Advice on designing a replacement circuit board for conventional trains

Perhaps you could start with telling us the basic structure you envision.  What capabilities is this device to have?  Are you varying track voltage to control the trains as in conventional mode, or are you running constant track voltage like command & control systems?

I'm having trouble visualizing what kind of architecture this will be.

Thank you Flash - those are awesome answers - thank you.

Gunrunner - The architecture is that you would have your conventional transformer turned all the way up at full constant voltage (as in a TMCC). The circuit board would exist between your power and your motor. The board is controlled by an iphone app. The benefits are it can offer Command Control style speed and accessory control, along with gameplay and control an iphone/ipad offers, and can be inserted in any train with a can motor (Pullmor would need its own separate board).

I initially decided to do this because I'd built this big layout with elevations and soon learned that conventional trains have speed management issues (difficulty maintaining constant speed going up and down hills) which can result in crashes.  The the expense of moving up to the Command Controls is considerable, the remotes seem a bit unwieldy, and in general it just feels like a commitment to go CommandControl both financially and technically. I wanted to make an affordable board you could just splice into any old train and get the same benefits (plus everything iphones/ipads let you do). 

We have a functioning board, but I need to make a rev2 that supports larger engines/power and accessories.

Again - I really appreciate info or feedback.

Well, you talk of the iPhone controlling it, what is the actual interface to the board, WiFi?

As far as accessory outputs, I'd consider having at a minimum, the mix that is supplied by the Lionel basic TMCC board, though you probably don't plan on supporting electrocouplers.

Front headlight, rear headlights, smoke unit control.  In the "nice to have
category would be marker lights and cab lights.  If you provide a 12V at 400ma, which is what the Lionel TMCC boards have, you can run most lighting schemes from that.  The smoke unit triac handles a couple of amps.

You'd do well to examine the capabilities of existing command locomotives, both TMCC/Legacy and DCS, and then make the tradeoffs.

I'm curious about your comment about the expense of moving to command, since you'll have to equip every locomotive with your board anyway.  You can equip an AC or DC locomotive for command with the ERR AC or DC Commander for as little as around $60. If you want sound, you have to add more money.  There are inexpensive ways to utilize your existing whistle for the ERR stuff as well.

You'll have to certainly be very competitive with that price point to be successful.

Thanks again Gunrunner Joe. Great info. I really appreciate it.

You buy a Lionel starter train for $229. Then it seemed like it was going to be about $400 additional to put together a TMCC system.  I played with the TMCC remote controls and they seemed primitive in terms of interface. There is SO much fun you can add when you throw an iphone/ipad into the mix and have well-programmed interfaces.

Now your train can have any simulated gameplay you want - tilt or touch control, simulations, automatic behavior, blow the whistle when it encounters a hill or when it gets near you - just a lot of fun stuff.  When I add a Christmas train or another HO layout I want to be able to still have this improved experience without buying another TMCC or DCC system for each.

So this project is trying to avoid that initial startup cost - so someone can just take any old conventional train and have a greatly improve the experience. It would be TMCC compatible (in that it could run on a TMCC layout) so TMCC users could have the same fun.

If TMCC offered iphone app control and no startup cost then I feel like this product would be competing directly with the $60 AcDcCommander.

Thanks again for the great info.  I appreciate it.

Well, you do have to remember that most of the audience that will open up a locomotive and install your kit will have a lot more than one locomotive.  So, spread that Legacy system cost over my 40 locomotives, and it's about $10/ea for the basic capability of command/control in addition to the locomotives.

So, if you are using WiFi to control a train, each train must have its own WiFi network interface with individual MAC address.  I will therefore assume that each train will be assigned its own individual IP address correct?

Each train controller would then have to have some sort of memory module to store network IP info and command info, plus a small processor to process packets sent and received from the train and a simple OS for configuration changes.

What kind of hardware are you using? Something similar to a Raspberry Pi?

For multiple train operation, what might the network map look like?

Or am I way off base?

I would also think WiFi would be the most "natural" fit for this but you probably could actually use bluetooth as well to pair the train to the device.  It would probably only breakdown as a wireless transport on really large layouts.

There were a lot of advantages in wireless, but we ended up going with bluetooth4 on this board.  The pairing is very simple for the user, multiple trains simultaneously, good range.

When I get it demoable I will post a YouTube video for you guys.