout at next year's Mansfield Show .....

 

Do please keep us posted as to how you get on.

[RedgateModels]

Chap said about 8 days airmail, I then of course need to find the time to build it LOL

[PatB]

Ironically, having started this thread, I've just used part of the proceeds of a (non-modelling) Ebay sale to order an Arduino Mega and a motor shield in order to have a go with DCC++. A DCC system for <$20 has a certain tightwad appeal .

[RedgateModels]

I've got a kit, just need to buy the time to build it LOL Plan is to build it into George's layout - will hopefully make it work better at the next Mansfield Show

[Oldddudders]

I know I’m not the brightest when it comes to matters electronic, but DIY DCC would seem to need NMRA-compliant soft/firmware to be of any reliable use with commercial decoders. I assume DCC manufacturers each have their own program - how does this work for DIY-ers?

[Right Away]

I can appreciate those who wish to go down their own DIY-DCC route. What would concern me is the potential risk to expensive decoders should a fault develop within the system.

[RedgateModels]

Well for me it will only be running a couple of locos chipped up with cheap Hornby or maybe Hatton's entry level decoders, so no real biggie

[Nigelcliffe]

I know I’m not the brightest when it comes to matters electronic, but DIY DCC would seem to need NMRA-compliant soft/firmware to be of any reliable use with commercial decoders. I assume DCC manufacturers each have their own program - how does this work for DIY-ers?

 

Exactly the same as commercial makers.  

There is a published (NMRA) specification for the DCC signal (shape of wave, encoding of data in the bits transmitted).  From that, write something which sends the correct bits in the correct order, at the correct voltage and frequency.  Within the standard, there are the absolute basics which are compulsory (or its not a DCC system!), and then there are numerous optional elements for the command station.  Optional parts might be "does it support accessory devices?",  "how many functions does it support?",  "which speed steps are supported",  "does it support aspect signalling (only a handful of commercial systems support this)"  and so on.      

 

But, writing ones own from the bottom isn't necessary.   For those who want to DIY, there are a lot of shortcuts because someone else has already written the software and made it publically available.   So, for the Arduino based "DCC++" system, you can buy an Arduino processor board for a few pounds, download the software from the internet and load that yourself onto the Ardunino.   Add the output power board (to get enough track current to run a loco).   Plug in, and it will work.   Or, if you want to start modifying the code to support some additional features, then you have the existing code as a starting point. 

 

Or for MERG members, there are several MERG designs which are available (some long withdrawn from the MERG kit range, but the design is still available), including their code.

 

 

Bear in mind that most commercial makers are, in reality, "one man bands" - there is one person at the company who actually knows how their system really works at the detailed could change things level.   Even the "big names" only employ a few dozen staff, and that includes office admin, packaging, stock control, customer support line, ordering, etc.. 

 

 

 

As for damaging decoders, well, yes possible in the extreme.  But that risk is also present in a commercial system if the output components are wrong, or the power supply is wrong. 

But, if someone understands the absolute basics of electricity and the output to the track is within the acceptable range for a DCC decoder, then damage isn't going to happen.  That output is very simple electronics, it can be a standard output/power module bought from somewhere.   The decoder might not respond because the DIY design is sending wrong instructions, but the decoder won't be damaged by the wrong instructions.  

[Oldddudders]

Thanks, Nigel. I might have guessed that since you chaps make JMRI available gratis, someone might have done the same for basic DCC. And I certainly understand the bolt-on nature of some facilities. My first DCS100 didn’t really cope with functions beyond f4, the later one goes to f28 I think. For silent layouts the first one remains perfectly adequate after 20 years.