It is neat.  You should be able to sell a lot of these.  I was pondering how folks like their smoke with a lot of the recent upgrades for smoke posted.  The problem with the fan driven Lionel upgrades is they really don't puff and have the full synchronization that MTH has with the integrated approach.  And you certainly don't get the flexibility to change chuff rates easily.

Adding a selection for 1, 2 or 4 puffs if possible would make this a neat additive for upgrades to TMCC if cost is reasonable.  G

I did think about programmable chuffs, but I can't figure how to do that properly. I'm still dependent on the mechanical or reed chuff switch for timing.  The problem is that as speed changes, you'd be hearing odd timing on the chuffs, and I suspect that would be a real turn-off.  If someone has a neat idea of how to generate multiple chuffs from one switch closure and keep the timing right, I'm all ears, I can't imagine how that would work.

Once you get running at a steady speed, it's pretty easy in software to generate the chuffs, it's getting to the speed without it sounding bad that's the hard part.

That being the case, I figure that I'll do some neat stuff that can be done properly with what I have to work with.  There's always the next project if I come up with a way to properly address programmable chuff rates.  One way would be to use an encoder on the flywheel and count pulses like MTH, we know that works...

That looks great! I don't know enough to offer any suggestions, wish I did. I have read a couple articles on the PIC chips in Garden Railways and they are pretty neat little chips. Had no idea the data sheet was 300 pages though, that probably makes it way over my head. They sounded a bit simpler in the articles. Thanks for sharing the project with us and I hope you keep posting your latest developments.  

Provide an option to leave the cab light on when in motion.  I'm not a fan of the cab light going off when in motion, perhaps it's due to all my stuff being old enough that if there's track power, there is a cab light, and I like it that way.  Maybe you can see if you can sense direction, then you could also change the lamp colors from green to red on the front and red to green on the back when the loco goes backwards.  Or even a way to program marker lamp color options, off/green/white, for the train status.  I think that would get a bunch of the scale guys on board.

Nice work John. I am guessing you are using the chuff switch to sense motion but how do you implement rule 17 lighting? If I understand the rule correctly the headlight is dimmed in response to an oncomming train or maybe a few other conditions. I don't see a wire that might be connected to serial data.

Pete

You should download the datasheet, it makes for good bedtime reading.

PIC12(L)F1501 8-Pin Flash, 8-Bit Microcontrollers Data Sheet (04/02/2014)

If nothing else, it'll mess with your brain!

sinclair, I am doing the cab light the same way that Legacy does them, it seemed to make the most sense.  If you want to leave it on when you're moving, probably best to simply wire it to power and not through this board.

Sensing direction would take more inputs, that's not in the cards for this round.  However, I have the basis for a more involved controller that will have expanded capabilities, I figured to get my feet wet doing this one as it's a feature set that seems to be in demand, and it's my first PIC project.

Pete, the Rule 17 isn't done correctly on any model that I'm aware of, Legacy and DCS just dim it when the engine is stopped and crank it up when they start moving.  Trying to dim it for oncoming trains is WAY outside the realm of reasonability with present capabilities.  I can't even imagine what that would take to do properly, but suffice to say that I'm pretty sure none of the control systems have the capability do to that as of now, and I'm sure this little project won't be trying!

Speaking of that, looking at the diagram are you capturing motion via the chuff switch?  I don't see any motor leads as inputs.  Also what dims the light, does the pic output have the ability to cut voltage in half?  G

On my TMCC conversions I like to have an indicator light to let me know that the smoke unit is on so that I don't run it dry. I usually add a red smoke box light in the cab when the smoke is on. I would be open to any other ideas. It would be ideal if the smoke shut off when the fluid ran out.

Yep, if the locomotive is chuffing, I figure it's still moving.  When it stops for two seconds (arbitrary value I selected) I figure it's stopped. Obviously, I can change that value to anything I like, but two seconds seems like a decent value for now.

I also selected a value of 100ms for the smoke fan to run for each chuff, it looks good on my bench test, but again I'll see how it looks in a real locomotive and fine tune it if necessary.  When the chuffs get to be 100ms apart, the fan run continuously.

The headlight is an LED directly driven by the PIC port and I use PWM to change the intensity of the headlight.  I figure when I build my first prototype to go into an engine I'll fine tune the PWM parameters, right now I'm using a pulse width of 15/256 for the low intensity and 255/256 for full intensity.  It looks pretty good on my board, but I might make the low a bit brighter when I see it in place in a locomotive.

Speaking of motor leads, in the future I'm thinking of coming up with a motor connection that will not only tell me when I'm moving, but from the polarity I'll be able to tell the direction as well.  That's for a future project.  I could do it with two opto-couplers, but maybe I can come up with something a bit neater.  It will have to be totally isolated, so there'll have to be optics in there at some point.  Basically, it would look like a dual circuit for the existing headlight input with both polarities covered, and they'd feed two ports the way I see it now.

I have no way of sensing the fluid level in the smoke unit, so I'd have an issue there. 

 John...

 Do you think you could sense fluid level by current draw of the element ?

 The less fluid the more current draw ?  might be something to bench test one

 day " in your spare time "..  Just a thought....

TAS used a thermistor to signal when temp got higher and that change in resistance based on the temp increase would shutoff the power to the smoke element.   G

That's way beyond what I'm doing this pass.  I don't know that current draw is a reliable indicator in any case, and I'd actually expect it to go lower when the element is dry.  TAS used to sense it using a thermistor next to the heater, when the temperature started to rise, they knew the fluid was getting low.  I have no idea how well that worked.  Truthfully, I just keep an eye on the locomotive and add fluid when the smoke tails off.

George, I see we were typing at the same time.

I think I see R2D2 on there  

Seriously, it looks like you are making good progress. It would be neat to have your own boards made (and know how to go about laying them out like that so you could have them made).  

I'm fine tuning the design, a couple of the components were the wrong size footprints.

I also decided on another "feature", when the locomotive stops, I'm going to run the fan on a lower voltage so the idle smoke comes out slower and not like the full speed running.  All it took is swapping a couple of pins and programming up the PWM for the smoke motor output.  I think that'll be another little nicety that will enhance the function.  At this point, I'm trying to look it over carefully and see if there's anything I missed, then I'll look for a place to get the boards at a reasonable price.

Sign me up,

but where will it be installed and how? I don't know squat about electronics.

I think there is a much simplier way to get chuffing smoke by installing either a diode or capacitor into the circuit between the smoke unit and the TMCC board that controls the smoke unit. One of the Lionel Service Stations has been doing this for a couple of years now.

Also it was mentioned that ERR is now suppling RS5 sound boards but there was nothing about this on their web site.

This is a solution, but it still is not ideal.  It stops the fan when the switch closes to generate the chuff.  This limits you to what ever the manufacture chuff rate is (normally one). 

It doesn't dynamically stop the fan like MTH does which gives a very distinctive puff of smoke on the chuff, it is not adjustable, and you run the chance of stopping the engine when the switch is closed which keeps the fan off despite the heater still being on.  G

These are installed in the locomotive and there is some wiring involved by necessity.

As far as adding a diode to mux the chuff, all I can say it "good luck".  If you use the new RailSounds 5 board from ERR, that is no longer a solution as both chuff inputs must be totally isolated from frame ground.  The chuff switch in that installation MUST be totally isolated from any other circuit, which is why my design has an opto-coupler to output the chuff. 

I also don't see where that diode runs the fan at idle, or provides the enhanced lighting effects I'm adding, but I guess I just don't understand the diode fix well enough.

Nice work John, very impressive.

Thanks, I hope to have some built soon, I'm looking for board suppliers now.

John,

This is impressive! Are you planning to make and sell these units when you've worked the kinks out?

I will most likely sell some, I've yet to see how much it's actually going to cost to buy all the parts, not to mention assembly.  That may limit my sales potential.  I'm assembling the BOM now, but I still don't know what boards will cost.

On the ERR site there is no mention of RS5. Is it somewhere else on their site?

The new single board RailSounds Commander is RS5 based.  I got that direct from Jon Z. at ERR, and using the new boards they do have the improved sound as well as the expanded crew-talk of RS5.  The difference over the older 2 board RS Commander is pretty obvious when you use them.

As far as my comment about the reed switch, you need to actually read the RS Commander manual. 

Clearly, the "simple" fix of using a diode to the smoke unit motor is not possible with these boards.

John,

This looks like a great project.  I cannot wait till it is released to manufacturing! 

John,

Your new board is great. May need to send some work your way.

I really like the sounds from my RS5 engines, but I don't like how Lionel set up the steam whistle.  You have no control over the whistle; when you press the whistle button you get one of three preset sounds.  i have found no way to sound forward, backing, or crossing signals. Does the new ERR boards have this same RS5 feature?

Thanks. Ron

TMCC control of the whistle is fairly limited, it's always been that way.

Still working on the board, I have a hand-wired prototype that I'm working out a couple of small issues with, then I can make some boards.   A couple of things surfaced when I moved from the prototype board to actual stand-alone components.

John,

You get this working and I'll definitely be sending you some work.  Would love some of these features in my JLC engines.

Enjoy your posts because I learn something everytime and you aren't afraid to expand into new horizons.

Awesome work!

Thanks guys, it's getting closer.  I'm trying to make sure all the issues are resolved before I spend the money to get the boards fabricated, it's more expensive than it looks.

Well, it's in the hands of the vendors now!  I ordered the parts and the boards for 10 sets, it'll be interesting to see how long it takes to get them.  Digikey has already confirmed shipping for the parts, so I'll have those long before the boards.  It wasn't as bad as I thought, $12/ea shipped for parts and boards.  Now I just have to see how long it takes me to assemble them when I get them, it's got to be faster than my hand wiring of the initial prototype!

I ended up using DipTrace software for the schematic capture and board layout.  Not perfect, but I'm getting better at creating parts outlines so I can do a board layout.  I wish they had more parts in the libraries, it would make the task easier.

Here's the prototype that I built to make sure the circuit worked.  Good thing as I left out a couple of things, glad I took the time!  The parts are obviously different as the real board is almost all surface mount, but the functions are the same.

I'm going to put this into a locomotive, might as well get some use out of it.

John fantastic work, looking forward to see your prototype in operation

Alex

As soon as I figure out what I have to stick it in.  I may have to do one of my pending upgrades to have a test victim.

I did order all the stuff for the real boards, and I'll have everything but the boards tomorrow it looks like.  I'm not sure when the board house will ship the boards, they're the pacing item.  It sounded like they have a pretty quick turn, we'll see...

Now that it's run on the bench, I'm kinda' eager to see it in a real locomotive.

Welcome to the club! I designed the TAS Puff 'n' Chuff circuit board PIC and software along with Lou Niederlander who did board layout. If I remember right it had the thermistor overheat control, switch input for chuff and  when the engine wasn't moving the smoke level was cut back. It also had steam or diesel mode. Diesel simply didn't chuff. 

I did work on a multiple chuffing circuit for somebody. You can divide the time between switch inputs for mulitple chuffs (2, 4 or whatever) but you have to guess until you get the second one.

I posted a thread in the general forum with a brief test video of the prototype board.

I don't control the smoke element power, so cutting back the heat isn't an option, but I do drop the fan speed.  The thermistor would be nice, but that not only requires controlling the heater, but also modifications to the smoke unit.  I didn't want to go there, maybe next time.

I thought about the multiple chuffing, but as the speed changes, it would be out of sync until it caught up, and that would sound very odd.  One way that I considered, and may yet do in the future, is to do a moving average and generate the chuffs from that.  That would allow smooth transitions between speeds without the spacing between chuffs suddenly changing in a step.

All things are possible, I just figured to start with a basic capability that I, and apparently quite a few other folks, would like to have.   I'm also getting up to speed on how to use the board layout tools, and to do a proper layout for manufacturability.  I'd rather learn those lessons on a small project and then move up the scale.

From the video thread, your regulator is apparently getting quite hot(?). 

In looking at the 5V regulator loads, you may be biasing the transistors too hard.  For example, on the headlight input, a 1K load is a 5mA full-scale swing.  You don't need such a high operating point to drive a PIC input.  Also, why do you need an opto (vs. just a transistor) as the grounds appear in common. On the output side, 470 base resistors mean full-scale swings of 9 mA.  I realize you want to drive the transistors into saturation but I don't think all the loads need so much base current.  If so, perhaps for a dime or so more per affected switch you could use an N-Fet which would require zero current and use the same SOT-23 (or that's what it looks like) footprint.  Also, the 330 ohm into the RS Chuff opto seems a bit low unless that opto output really has to drive a lot of current.

Bottom line. I'd think you could reduce regulator current by, say, 25+ mA which might bring the heat issue under control.