Question re: Consisting and Speed Matching

[shark_jj]

I am using a Digitrax System with a DCS 100 Command Station and a DB 150 Slave.  I currently run 22 trains out of staging with between 2 to 5 locomotives on each train.  These trains all leave staging, transit the layout, and return to staging.  The locomotive consists are never broken up during operations.   As a result, rather than consisting the locomotives, I give all locomotives the address of the lead unit, thus only using one slot in the DCS100.  I have just started the process of speed matching the locomotives.  Locomotive consists may include units from Kato, Atlas, and Intermountain, with either Digitrax or TCS decoders.  I am using JMRI to speed match.  I am assuming that the speed table is written to the locomotive decoder and that no information is retained on the Command Station.  The question is:  When I call up a set of locomotives, say #1000, which has two locomotives, say a Kato and an Atlas, which have two very different speed tables, and I send out a signal via Throttle and the Command Station to go to speed step 20, will the two locomotives respond according to their separate speed tables.  Is their anything I am missing, or should I be consisting these locos so that they get independent signals for speed.  I don't have a real in depth understanding of exactly how this process works, and this is well beyond anything in the manual, so any insight that anyone can share would be appreciated.

[peteski]

You got it.  Speed tables are strictly kept inside the decoder and the command station doesn't know or care about them.  When the command station sends control packets over the rails to a specific loco address, it will send whatever speed step is dialed up on the throttles. All the decoders which are being addressed by that packet will then receive it and adjust their speed accordingly to what is set in their internal speed table for that speed step.

[shark_jj]

Excellent, thanks Peteski, its nice to get confirmation that you are on the right track (pun intended)

[bnsf971]

If you do set all the engines to the same address, don't forget to turn the bell and horn volume to 0 on the trailing units, if you are using sound. It can get noisy pretty fast when you hit the F1 or F2 button.

[Greg Elmassian]

Is there a reason no one is recommending advanced consisting, so the bell and horn work correctly automatically for the lead loco?

Greg

[jdcolombo]

Is there a reason no one is recommending advanced consisting, so the bell and horn work correctly automatically for the lead loco?

Greg

Advanced consisting is a kludge that some DCC systems have worked around, but not Digitrax.  You have to program CV19 in each engine with a value from 0-127 and then address the consist with that value, rather than the lead engine number.  NCE gets around this (I believe) by allowing you to alias the consist with the lead engine number.  Not so for Digitrax.  Plus there is no easy way to create an advanced consist with Digitrax other than actually programming CV19.  Again, NCE (and some other systems, like ESU's ECOS) get around this by allowing you to consist via a simple throttle command.

But advanced consisting itself is not a very good option.  You should be able to set up a consist in the command station and tell the command station what functions in each consisted locomotive should respond to throttle commands.  This is 2017, not 1989.  When a phone has enough processing power to emulate a 1990 Cray supercomputer, the fact that command stations can't implement this kind of "record keeping" is simply crazy, and shows how far behind the interface curve DCC really is.

John C.

[peteski]

Advanced consisting is a kludge that some DCC systems have worked around, but not Digitrax.  You have to program CV19 in each engine with a value from 0-127 and then address the consist with that value, rather than the lead engine number.  NCE gets around this (I believe) by allowing you to alias the consist with the lead engine number.  Not so for Digitrax.  Plus there is no easy way to create an advanced consist with Digitrax other than actually programming CV19.  Again, NCE (and some other systems, like ESU's ECOS) get around this by allowing you to consist via a simple throttle command.

But advanced consisting itself is not a very good option.  You should be able to set up a consist in the command station and tell the command station what functions in each consisted locomotive should respond to throttle commands.  This is 2017, not 1989.  When a phone has enough processing power to emulate a 1990 Cray supercomputer, the fact that command stations can't implement this kind of "record keeping" is simply crazy, and shows how far behind the interface curve DCC really is.

John C.

John,
I think you got your consists mixed up.  There are 3 types:

1. Basic (simply use the same address for all the locos). This consist is portable.

2. Command Station Assisted Consist - the command station keeps the consists parameters internally (each locos address, direction, etc.). This consist is not portable and not very configurable as far as functions go.

3 Advanced Consist - the newest and most flexible consist. The consist configuration is kept inside each decoder.  It is enabled by CV19 set to a value other than zero.  There are several function mapping registers on the decoders dedicated to configuring all sorts of parameters (like direction, or function mappings or which loco responds to which functions, which unit's headlights should be kept dark, etc.).  This consist is also portable.

[jdcolombo]

Hi Pete.

No, I didn't get them mixed up.  I noted that you must program CV19 with advanced consisting and use that value to address the consist.  But I probably transitioned to my take on Command Station assisted consisting without enough of a warning that I was segueing to that in the last paragraph.

The reason I think advanced consisting is a kludge is because you can't use 4-digit addressing with it, unless some sort of aliasing is provided in the command station.  That's bad engineering from my viewpoint.  Anyone that knows anything about railroads knows that the numbers of engines go way beyond 127.  The way advanced consisting SHOULD work is to allow the lead engine number to be the consist number.  That would require 4-digit addressing, which would require more than CV19.  Whoever thought up the advanced consisting system within the NMRA specification should be tarred and feathered for that omission.  It's just stupid, plain and simple, like the whole notion of two-digit addressing was stupid (and which no one now uses).  Yes, you can overcome it with aliasing, which NCE does, but then you have a hybrid of advanced consisting and command-station-assisted consisting.  And it means that if you take that consist to a Digitrax system (or any other system that does not implement NCE's version of aliasing), you can't use the lead engine number as the consist address - you have to use the CV19 address.  Who can remember that?  I've got 20 consists on my layout, each made up of three engines.  I can easily remember to key in the lead engine number to operate the consist.  But I can't remember that the consist with engines 498, 522, and 426 has to operate on address 06.  I'd have to print out a "cheat sheet" with the consist numbers on it, and then my operators would go crazy.

This is foolish.  Moreover, I think the "portability" of advanced consisting is highly overrated.  Who cares about it?  I take my 3-engine consists over to a friends' house who has another Digitrax system, key in 498 to select the lead engine, then add 522 and 426 to the consist.  Takes 10 seconds flat, and I'm ready to roll.  Portability schmortability.  It takes longer to think about which engines were in which directions when setting up an advanced consist on another layout than it does to plop three engines down, key in the numbers, and set up a new consist using command-station-assisted consisting.

Finally, while the idea of having hard-coded decoder-level information on how engines should respond to function buttons sounds great in theory, it's not in practice unless you decide that you want every single engine to respond exactly the same way to the F keys, in which case you can set them all at the initial programming stage and leave them.  But that's rarely the case operationally.  Suppose, for example, that I have a 3-unit consist.  The way I want the headlights to work is for the middle unit's lights to be off, and the lead and rear units to operate only their headlights directionally.  OK, I can program that behavior into CV's 20-21.  But now let's suppose that I want to cut off the rear engine on that consist and have only the two remaining engines be the consist.  Now the middle unit, which had lights off in my 3-unit consist, is the rear unit, and now I want it to respond to lighting commands the same way as the lead unit.  But its not programmed to do that.  If I want that to happen, I have to reprogram CVs 20-21 to get that behavior correct.  This is easy?  This is flexible?  You're kidding, right?  OK, so I suppose I could get around all this by simply remapping the headlight and rear light on each unit to different function buttons so that I can control them separately for each unit, instead of having them respond to F0.  Fine, except everyone else in the world expects to press F0 on the throttle and have the lights come on.  So now I have to educate a group of operators about how to operate headlights and rear lights on my special layout with my special Advanced Consists.  No thanks.

So I just don't find Advanced Consisting inherently better than command-station-assisted consisting.  However, the big problem with command-station consisting today is that the command station can't keep track of how the engines in a consist are supposed to respond to function buttons.  The only way to do that today is to use Advanced Consisting.  But you could also keep that information at the command-station level; it's just memory and programming.  There should be a way to tell the command station "For the consist consisting of engines 498, 522, and 426, the lead engine and the rear engine should respond to F0 by using a directional headlight, with rear lights off.  Only the lead engine should respond to F1 (bell) and F2 (horn)."  Etc.  If I have three engines with ESU LokSound decoders, I should be able to tell the command station that only the lead engine should respond to F1 and F2, but all three engines should respond to Drive Hold on F9 and the brake on F10.  And I should be able to set this up once, and have the command station remember it.  And it should be done in some kind of logic fashion so that if I cut an engine off a 3-unit consist to make it a 2-unit consist, the command station recognizes that I've done this, and automatically applies my consist Function rules to the "new" rear engine.  This can be done - it's just memory and programming.

Now I'll admit this does present some issues for "portability" - I'd have to have some way to import that function-response information to a new command station if I take my units and run them on the club layout with a different command station.  But this is the era of cell phones with more processing power than we used to send men to the moon and ubiquitous wi-fi.  Command stations should all be equipped with wi-fi capability, and there should be an app that can download your consist information to your phone, then upload it to a new command station when you go run at a friend's house or on the club layout.  Problem solved.  And such a system would be far more flexible, portable, and usable than the Advanced Consisting method.

Admittedly, the big problem of addressing Advanced Consists with the lead engine number could also could be solved by changing the NMRA standard to provide 4-digit addressing for advanced consisting.  But the liklihood of that happening is about zero.  It could also be solved, I suppose, by each manufacturer introducing an aliasing system.  But to implement that, you'd still presumably have to remember the CV19 address so that you could then alias it when running on a different command station.  What I've suggested with respect to command-station-assisted consisting could be accomplished tomorrow with off-the shelf hardware products (e.g., built-in wi-fi in all command stations) and some computer programming expertise to create an import/export app for iOS and Android, along with a wee bit of cooperation among DCC manufacturers to adopt some sort of common data transfer protocol to export and import consist data.  That would be true portability that the end user wouldn't have to think much about.

So at the end of the day, I just disagree with the notion that Advanced Consisting is somehow inherently better than command-station-assisted consisting.  It may be better on an NCE system because NCE has worked around the inherent limitations of the CV19 address; it ain't better on a Digitrax system, or an ECos system (you should read the current thread over at the ESU Yahoo Group on how to get Drive Hold to operate in a consist using an Ecos system), or a Zimo system, etc.  The real problem is that command-station interfaces are still built like it's 1985, instead of 2017.  Apple or Samsung could fix this in a heartbeat (well, OK, maybe a couple of months).

John C.

[peteski]

Ok John, I got it now.

However, I still am quite comfortable with the consist-specific info being kept in the decoder (rather than in command station). I do like the portability, especially with several units in the consist with complex lighting and/or sound customizations.  It is also very handy when breaking up a consist (as soon as CV19=0 the loco goes back to being a standalone loco with all of its original settings).   But I do agree that a single-CV addressing scheme is less than ideal.

I also agree that capabilities many DCC command stations are quite archaic.  Could there be a better way to consist? Yes, but I don't see that happening anytime soon.  Hopefully I'm wrong.

So I guess we just have to agree to disagree about usefulness and flexibility of advanced consisting. 

[jagged ben]

Preach John!   

Lol, I have never even tried Advanced Consisting.   For pretty much all the reasons you stated.  I do have a couple comments.   

One is that in Digitrax Unviersal consisting the function messages are not sent to all locos in the consist, only the motor commands are.  Which is how it should be, in my opinion,  especially if you're not running sound.  Me, I always thought that the default setup for lights was dumb (all right, what I mean is not prototypical), so I started programming the rear lights on my locos to F1 some time ago.   With Digitrax throttles with two locos selected at a time I always had the rear loco selected on the left if I needed to quickly switch to it to turn the rear light on and off. 

Of course along came sound and now I'm trying to figure out what to restandardize my functions on.  But whatever.  Other people don't run my locos hardly ever and if they did I'd make a cheat sheet.  I can do what I  want.

What I really think the solution is is better throttles to work with universal consisting.  I have somewhat outsize hopes for app programs.   I'd love to be able to add a couple knobs  for speed control to a touchscreen device that otherwise lets me flexibly setup what I want to see and control.  Unfortunately I do not see that much incentive for someone to set up exactly what I'd want and I  don't really have all the skills myself.

[John]

Whoever thought up the advanced consisting system within the NMRA specification should be tarred and feathered for that omission.  It's just stupid, plain and simple, like the whole notion of two-digit addressing was stupid (and which no one now uses).

There are 8 bits in a byte 

[robert3985]

Just to get things straight and quell any rumors after reading John's and Peter's posts that you have to have a novel-sized-set of instructions to consist locomotives using Digitrax...As everybody knows, I use Digitrax.  Okay...I have several consists that always run together.  When consisting them I use the Digitrax universal consisting feature and, before anything, I consist them to run together (using the 4-digit lead unit's number), then fool around with speed tables etc. to get 'em to run as close to the same as possible, even if they're the same engines, same manufacturer, bought at the same time, with the same brand of decoder in 'em...they almost ALWAYS run differently from each other. 

However, consisting with Digitrax universal consisting is easy...or am I in The Twilight Zone here???  All I do is call up the lead unit using the right throttle knob on my Super Throttle, then call up the next trailing unit using the left throttle knob, make sure they're running the same direction, run the lead unit down to couple up with the stationary trailing unit(s), then press the MU button on the Super Throttle, and answer the question the throttle asks me.  And I do it the same way with every unit in the consist, releasing the freshly consisted trailing unit that was previously being controlled by the left throttle knob, calling up another engine on the left throttle knob...making sure it's going the same direction as the other consisted units, back the consisted units (being controlled by the 4 digit address of the lead unit) on to the stationary B or trailing A, and push MU again, and answer my Super Throttle's question. (I don't have one in hand right now and I can't remember for sure, but I think it's a simple yes/no question, answered by pushing the "yes" or "no" button.).

Takes me about three minutes to consist a three-unit lashup...maybe less if I don't have to put engines on the track.

I've got sound in the trailing unit and lead unit, and no sound in the B's in my F's, E's, PA's, FA/FB's...and I can hear the sound just fine...my inertia settings work just fine with each individual engine so they start up and stop the same, go the same speed and retain their individual sound and light programming.

De-consisting is just as easy, and I don't have to program in their individual 4-digit addresses because they retain them in the decoder (I assume) or somewhere.  But, while consisted, they all run using the 4-digit number of the lead unit.

What's the deal with messing with CV19 and not being able to use 4-digit addresses, and "advanced" consisting??  I may not be using "advanced" consisting, but what I am using works GREAT for me, and it's as simple as can be to do.

Seeing no need for advanced consisting, maybe John can tell me and others what the advantages of it are (if any) over what I'm doing now???

EDIT:  Okay...so what in the content of this post deserves a thumbs down vote?  I don't get it... 

Cheerio!
Bob Gilmore

[jdcolombo]

There are 8 bits in a byte 

Yes, but you can use TWO bytes, like they did for 4-digit addressing.

John C.

[jdcolombo]

Just to get things straight and quell any rumors after reading John's and Peter's posts that you have to have a novel-sized-set of instructions to consist locomotives using Digitrax...As everybody knows, I use Digitrax.  Okay...I have several consists that always run together.  When consisting them I use the Digitrax universal consisting feature and, before anything, I consist them to run together (using the 4-digit lead unit's number), then fool around with speed tables etc. to get 'em to run as close to the same as possible, even if they're the same engines, same manufacturer, bought at the same time, with the same brand of decoder in 'em...they almost ALWAYS run differently from each other. 

However, consisting with Digitrax universal consisting is easy...or am I in The Twilight Zone here???  All I do is call up the lead unit using the right throttle knob on my Super Throttle, then call up the next trailing unit using the left throttle knob, make sure they're running the same direction, run the lead unit down to couple up with the stationary trailing unit(s), then press the MU button on the Super Throttle, and answer the question the throttle asks me.  And I do it the same way with every unit in the consist, releasing the freshly consisted trailing unit that was previously being controlled by the left throttle knob, calling up another engine on the left throttle knob...making sure it's going the same direction as the other consisted units, back the consisted units (being controlled by the 4 digit address of the lead unit) on to the stationary B or trailing A, and push MU again, and answer my Super Throttle's question. (I don't have one in hand right now and I can't remember for sure, but I think it's a simple yes/no question, answered by pushing the "yes" or "no" button.).

Takes me about three minutes to consist a three-unit lashup...maybe less if I don't have to put engines on the track.

I've got sound in the trailing unit and lead unit, and no sound in the B's in my F's, E's, PA's, FA/FB's...and I can hear the sound just fine...my inertia settings work just fine with each individual engine so they start up and stop the same, go the same speed and retain their individual sound and light programming.

De-consisting is just as easy, and I don't have to program in their individual 4-digit addresses because they retain them in the decoder (I assume) or somewhere.  But, while consisted, they all run using the 4-digit number of the lead unit.

What's the deal with messing with CV19 and not being able to use 4-digit addresses, and "advanced" consisting??  I may not be using "advanced" consisting, but what I am using works GREAT for me, and it's as simple as can be to do.

Seeing no need for advanced consisting, maybe John can tell me and others what the advantages of it are (if any) over what I'm doing now???

Cheerio!
Bob Gilmore

Hi Bob.

I use Digitrax, too, and I didn't mean to imply that using Digitrax's Universal Consisting (command-station-assisted consisting) was difficult.  Just the opposite, which is why I think Peteski overvalues the "portability" part of Advanced Consisting. 

But here's the advantage of Advanced Consisting.  With AC, you can set two CV's that tell each engine in the consist how to respond to Function commands.  Honestly, I've never needed this before even with sound.  Like you, I have two sound units in a 3-unit lashup.  I press F8 to turn the sound on for the lead unit and the rear unit, and that's that.  Under Digitrax's command-station consisting (Universal), function commands are only sent to the unit selected on the thottle, which will be the lead unit.  So if I press F2, the horn blows on the lead unit (only) and that's just the way I want it.  All good, no problems.

But now assume that you are very interested in trying out the Full Throttle features on the ESU LokSound.  Full Throttle allows you to vary the prime mover notch independently of speed (yes, you can also use manual notching, but that is VERY cumbersome, and again requires doing this for each individual locomotive in the consist using Universal Consisting).  With Full Throttle, you press F9 (Drive Hold), and the engines stay at the same speed; the throttle knob can now be used to vary the notching.  So if you want your engines screaming in Run8 while going 12mph up Sherman Hill, no problem.  They reach the top of the hill.  You disengage Drive Hold, and now you have speed control for coming down the hill.  Set it at 60mph for the downhill run, engage Drive Hold again, and cut the notch to Run 3 or whatever.  It sounds complicated, but it really isn't, and adds a great deal of realism to operations with grades and heavy trains.

BUT . . . you can only do this in a consist if EVERY LOCOMOTIVE IN THE CONSIST responds identically to F9 and F10 (the independent brake).  You can do this with Advanced Consisting: you set CV20 and CV21 so that each locomotive responds to F9 and F10 commands sent to the consist address (the address in CV19).  Now when you press F9, all three locos enter Drive Hold; when you press it again, all three locos return to normal throttle control.  This is the ONLY way you can set things currently to use the FT feature of the ESU LokSound.  You can't do it with Universal Consisting, because you'd have to select each engine individually on the throttle to send the F9 command.

But what I was telling Pete, who I think agrees with this part of my post, is that there is no reason other than antiquated interface design that the command station couldn't do this.  You COULD have a command station that would keep track of not only what engines are in a consist (like Digitrax does now), but also how each engine should respond to Function commands - and indeed, how the command station should send those commands to each engine.  It's just programming.  Right now you can't do this.  But you COULD do it, if manufacturers would build this capability into their command stations.  It isn't really even that hard programming wise - it would involve a couple of lookup tables that the command station would check when function buttons are pressed and it would all be seamless to the end user.

Peteski doesn't agree with me, but I think Advanced Consisting is an answer looking for a question.  In my view, the real problem is antiquated command station design.  Fix that, and there is no need for Advanced Consisting.

John C.

[jdcolombo]

BTW, Jagged Ben might be correct in that you could also do what I think should be done with "smart" throttles.  Suppose, for example, you had a THROTTLE (rather than a command station), that knew there were three engines in a consist, that 498 was the lead engine, and 426 was the rear.  It would be trivial for an appropriate throttle app (correctly programmed) to then send F1/F2 commands to the lead engine only, but F0 and F9 commands to all engines in the consist.   You'd have to have some way to input the information on how you want engines in a consist to respond to F commands, but you'd have to do that with my command-station idea, too.  A smart app on a wi-fi throttle that has 16gb of internal memory (e.g., a very cheap smart phone) could do this, I think.  Again, just programming and memory.

John C.