Unusual DCC circuit request...

[Phoebe Vet]

Interesting OT observation:

Not one person asked "who is Rube Goldberg"?

We must all be very old in here.

[blownout cylinder]

Phoebe Vet

Perhaps Rube Goldberg would have been a better example.

Somebody probably would have complained about that too.

[Allegheny2-6-6-6]

 Simple solution: Do what the UP did in Cheyenne park em all head in with nice easy to read road numbers on the back of the tenders. But don't use a Russian pencil and if that doesn't work smash it with one of those $750 US government hammers.

[Phoebe Vet]

I didn't claim it was true.  I said the thread reminded me of the story.

Looking for an incredibly complicated way to perform a simple task.

Perhaps Rube Goldberg would have been a better example.

[CSX Robert]

Phoebe Vet

Early in the space program NASA discovered that ball point pens need gravity to work and so could not be used in orbit.  They spent more than a million dollars developing a pen that would write in zero G.

The Russians used a pencil.

Urban myth. NASA did not develop the pen, it was developed by a private company with no funding from NASA. Also, both Russian and US astronauts used pencils initially, but there are good reasons for not using a pencil in a spacecraft.

[Phoebe Vet]

CSX Robert

Phoebe Vet

This entire thread reminds me of the NASA zero G ball point pen story.

How?

Early in the space program NASA discovered that ball point pens need gravity to work and so could not be used in orbit.  They spent more than a million dollars developing a pen that would write in zero G.

The Russians used a pencil.

I would hang a small whiteboard on the wall and just write down the address of the engines that are parked in the roundhouse.

[CSX Robert]

Phoebe Vet

This entire thread reminds me of the NASA zero G ball point pen story.

How?

[Phoebe Vet]

This entire thread reminds me of the NASA zero G ball point pen story.

[AlienKing]

Here's a crazy idea:

Turn your roundhouse into a giant multiplexed programming track.

You'd have a dpqt switch that toggles between track power and programming track for all the tracks around the turn table. The track power on the table itself will shut off when in programming mode to prevent "accidents".

The tricky part is creating a computer controlled set of relays (the mux) that would supply programming track power one track, read back the CV(s) for the loco number, and then show that on the board above the corresponding stall. You could create a script that would cycle through each of the stalls on startup and each time the switched is flipped from run to programming.

[tricky_trev]

as an indication, the components for the current detector = US$3. A PIC costs US$1.

i used the current detector as the basis of my signalling system (i have about 100 circuits around the layout - cheaper than an expensive loco)

time in a hobby costs nothing.

[Allegheny2-6-6-6]

 . We looked in to RFID tags on the bottom of each piece of rolling stock in the club so the dispatcher could tell the exact location of every piece of rolling stock and locomotive at any given time. Yup just like the prototypes do except instead of being located on the sides of each car we were going to place a tag underneath.and have readers placed in strategic places around the layout. That idea went south when we calculated the cost involved.plus there did seem to be other issues such as others have mentioned but it's a growing technology so who knows whats on the market now.

So we use plan A: magnets on a board. The kind you can slip a little piece of paper into with the road number or car number. There are two per piece or rolling stock and locomotive. When a locomotive is in a particular yard or siding or the Round house the corresponding magnet is placed on the track diagram. The dispatcher does the same thing on is board with the aid of an assistant dispatcher. So at any given time during operations when there can be up to 40 trains running if you want to know where car ABC123 is you check with the dispatcher and if it's not where he says it is then someone dropped the ball. It usually isn't too hard to find out who the guilty party is. Even during a work session or any other time if there is a trouble ticket as well call it for a car or locomotive that may have ad some issues during the last op session the repair crew can go to the dispatchers board and find exactly where that piece is. With several thousand pieces of rolling stock and I think we are some where around 140 operational locomotives this system is an absolute must.

[CSX Robert]

jwhitten

No way-- you can go an easy 8 to 15 inches with el-cheapo off-the-shelf stuff. I've done it myself with RFID stuff I've gotten off ebay for less than $50 bucks...

John, I think you misunderstood Randy's point. The problem isn't not enough range for the reader, but too much range, leading to the reader reading tags on nearby tracks as well as the one it is intending to read.

[jwhitten]

rrinker

There are some people working on using RFID in model railroad applications. The biggest hurdle so far is the range - 2-3" of seperation in HO scale it still close enough that multiple tags respond to the sensor.

 

 

No way-- you can go an easy 8 to 15 inches with el-cheapo off-the-shelf stuff. I've done it myself with RFID stuff I've gotten off ebay for less than $50 bucks. You can do it with the 125KHz stuff and the 13MHz stuff. The range varies a little but well within usable limits. It is true that a lot of it hinges on good antenna design which can make it difficult to hide the thing.... well, not really. Anybody who can hide a Kadee uncoupling magnet can hide an RFID reader antenna... IMO :-)

 

John

[jwhitten]

RF&Prr

  I am interested to know if anyone has done this.....

Is it possible to build a circuit that will read a DCC signal and extract just the engine address only? This circuit would need the ability to read all engine addresses being sent to a roundhouse stall track. 

  My application would be for a 12 stall roundhouse, so one knows which engine is currently in which stall.  On my control panel would be twelve - 4 digit numeric led displays, numbered according to the stalls.  These 4 digit displays would then let me know which stall engine number xxxx is currently in.

 John

 

 

If I were going to try something like this, I think I'd see about doing it with RFID. Reader/Writers are cheap and so are the RFID tags-- you probably would want the "grain of rice" type tags which could be located nearly anywhere on the loco and painted any color you like. Then you would have a unique ID associated with every loco which you could then read with one or more readers wherever you want them, and manage the data read however you like.

[CSX Robert]

RF&Prr

Suppose you use an IR emitter detector scheme.  IR emitter (facing downward) in each engine set to a specific freq. and have the detectors mounted in the stall tracks.  You could build the emitter circuit small enough to reasonably fit into a steam tender, but a diesel would be harder to do.  If each engine had it's own specific freq. then the detector would know the engine #.   The detector circuit would have to be able to detect all frequencies used and then transmit the proper decimal code to the LED display drivers based on which freq it has detected.   Seems possible and not to costly.  Any thoughts?

There already is an infrared emitter/detector scheme available, it is the Uhlenbrock LISSY system that I mentioned in my first post in this thread. Here is some more information about it.

[rrinker]

 There are some people working on using RFID in model railroad applications. The biggest hurdle so far is the range - 2-3" of seperation in HO scale it still close enough that multiple tags respond to the sensor.

 Another option is Uhlenbrock's LISSY. This is an IR transponder system a small chip with a unique ID, like an RFID tag, is installed on each loco and an IR LED tramsmitter/receiver pair goes between the rails. As the transponder passed over the track sensor it broadcasts its ID. The transponder cna be programmed like a mobile decoder to whatever ID you want to use, the fixed side of the system can each monitor two tracks (2 sets of sensors connect to each box). It's a Loconet device (yet another bit of wondeful gadgetry that works with Loconet but is not a Digitrax product) and support for it is in JMRI. WHat ther isn't is much information in English, which is rather wierd. ESU offers their site in multiple languages, Uhlenbrock is German-only. The one site I foudn with any information in English is 6 years out of date, but they do mention at the bottom 4 dealers in the US. The Lissy manual on the site is the German one, but the pictures give a pretty good idea how it works and how to install it. http://www.rjftrains.com/intellibox/uhlenbrock.htm I wish there was more information available - this would seem a great way to go, a transponding system that is independent of the DCC track signal. Short range, and localized to a section of track (point detection vs "anywhere in the block" which is exactly what you'd want on roundhouse stall tracks.

                            --Randy 

[jrbernier]

John,

  I agree, getting away from using the DCC as the detection mediem is a good idea.  What about using small RFID tags(like those in pets)  They are small enough to put in an engine and can be read by a small proximity scanner - there is lots of HW/SW for this stuff.

Jim

[RF&Prr]

 Well, I am mostly steam, but do have several diesels too.  I have a good understanding of electronics and am quite computer literate (MCSA).  The cost of this roundhouse engine detector is going to be a factor for me, so I am thinking that trying to work out a circuit to cypher the DCC loco address is going to be difficult, tricky and costly to do.  

Now , for my next idea that is totally independent of any DCC system....

 Suppose you use an IR emitter detector scheme.  IR emitter (facing downward) in each engine set to a specific freq. and have the detectors mounted in the stall tracks.  You could build the emitter circuit small enough to reasonably fit into a steam tender, but a diesel would be harder to do.  If each engine had it's own specific freq. then the detector would know the engine #.   The detector circuit would have to be able to detect all frequencies used and then transmit the proper decimal code to the LED display drivers based on which freq it has detected.   Seems possible and not to costly.  Any thoughts? 

 John

[locoworks]

idea 'B' ?   assuming you are all diesel stock,  just select likely numbers and switch the lights on and off or off and on to confirm the correct loco is selected. i'm sure you will be able to tell what numbers go with which loco types?? so it may take a few trys to get the right number??

[CSX Robert]

tricky_trev

engine #1 would need to do do the same acceleration as engine #2 to confuse the system, which is almost impossible.

It depends on how busy the layout is and what the other engines are doing. With several operators working simultaneously, some of them swicthing out industries, it may not be that hard to confuse the system.

[tricky_trev]

if we put a current detector & PIC (basically a modified stationary decoder) under the turntable and one of the same in each of the stalls, then we can "see" engine #2 start moving since we can read the throttle command, loco number, and monitor the current draw rise. linked with the adjacent PIC, we have enough inteligence. we probably have to stop the loco on the turntable.

engine #1 would need to do do the same acceleration as engine #2 to confuse the system, which is almost impossible.

any refinements?

[CSX Robert]

tricky_trev

...Have you got any PIC programming know-how or a PR3? You need to monitor the DCC packets on the layout (very easy to do), and monitor the feeder droppers for the stalls with a sensitive current detector (i use a simple op-amp design).

if you monitor the packets along with the status of the current in the stall (train entering =1 and leaving =0), you can have a very reasonable and accurate guess of what loco is there. store it in a latch/memory or text file on the PC and you will "remember the status". sounds like a cool project i should add to my layout as i have the same problem...

This would only work if you do not have other engines running on the layout when you are pulling an engine into a roundhouse stall. If you have engine #1 running on the layout and engine #2 pulling into the stall, how do you know which one is pulling into the stall?

[tricky_trev]

actually, this is one of the coolest questions i've ever seen on this forum for a while....:-)

i've considered transponding before on my layout, but even as an experienced electrical engineer, the concept is still too young and complicated.

However, here is a non-transponding way i've thought of for you (the concept only really works for a roundhouse):

Have you got any PIC programming know-how or a PR3? You need to monitor the DCC packets on the layout (very easy to do), and monitor the feeder droppers for the stalls with a sensitive current detector (i use a simple op-amp design).

if you monitor the packets along with the status of the current in the stall (train entering =1 and leaving =0), you can have a very reasonable and accurate guess of what loco is there. store it in a latch/memory or text file on the PC and you will "remember the status". sounds like a cool project i should add to my layout as i have the same problem.

i hope i've inspired you now ;)

[MisterBeasley]

locoworks

  it would be easier to fit a couple of webcams and look at whats in there. ??

Miniature video cameras would work, but if you have similar engines which you can only distinguish by road number, they probably won't have the resolution you would need to read the numbers.

What would the prototype do?  My guess is, they'd get a whiteboard and write the engine numbers on it.  In the steam era, it would be blackboard and chalk.

[locoworks]

 it would be easier to fit a couple of webcams and look at whats in there. ??

[jeffrey-wimberly]

RF&Prr

Is it possible to build a circuit that will read a DCC signal and extract just the engine address only? This circuit would need the ability to read all engine addresses being sent to a roundhouse stall track. 

I find that a notepad and pencil works just fine for me. I also have a label on the bottom of each loco that tells the loco address and the type of decoder installed.

[60YOKID]

 

RF&Prr

Is it possible to build a circuit that will read a DCC signal and extract just the engine address only? 

John, One way would be to read each individual isolated track with JMRI Decoder Pro. I don't know how you would streamline the process or display for the 12 tracks though.  It would probably require a polling type of software program.  Just a thought.

[cacole]

A German company (Dietz Modellbahntechnik) has proposed the SUSI (Serial User Standard Interface) protocol for DCC decoders, and the NMRA has assigned an interim Recommended Practice to this technology which will use CVs 897 through 1024 to control additional advanced functions; however, no other DCC manufacturer has yet adopted the SUSI specification.  If it does ever come about, this may be what you need..

[CSX Robert]

RF&Prr

Again, I do not intend to buy / replace 40 plus decoders that can do transponding.   There is a way to do this using the DCC commands on the track and not having to purchase transponding decoders.

There is no way to do this just by stripping the address from the DCC packets, unless when you are running a train into an engine stall it is the ONLY engine on the layout being moved at that time. If you have another engine moving anywhere on the layout it's commands will be seen on the stall track as well as the engine you are running into that stall.

If you made sure that all other engines are stopped and released when you run an engine into a stall then this idea could be made to work; however, it would not just be a matter of building a circuit, the circuit would require a processor that is programmed to read the address.

Here are your options:

Transponding - You would have to install transponding zone detection for the roundhouse stalls. You would not necessarily have to replace any decoders, but for engines with decoders that do not have transponding you would have to add a Digitrax Transponding function decoder.

RailCom - One neat thing about RailCom is Lenz makes an LRC120 RailCom detector that will display the address of the engine in the isolated section of track. Again, every engine would have to have either a RailCom equipped decoder or an add on RailCom transmitter. Also, you have to have a RailCom cut out device on the booster feeding the roundhouse stalls.

Train tracking software(Such as RailRoad & Co.) - This would require that your whole layout be setup for block occupancy detection, but it would not require any changes to your engines.

Uhlenbrock LISSY system - This is a system of infrared transmitters and receivers for engine location detection. The transmitters are mounted on the bottom of the engine and the receivers are mounted between the rails.

RFID - There is no ready made system for model railroads, but it could be adapted to it.