Power, Comms and Control

How do you power it? (battery, track power pickups, etc)

How do you communicate with it?  (DCC chip, radio, ultrasonic, infrared, etc).

How do you control it?  (How do you identify which end of which car?)

Questions

All good questions Dave. Lets see what the collective can come up with.

For the communication method I had thought a dedicated decoder strictly for uncoupling and perhaps braking in the caboose.

Any ideas to your own questions?

Bernd

Don't care

I don't have any answers because I wouldn't ever buy it and since the era and type of models I buy aren't "commercially" offered, I would suppose that I would never have to worry about it.

But how do you control it?

Even if you get all the mechanical and electronic stuff figured out, it will still be a PITA to use remote uncouplers, because any method of determining which decoder and which end of the car to activate will be complex and awkward.  If it makes your operations more difficult, it won't be worth the trouble.

Actually, Controlling it

Actually, Controlling it might be the easy part. You could always just include an IR Receiver in the activator circuit and then make the IR emitter (controller) built into a hand-held wand which you simply point at the coupler you want to uncouple and push the button. The activator senses the IR signal and activates the coupler. I doubt anything fancier (i.e. DCC Enabled) would really be all that necessary since its not often you want to uncouple something while the train is in motion. And this would greatly simplify the electronics and reduce the size requirements for the electronics package. Another plus is that many types of plastic are practically invisible to Infrared light which would make it a lot easier to disguise the receiver.

John

...

Bernd, I think you're on the right track...

The Kadee coupler is almost already as it is for an operating coupler system.  if you were to take that "brake hose," straighten it, and then bend it up 90 degrees, and then rotate it 90 degrees [so that it's normal to the centerline of the car] you have an effective lever arm attached to the knuckle of the coupler.  Now, attach a solenoid or other motion device to this arm with a wire - power on, solenoid energized, coupler open; power off, solenoid deactivates, coupler automatically closes - all thanks to the "evil' kadee coupler spring!  This answers the mechanical question and it also answers the "what happens when you lose power" question.  It's also backwards compatible to all pre-exisiting kadee couplered equipment.  Feasible?  I don't know much about solenoids - only that this setup would be able to use a number of control arms.

It would be slick if we could do the communication wireless, while feeding the power to the car through the trucks.  I think this would be best only because information in the track is very sensitive to dirt.

I think it would be easy to design a 4 function decoder that fits in a sill or under an end - a postage stamp or a stick sort of decoder.  It would be surface mount circuit board, a design file that has 100 or 1000 on a card, which is then cut up by a lasercutter at the end of the assembly line. Hence, a DCC decoder is only as expensive as the run.  I think decoder design will be best handled by the big players, and they will, once the need becomes widely apparent.

I do believe DCC is only cumbersome and awkward if one insists on continuing to use an operating system that isn't even on the same level as DOS.  Yes, we have to move to the GUI.

Once you're in a GUI interface, couplers and cars and car orientation is all a manner of software engineering.  the one thing I foresee the user having to do is walk the train once and correct the car ends, so that the system knows which coupler you are activating when you say "THAT one."  You'd do this by walking the train once, "connecting brake hoses."  As you come to each car, you tell the system which end has the brakewheel, and the system correctly correlates the icons with each coupler.  Your manifest appears as a "box" full of car icons.  You scroll down to the car you want to disconnect [the car icons scroll up and down], activate the car [highlights the icon], then activate the proper end, and the system sends out the right information to open that coupler.  If the system is really smart, it activates both that coupler and the one attached to it.

Once we have common use of the GUI, I can easily foresee a time where we operate using tablets.  This would allow us to open our switch list on the same pane where our throttle interface is, our consist information is, and our manifest information is.  This does away with all that operations paperwork, even the time table and the fast clock, because these things would all appear within the tablet pane.  But again, this is all software engineering...

The only thing I don't like about the infrared wand is the fact that I'd then have to carry two devices to interact with the trains, or carry along a second operator.  This is like carrying a phone, a camera, a calculator, and a clock, when I could carry a single device with all devices converged together.  If I'm carrying a GUI tablet with me to operate my consist [you'll see more and more of this as smart-device technology becomes cheaper] then it'd only make sense that I'd operate the couplers through that device as well.

Anyhow...we'll see this unfold in my lifetime, of this I have no doubt.  DCC came around about 1993?  So I figure Remote Couplers [DCC+ => DCCNEXT => DCCAFTER, perhaps] will be standard operating items by around 2027 - you won't be able to buy a car without it at that point. provided things stay constant.  You might say that's eons away, but I still remember where I was in 1993...do you? Blink of an eye...

Just to be different...

Let us assume for a moment that the Kadee coupler is not the only readily available coupler on the market.  Let us further assume that the Sergent Engineering coupler is a viable option.  I have given some thought on how to use them with passenger cars with full diaphragms, and my solution would be to install a small electromagnet over the coupler.  The magnet could be energized when the uncoupling wand is waved over a sensor at the end of the car, or remotely via an output (with or without a relay, depending in the current needed) from a DCC decoder.

In other words, if the electromagnet can be made unobtrusively small and still work, the Sergent coupler lends itself very neatly to remote uncoupling.

Another idea, along the lines of what Benny was suggesting with modifying the trip pin, would be to bend the pin so that it is transverse to the track, and mount an electromagnet on the end of the car.  When the magnet is energized, the pin is pulled toward the center of the car, and the knuckle opens.  You could also extend the pin toward the center of the car (reversed from the normal orientation) and possibly extend it to better hide the car-mounted magnet.

A third option would be to utilize the split body design of Accu-Mate and Atlas couplers, where the top and bottom half (one of which holds the "thumb," the other the "knuckle") rotate relative to each other to uncouple.  By extending the bodies toward the rear, possibly with steel wires, it would be possible to make convenient lever arms on which some force could be applied to open the knuckles.

As long as I am being different, I would not personally care for a GUI and the need for sequencing the cars in my train in real time with every single move.  To me, it seems overly complicated, unrealistic, and too much like a video game.  I would prefer the ability to select a car by reporting mark and A or B end and activate the "cut lever."  That way, there would be no need for software to track my train, and no associated confusion when things get out of sequence.  If I have to update the software with every move, it's too much like work, and not enough like fun.  Even the prototype does not require that I report to the clerk every single time I couple or uncouple or move a car.

..

I agree on this about Sergeants; in a lot of ways, the MTH coupler does precisely as you suggest; a magnet hovers over the coupler and lifts the ball.  It's mounted on a piece that extends out of the pilot when activated.  The main issue is, it's MTH and it's covered by a patent.  Heck, most of these uncoupling systems you see emerging right now are indeed all covered by patents!!!  I foresee another lawsuit in the future when people try to reconcile who owns what on these couplers...

The whole idea behind software is to simplify operations to the point where all your concerns and focus can be applied to the trains and to running the trains, not to the Game of operating the railroad [the paperwork, timetables, fast clocks, etc].  It provides the Operations Gurus with all the tools they need to make the Operations game as complex and as intricate as they wish, but from your vantage point as the operator or even dispatcher, it's as easy as it is if it were just moving cars around...

I want you to imagine using DOS to dial up a specific printer on a network with 6 printers where you have to remember the complete address of each printer to use it when you want to print, in addition to the code to command the printer to print what you are working on...  Now imagine doing it with a GUI [i.e. Printing a word document from Windows or MAC].  That's how easy I foresee this Car Operating GUI versus crunching numbers on the button boondoggles.  You type in 6 numbers, I scroll down to the car in my consist tap the car icon once to highlight it.  You select coupler A and then press F1; I simply tap the coupler icon on the A end of the car and the coupler comes undone.  You deselect the car and put the car card in the siding box, I drag and drop the car icon into the Industry Siding box for that industry, thereby removing it from my train and putting it in the folder for that industry siding.  When I open that folder, I can see the other cars in that siding, and drag and drop the cars I'm picking up into my consist box.  We get the same effort done, I simply have more time to watch the trains and need less time to play with the throttle.

It has to be so easy and quick to set up that it's just the first thing you do when you pick up a train, just as the first thing I do when I get a train on the current system is I go through my car cards and determine A) has my train been blocked andb) What goes where and then C) I look at my train Order to determine my service towns before D) looking at my clock to determine my departure time.  From there on out, no more pain.  And it has to be simple, or people won't use it!  It can't be too simple, though, or the users will quickly outgrow its utility, and the people who like to go further than bare basics won't use it either!

My last words before I'm out...

I am not going to itemize each car in my train to some electronic box before I leave the yard.  "Smart phone" is an oxymoron.  If I wanted to update electronic lists with every move, I would simulate my layout on a spreadsheet.  And I don't think designing something for couplers other than Kadee or similar is wise, since there is simply not enough people interested in Sergants (though I do admire the concept) and plastic couplers are frought with pain and suffering, from my experience.

As soon as you have to change everything you use or complicate how you have to do things, it becomes a hard sell.  The problem is how to make the transition easy to do and to use (and inexpensive??  good luck there...)

Another long reply

Benny, I think we're thinking of different situations.  For a through freight it's reasonable to maintain a standing order and direction library for the entire train.  The trouble from my point of view comes when one settles into some heavy switching.  I don't care for car cards or waybills, preferring a simple switch list which indicates what cars go where, and what cars are scheduled to be pulled.  Even with car cards, I do not re-order them with every single move that I make.  I set them down somewhere, do my work, and then sort the cards to indicate the new standing order.

While that would be practical from the point of view of operations, if you want to use the system to remotely uncouple a specific car, you would either have to know what car it is (presumably using the reporting marks and number, just as I suggested earlier, and completely eliminating the need to know the standing order in the first place), or keep the list up to date with every move.  Unless it is automated, possibly by RFID, transponding, or some other means of either tracking every car or allowing the cars to report their relative locations, the user would be left with the task of making a move, updating the computer, making a move, updating the computer, making a move, updating the computer, ad nauseam.

As for the argument about DOS vs. Windows, you are talking to a guy who cut his teeth on DOS, is familiar with (and used for many years) OS/2, uses Linux, and misses the command line as a useful interface with Windows.  There are some things which are simply much more efficient with a command line interface, if the user knows what they are doing.  Obviously that efficiency comes at a price - training.  Added layers of complexity are frequently (though not always, by any means) simply crutches to allow the users to be less competent, and the result is that things only work when they work perfectly.  Uninformed users make stupid mistakes, and the software has to become ever more complex in order to be ever more idiot-proof.  there has to be a balance somewhere, and there has to be an opening for competent users who want streamlined efficiency without all the bells, whistles, and protections.  Similarly, I prefer a manual transmission in a car, know what the sounds my cars make are and maintain or repair the vehicles as needed, and I keep my fancy all-auto digital camera on full manual mode.  I dislike giving up control to a computer that may or may not have the same ideas about what needs to be done, and I want to be able to understand whatever is going on.

In the case of a model railroad, if I want to refer to a specific piece of rolling stock, I would do so most effectively by the reporting mark and number.  Just yesterday, we had to set two bad orders out of our train.  While we were conscious of the fact that they were lines 27 and 68, our instructions were to set out the cars with the specified numbers.  Of interest to this discussion, the reason we do not use just the line number is that it is notoriously difficult to maintain an accurate list, and simply counting cars would result in the wrong car being pulled, cut, or otherwise handled.  If all of the prototype railroads in the history of railroading have not managed to make it happen, it seems rather bold to think that it can be done on a model, with a much more limited budget and market.

But, to get back to the original point, let's talk about remote uncoupling, not GUI interfaces and layout databases. The point is to discuss what can be done, how, and whether it would be practical and/or desirable.  Unless you feel that it's an all or nothing deal, let's keep this thread about couplers.

Good plan...

Not hard to do!!

I do think the Kadee augmentation would be the most effective uncoupler - cross compatiblity, closed natural state, "operateable" without the active mechanism - and I dare believe Kadee is working on it as we speak.  I know they got the G scale version working, so the smaller versions, I be they're in the pipe line...we'll see, aye?

Oh come on, it is not that

Oh come on, it is not that big a pain in the butt. Right now with existing tech I could do this in a couple seconds.

You simply use the last 4 digits of the car number to id the car that give you something like 9999 cars and engines if you have more cars then that you most likely cant afford remote uncoupling anyway.

As for the end ALL cars have a A and a B end the are identifiable on the prototype so you can identify them on the models.

So assuming that you need one button to uncouple one coupler the operation with DCC would go like this.

Select the car number (last 4 digits) once the car is selected you press either 1 or 2 (depending on what car you want uncoupled. On some systems you can do this with aux functions or you can do this as you would an engine.  

It is not that complicated. If it is complicated then you need a new DCC system

-Doug M

Mechanical first

Benny your running off in to the ding weeds again trying to put the cart before the horse. Your input is a appreciated, but lets keep it on track (pun intended). First we need to come up with a reliable coupling-uncoupling device or activator.

I like Ken's idea of using a passenger car. Lots of room and they won't be interchanged as much or at all in a club or even home situation. I also like the idea of using the Sergeant coupler. Now I'm not partial, I use mainly Kadee now. I have purchased the Sergeant couple and find it looks great and works great.

So lets assume modifying a set of passenger cars with Sergeant couplers. I also like the idea of using a laser pointer. I've played with those and understand how they function. I had built a infrared laser and was going to apply it for walk around speed control. Shortly after that idea Lorel Joiner came out with his article in MR about building one. The beam can be modulated so you can send coded signals to a receiver. You would need a tone encoder and a tone decoder. Some are probably asking what a tone decoder is. Any of you that have had experience with a land line phone know that one. It's the sound the different buttons make when pressing them. They are generated by a tone encoder. Now before you DCC guys get upset I'm just suggesting one method of control at the moment. Not all have DCC on their layout yet. Now, if you don't have DCC then you don't have constant power on the track and this proposed device would not work, thus the need to carry a battery on board. Could it be a rechargeable, yes. or just an ordinary battery. Voltage to be determined.

The above musings would be the first stage of development. That's about as far as I have gotten on mulling over this idea. This same concept could be applied to the braking idea of at least the caboose.

Further thoughts from the forum?

Bernd

A possible problem

I think I may have come up with a possible problem with remote uncoupling using Sergeant couplers. When uncoupling only one coupler will be opened. Which will it be, the car that's getting uncoupled or the car being uncoupled from? If the car being uncoupled from is left with a closed coupler and the car that will get coupled to it also has a closed coupler how will they then couple? This would not be the case with a KaDee coupler.

Any agreements on this statement or other thoughts?

Bernd

...

Bernd, I'm standing in a patch of weeds today where the mainline will be tomorrow!!

Here's my solution: utilize the Kadee coupler - there's my cart...and then I moved to the most logical means of controlling that coupler: DCC - there's my horse.  To tie the two together, I'd use any device that pulls back the coupler bar on the Kadee coupler, activated by an electrical impulse - there's my tongue and yoke.  The problem now is identifying the components and then getting the components small enough, but this is where industry steps in and brings their trained mechanical engineers, software engineers and advanced manufacturing techniques to bear.

This coupler system would then be backwards compatible with the most popular coupler on the market.

I have become completely unapologetic towards those who are still un-DCC. they don't have to use DCC, but it's in their best interests to become educated on it because it had might as well be an industry standard now.

How we move that digital signal is still a question of debate; I believe wireless is the route to take. 

I like the laser pointer

The laser pointer has a nice advantage, in that you don't need to know the identity of the car you are aiming at.  The disadvantage is that it will need to be visible to work, so no uncoupling inside a building.  Two things:

Why do you need to modulate the laser or send any signal at all?  If each car has some sort of target or receiver on it, just aim at that to open the knuckle.  That way the only thing you need to have on-board is the uncoupling mechanism and some sort of sensor which is preferentially sensitive to the light from the laser pointer.  Making it work with any commercial laser pointer would be a plus, since they're a dime a dozen these days (we even have some shaped like mice for the cats to play with).

As for which knuckle on a Sergent, follow the prototype practice.  It is generally a good idea to open the knuckle on the car you're leaving, although I was taught early on to open every knuckle I see, since sooner or later somebody is going to couple to it.  I see the Sergent as being prototypical in function and use as well as in look - the prototype knuckles don't open themselves, or center automatically either.  With all that said, I recognize that the Kadee and Kadee clones are the current standard, so any solution will probably have to work with them if possible.  On the other hand, it may turn out that remote uncoupling is a highly desirable feature which can best be implemented with Sergents, and the hobby will move in that direction, much like it did with Kadees decades ago.

Also, it's funny that you picked up on the passenger cars in my idea as part of the solution.  To me, they were the problem - you cannot get the uncoupling wand in there on a car with diaphragms.  I didn't even think about the inherent advantages.  Funny how things change as soon as you shift your point of view.

See

See http://www.unionvilledepot.com/

(apologies if this has already been pointed out)

(apologies for the bad pun in the previous sentence)

I'm a little shocked by the prices tossed around.

For example, the "nano" servo mentioned earlier. Looks like $32.00  for that alone. Laser pointer coupler receiver above asking $32.00 also. 

Looks like many of the "you could do this" ideas on the thread would come out at around $100.00 per car, by the time you added the mechanical interfacing parts, although I suspect although a "Bachmann" type of Chinese all built-in to the car at time of manufacture solution would only add about $20.00 extra in today's pricing models..

I'm old enough to be still living the '70's (in my mind ), so my personal interest "and perceived value"  target for a "gizmo" would be about $5.00 per car. But then my interest in the hobby has always been on the "how can I make it in quantity, rather than only afford one" side of the equation.

An Old Phrase...

There's Nothing new Under the Sun here...

I dare say the Main Manufacturers all have something under their hats right now, they're just wrinkling out the last details.

Andy, your thoughts are right in line here.  I am remiss to recall the early days of DCC, when a simple decoder war $50 or $100, and all it did was move the motor and turn on the lights.  And when those threads rolled [DCC v DC] my response then was "It's too expensive now, but when the price comes down, it will be a no-brainer."  Which made the DCC people unhappy, because they wanted people getting on NOW, and it made the DC people unhappy because they wanted to hear how DCC would go AWAY forever, not grow in strength.

I will say I am alarmed at the rate new freight cars and passenger cars have accelerated in price, mainly because I don't see the addition of metal wheels, Kadee clones, and cleaning up the paint to modern standards to be worth a 200% or 500% increase in the sticker price.  But if you added an operating coupler and a DCC decoder, ok, now I understand that increase - there's something in there of new value.

Ideally, if Kadee were to team up with Kader to bring operating couplers to the market just as Soundtraxx teamed up with Kader to bring economy DCC sound to the market, I think we'll see the most effective system and at the best price point.  Kadee would develop the mechanical circuit and then relay the decoder needs to Soundtraxx through Kader.  Soundtraxx designs the new minimally sized decoder, Kader then makes them using their economy of scale production arm and puts the whole thing together in their factories.  Kadee collects royalties, Soundtraxx collects royalties, Kader builds it and collects profits and dishes out royalties; all three go up together.  Heck, if that were to be done, at that point you may even see a Kadee Coupler on every Bachmann product - now how about a win there?!

This is how the system is Supposed to work!  Otherwise, you end up with things like the MTH/LIONEL/QSI Lawsuits, which really sucked the wind out of those sails and soured more then a few to at least one of those names.

The problem with making it happen

I think the main impediment to remote uncoupling is going to be design inertia.  If you think about it, most of us build our layouts so that we can reach any conceivable point at which a car might need to be coupled, uncoupled, or re-railed.  With that in mind, remote uncoupling is mainly a neat gizmo, but not really a practical necessity.  Only when remote uncoupling becomes common, or at least common enough to be viable for an entire layout, will layouts be designed to utilize and even require the feature.

It's much like the early days of DCC, where layouts were already built for DC operation.  Only when DCC reached the point that it was easy to get and reasonably inexpensive did people start building layouts that fully integrated the additional capability, and thus required it from the start.  I'm thinking about helper districts, engine terminal switching, and other things that were challenging if not impractical or impossible prior to DCC.

I am currently in the "It's neat, but I don't need it" camp.  In a few years, who knows?  I can see how the ability to open a knuckle without the need for a pick or magnet would be pretty handy.  If it were practical for $5/car, I might be interested.  Much more than that, and the cost outweighs the benefits.

Thanks Scarpia

for re-posting that link. I remember it now from the previous un-coupler thread. Looks like the electronics are already done. The only thing to do is come up with a mechanical design that's reliable. And I'm still thinking brakes in the caboose along with this.

@Ken,

As far as the modulated laser, I was thinking coupling = one frequency, uncoupling = second frequency. But your right it seems over kill here.

I was wondering if you'd speak up and tell us how the prototype does it. I figured it would be the car from which the other car gets uncoupled from. I think a mechanical linkage can be developed.

The reason for picking the passenger car wasn't so much of getting the wand between the cars as it was that you can use the windows as a place for a laser light receiver.

@Andy,

Remember when 512 bits of memory was more than what a 500 gigabit hard disk costs. My Atari 800 cost more than the computer I sit at now and I can do lots more with this new gadget. So once modelers get interested in this idea and buy the product, if it gets on the market, I'm sure the price will come down just like they did with computers.

@Benny,

Benny I think you missed your calling. You should have been an organizer. Why don't you get all the different manufactures together and present this project and see if they'll go for it. Then you can collect the royalties. What da' think?

Bernd

Modified coupler

While outside doing some odds and ends, like cleaning the lawn mower and covering the boat because it's supposed to rain I was thinking of couplers and how to automate them.

Well here's an attempt at a prototype for the remote coupling/uncoupling project.

A modified KaDee #5 and the wire that will one link of the activation system. Mind you it's just a start.

I then stuck it on a piece of clay for an end view.

And the wire slipped over the coupler pin.

Next, how to mount the wire under the passenger car. I've got some ideas, but it'll be a while for that.

Bernd

Why not try to improve on this

Why not try to improve on this since its already a working design. You guys seem to be working too hard. I have seen the laser design and thought it was neat. However, not very practical. The sensor placement would be a huge problem. Shining a laser at a string of cars what would stop accidental uncoupling? Muscle wire is expensive, maybe $15.00 for 3ft piece but your only going to use about 2" for each coupler. The rest of the components wont cost much. As far as decoders go all that is needed is a "Function" decoder. A bulk order could get the price down quite low. 

Trying to figure out which end of the car has the "A" or "B" uncoupler is easy. DCCUncoupler uses the end with a brake wheel as an easy site point to determine which end is which. I have used them and all I can say is "Awesome".

Rob Teed

New Design

Rob,

First thanks for posting your video again. You show another method than mine that can be used. The more different designs the better. This will let people try them out and then implement the one that works best for them.

The laser light would only be on when you press the bottom. I really don't see how you would uncople accidentaly. Also you could aim it at the center of the car. Right now I'm working on experimenting with passenger cars so choosing the right car shouldn't be hard, plus they have clear windows that a laser will shine through.

I played with muscle wire when it first came out, IIRC Mendotronics was the first that offered a kit. For those that are interested in muscle wire here's were you can get it.

http://www.ebay.com/itm/MONDOTRONICS-MUSCLE-WIRE-PROJECT-BOOK-SAMPLE-MUSCLE-WIRE-KIT-/110909657307

Another source for this wire is here: http://www.jameco.com/webapp/wcs/stores/servlet/Product_10001_10001_357472_-1

A little over 5 feet costs $29.95. A lot of wire. Also it can be bought in several thickness'.

The one thing I don't like about the muscle wire is that the only way to attach the wire is crimping. Can't solder it. Also the length needs to pretty close so you can get the same travel. But as I said the more iterations the better. Let the modeler choose which system will work for him, plus somebody may build on one of our systems.

Right now I'm developing a system that will work with truck mounted couplers as in the picture below. These are from a set of Athearn aluminum fluted passenger cars. I think it would be a little hard to mount your design on these truck mounted couplers.

I think that one design will not work for all cars. I'm looking at using a micro servo for activating the coupler. Right now I'm working on the mechanical end of the design.

Also any body with passenger cars that could post a picture in this thread of how the couplers are mounted on your car would be a great help.

Bernd