One or two DCC programming questions

[keeper]

Not really related to the question, but maybe for understanding what I'm doing (right or wrong....  )

This is my set-up:
Small loop as my programming track
Windows 10 computer
JMRI Decoder Pro and ESU LokProgrammer 5 installed
Power plug for
Digitrax PR3 and
the hardware part of said LokProgrammer
Connection from the computer to either the PR3 or LP by USB cable. They are both not connected at the same time.
Wires to the track

I haven't done much more than changing addresses so far, but I find it way easier than doing so with the handheld (DT4xx).

1.
When I enter an address everything is fine.
When I enter an address and a permanent consist address, the loco will only work with the consist address. It won't do anything with its own address. At least with the test throttle of the programme.
Maybe I don't get it right, but is this how it works?

2.
I haven't done any speed matching yet, especially non sound with sound units (both with ESU decoders). Is there a best practise?
Or is it more or less try and error?
When I search for that, there are so much answers, it is not easy to find the good ones.

Oh, and I have an issue which I don't know how to solve, because I tried what I have found so far.
The test throttle of the JMRI Decoder Pro does not work. @Ed Kapuscinski mentioned on another thread to upgrade the firmware of the PR3 which I did. At least the programme for doing so is telling me that the latest version is installed. But no loco is doing anything when I try to use the test throttle.
I hope a PR4 ist not the last solution...

Thanks for reading this. I hope it is not asked too much.

Thomas

[John]

Oh, and I have an issue which I don't know how to solve, because I tried what I have found so far.
The test throttle of the JMRI Decoder Pro does not work. @Ed Kapuscinski mentioned on another thread to upgrade the firmware of the PR3 which I did. At least the programme for doing so is telling me that the latest version is installed. But no loco is doing anything when I try to use the test throttle.
I hope a PR4 ist not the last solution...

Make sure the PR3 is not in programming mode -- otherwise you won't get any commands to the command station -- the top button changes modes

https://www.digitrax.com/tsd/KB169/pr3-operational-modes/

[skytop35]

With regards to question 1, yes that's how it works. If you program a DCC consist address with the Lokprogrammer, the engine will only respond to the consist address and no longer responds to the locomotive address.

[keeper]

Make sure the PR3 is not in programming mode -- otherwise you won't get any commands to the command station -- the top button changes modes

https://www.digitrax.com/tsd/KB169/pr3-operational-modes/

Ok, I will double check the correct setting, stand alone programmer, no command station, working with JMRI.
So far reading and writing decoder setting are no problem.

[keeper]

With regards to question 1, yes that's how it works. If you program a DCC consist address with the Lokprogrammer, the engine will only respond to the consist address and no longer responds to the locomotive address.

Ok, I understand.
So you can't uncouple one loco of the consist and move it away under it's own address.
Where is the advantage of having a permanent consist address? Or can this address be deactivated while operating with a handheld?

[mmyers]

Ok, I understand.
So you can't uncouple one loco of the consist and move it away under it's own address.
Where is the advantage of having a permanent consist address? Or can this address be deactivated while operating with a handheld?

The only way I'm aware of deactivating a Loco from an advanced (Decoder assisted) consist is to program CV19 to 00. That can be done while programming on the main.  It can be removed from a  Universal consist on the fly with a Digitrax throttle by having it selected on the left knob, pressing MU then the - button.  The best way is to not consist the engine and run it on its own address as a helper. The two knob Digitrax throttles work well this way with the helper selected on the left knob and the train on the right knob.
Martin Myers

[peteski]

Manually setting CV 19=0 (wither using POM or Service mode on the programming track) will remove it from the advanced consist (decoder will again behave as a standalone decoder), but if this consist was originally set up using the command station consisting procedure, the command station  will nto be aware that the loco was removed from the consist, That cna cause some unforeseen problems later on.

[jdcolombo]

A bit more background info that might be helpful.

Digitrax command stations can use two different kinds of consisting.  "Universal Consisting" is Digitrax's preferred consisting method.  With this method, the command station keeps track of the engines in the consist, but the engines receive commands from the command station at their "regular" DCC address.  So, for example, let's say you consist three GP9's, with engine #'s 561, 562 and 563.  With Universal consisting, the DIgitrax command station "knows" that these three units are in a consist and if you increase the throttle setting, it sends identical speed packets to all three engines at their regular DCC addresses: three separate packets, one for each address.  The advantage of Universal consisting is that you can have basically an unlimited number of consists (limited only by the command station's ability to keep track of them, which for a DCS100 or 200 command station, is essentially unlimited).  It also allows you to address commands to each individual engine easily.  If you want to turn on the headlight of the lead engine (561), you just press F0 while that engine is in the address line.  If you have a sound decoder and press F2 for the horn, only the lead engine will respond to that command.  You can send commands to consisted engines by calling up their number and pressing the appropriate F key for that command.  Etc.  The main disadvantage of this system is that multiple packets are required to control the consist, since each engine must be addressed separately at its own DCC address.   A second "disadvantage" is that if you remove the engines from your layout and run them on a different layout, they will no longer be consisted, because it is YOUR command station, and ONLY your command station that "knows" about the consist.  Any other command station will think these are simply separate engines.  Not usually a big deal, but if you are a club member and consistently move engines from one layout to another, it could get annoying.

Universal consisting is not part of the NMRA DCC standard.  Instead, the NMRA standard supports "Advanced Consisting" which uses CV 19 to program a "consist number" that is kept in the decoder, not the command station.   When you program a consist address from 1-127 in CV 19, the decoder responds ONLY to the consist number.  This means that things are simpler for the command station: it doesn't need to keep track of the engines in a consist.  Instead, it just sends a single packet of information to the consist address, and all locomotives with that same consist address respond to that single address.   But Function keys are not included in this system automatically.  Instead, CV's 21 and 22 determine how a decoder responds to a function key when you are using an Advanced Consist address.  By default, function commands are directed only at the individual locomotive address.  That means that to turn on a headlight while using a consist address, you would have to dial up the address of the individual locomotive to do that.  CV's 21 and 22 override this default system, but then all locomotives so programmed would respond the same way.  For example, if you program CV22 in your engines to respond to F0 at the consist address, ALL the consisted locomotives will turn their headlights on.  That's not exactly useful, either, and CV 21/22 operation requires programming all the individual locomotives separately to enable whatever functions you want to enable.

Some DCC manufacturers (e.g., NCE) get around the limitations of advanced consisting by having the command station keep track of the advanced consist and what engines are in it, and then allow you to, for example, independently address an engine in the consist from an advanced throttle so that you can easily break up a consist or "turn it around" with the rear engine in the lead.  It also translates the engine number to the consist number, so you don't have to remember the consist number for DCC operation.  But it's sort of a kludge and actually more complicated at the command station level than what Digitrax does with Universal Consisting.  Nevertheless, the advanced consist method DOES simplify the number of commands sent to a consist, and therefore can avoid "latency" if you are operating a large layout with lots of consists receiving commands simultaneously.  And because advanced consisting is contained in the decoder, and not the command station, the consist "lives" even if you move the engines from one layout to another.

Each consisting method (the Digitrax Universal method or the NMRA Advanced Consist method) has advantages and disadvantages.  For example, if I consist three locos with sound decoders on my Digitrax system, I have to remember to turn on the sound (hit F8) for EACH engine before an operating session.  With advanced consisting, I'd just call up the consist number and hit F8 once (assuming I had programmed CV21 to have the engines respond to F8 at the consist address, and all the engines would turn on their sound.  And certain advanced functions of ESU decoders (e.g., Drive Hold) really will only work correctly with advanced consisting and proper programming of CV21 or ESU's own command station. 

I think it is fair to say that given how cheap digital memory is these days and how cheap CPU's have become, keeping track of consists at the command station level, instead of the decoder level, is probably easier and more flexible IF the interface is designed well (Digitrax's interface, frankly, sucks, but so does NCE's.  If you want something more user-friendly and modern, you'll need to move to one of the European systems like ESU or Zimo).  But like most things, this is really more of a "what you get used to using" thing.  I use Digitrax, so I find advanced consisting a PITA - unless I want to use Drive Hold on an ESU LokSound consist, in which case advanced consisting is really the only way to do it.  But I also operate on a friends NCE system, which uses advanced consisting with command station help, and it works well 99% of the time (not so well when you have a single engine that has the same DCC address as a consist address the command station has randomly assigned . . .)

John C.

NOTE:  Edited to correct errors regarding the CV used to control consist addressing and CV's 21 and 22.  More explanation provided in my second message below.

[peteski]

John,
I'm not sure if I follow you with the Advanced Consist and address in CV 17 and 18.  Those registers hold the "regular" 4-digit (long) address. Any loco using long address will use those registers to hold its address. Unrelated to consisting.

Did you mean CV 19 (and 21.22)?

[jdcolombo]

OOPS.

Yep.  Got my address CV's mixed up.

I've now edited that post to correct it.

As I noted in my original post, both consisting systems have problems.   For example, if you use a consist address to run a consist with a Digitrax system, you would have to address the lead engine at its "regular" DCC address to turn on the headlight.  OK, so let's say you do that, then dispatch that address and acquire the consist address to move the consist.  Now let's say you want to do a run-around move.  You would uncouple the consist from the train.  Then you would need to acquire the former lead engine's address to turn the headlight off, then acquire the former rear engine's address to turn it's headlight on, then re-acquire the consist address to actually move the consist.  You COULD use CV22 to have F0 respond to the consist address in the front and rear engines, but this requires programming CV22 in EACH ENGINE before you use them in a consist, and then the rear engine will have its headlight on while the consist is moving.  That isn't prototypical, but might not bother a particular operator.  And if you're using a sound decoder, you basically HAVE to program CV21 to have F2 respond to the consist address to make it work.  But if you do this for, say, both the lead and rear engines, then BOTH those engines will blow their horns when you press F2.  So you probably don't want to do that, but if you then want to do a runaround move that puts the rear engine in the lead, you'd have to reprogram CV22 using ops mode programming on the main so that the correct engine blows its horn.  Operationally, this is a nightmare if you're using Digitrax.

NCE's system avoids all this complication by using the command station to "insulate" the operator from the background CV programming.  When you form a consist with an NCE system, it just automatically picks an unused consist number from the 127 available in CV19, programs the engines with that number as you add them to the consist, and then allows you to address the front and rear engines directly for things like turning the headlight on and off or blowing the horn.  But watch out - if you have an engine that is programmed to a 2-digit address and the NCE command station picks that address for a consist address, suddenly your stand-along engine will start running with the consist.  And even with NCE, you still have to program CV's 21 and 22 to make all engines respond to certain commands, like F8 to turn on sound. 

All this is a result of trying to conserve digital memory and processing efficiency back in the days when memory was expensive and processors were infantile.  Today, digital memory is cheap and processors that have a 1/4" square profile are as powerful as the room-sized computers used in the Apollo moon missions.  So it seems to me there should be an easier way to do all this, probably via command station memory and processing and a better interface.  I think the new DCC systems from ESU and Zimo might accomplish most of this, although I haven't actually used them in an operating situation, so I'm not sure.

John C.

Edited to correspond to edits to prior message and eliminate duplicate information

[keeper]

John, thanks for the info. I understand some things a lot better now.

I'm running my layout with a Digitrax Empire Builder set (DT400 and DB 150 command station).
Most locos have a Digitrax decoder.
That's why I thought programming the locos with the above mentioned set-up might be a good idea.
Digitrax decoders are becoming more and more expansive here in Germany and so I started to buy ESU decoders and
I find the LokProgrammer easier to use than JMRI with the PR3.

With the info from John, a command station from ESU might be something to consider. There is a Loconet converter available from ESU, so all existing stuff should be usable.

But first I should think about the consist variants before anything else.

Thomas

[peteski]

European DCC systems might handle consisting differently than the Amarican ones.  While the 3 types of consisting remain the same, the way they are created and handled by the command station might be different.  While I have no personal experience with either ESU or ZIMO DCC systems, I belong to the groups.io groups for both systems, and reading posts there makes me think they do things differently. If you are used to Digitrax, you might have to learn how they handle consisting.

If you are thinking of purchasing either ESU or ZIMO DCc system I recommend studying the manuals, and maybe even joining those groups for questions specific to those systems.  Having said that I believe that few members here own the ESU DCC system, so they might be able to give you more specific info on it.

[keeper]

One thing thing is for sure, I won't buy a Zimo system just because of the price tag.
From what I have found out so far, and depending on the hardware, Zimo is at least twice as expensive as ESU.

[peteski]

I have heard of another European DCC manufacturer in Netherlands: Digikeijs

[keeper]

Looking at the clock, are you ever sleeping?