A couple of weeks ago, I downloaded JMRI Decoder Pro and created a roster of all of my locos. Several were in consists, so I went through the process of killing the consist before adding the individual locos to the roster on Decoder Pro. At that time, I noticed that even though the consist was no longer there, the consist number was still in CV19, so I had to zero it out. This only happened on the sound decoder (QSI and Tsunami) consists. On the non-sound decoder consists, the consist number in CV19 was zero, as expected.
Now, when I re-create a consist with locos with sound decoders, the consist is created, but the locos don't respond. They don't respond because the number in CV19 remains zero. This only happens with sound decoders. So, using POM, I key the consist number into CV19, and then the consisted locos all respond. The non-sound decoders correctly add the value to CV19 when setting up a consist.
Any idea what is going on here?
Rich
Alton Junction
Rich.
What system are you using? Lenz does things different than NCE and Digitrax.
Pete
I pray every day I break even, Cause I can really use the money!
I started with nothing and still have most of it left!
Are you NCE, Digitrax, or other? I don't know about D and O, but NCE stores the consist alias number in the command station. Are you deleting the consists using decoderpro, or deleting through the command station? And are you making up the consists using the command station or decoderpro?
And are you deleting the consist using one of the cab numbers on the consisted locos, or deleting the alias number?
I don't know for sure, but maybe something is getting confused.
I do know that I have had the same problem as you with some decoders where I delete/clear the consist using the loco number. I believe that if I delete the consist using the alias number all the CV 19s return to zero, but I'm not 100 percent certain. I have gotten into the habit of checking if the loco moves using its own address after deleting the consist just to make sure that nothing has been retained in CV19.
In any case, if you try to reconsist a loco whatever new number you try to put in CV19 should overwrite the existing value.
maxman Are you NCE, Digitrax, or other? I don't know about D and O, but NCE stores the consist alias number in the command station. Are you deleting the consists using decoderpro, or deleting through the command station? And are you making up the consists using the command station or decoderpro? And are you deleting the consist using one of the cab numbers on the consisted locos, or deleting the alias number? I don't know for sure, but maybe something is getting confused. I do know that I have had the same problem as you with some decoders where I delete/clear the consist using the loco number. I believe that if I delete the consist using the alias number all the CV 19s return to zero, but I'm not 100 percent certain. I have gotten into the habit of checking if the loco moves using its own address after deleting the consist just to make sure that nothing has been retained in CV19. In any case, if you try to reconsist a loco whatever new number you try to put in CV19 should overwrite the existing value.
Sorry, I should have mentioned the system. I have an NCE PH-Pro 5 amp, wireless.
All of my consisting, making up consists, killing consists, are all done through the command station. I only used Decoder Pro to create a roster, but I haven't done anything since with Decoder Pro.
When I kill a consist, I begin my deleting the rear loco by long address. Then, I kill the consist by referencing the consist number, not the lead loco address.
In order to kill a consist with the NCE PowerHouse Pro, all locomotives in the consist must be on the track when you press the CLEAR key, and you MUST know the consist address (which locomotive number is being used to call up the consist). If you don't do this correctly, the consist alias number will remain in the decoders and the command station.
After doing this, I sometimes have to program each individual locomotive's CV19 to a value of zero to make sure the consist has been cleared from the decoder.
At our club layout, we have over 290 decoder-equipped locomotives entered into the database, and members frequently kill or create consists without going through the proper procedures, which causes the command station consist stack to become overly cluttered with data that is no longer relevant.
Something we've learned the hard way: NEVER use a short address of less than 128 for a locomotive decoder, because the command station uses these numbers as the consist aliases. If a locomotive has a short address of, say, "10" we program the decoder to "1010" or "1000" to make sure it is a LONG address number higher than 128; otherwise some really strange things can happen when someone tries to run a locomotive with a short address that the command station has assigned as a consist alias. CRASH! BANG! Runaway trains!
cacole In order to kill a consist with the NCE PowerHouse Pro, all locomotives in the consist must be on the track when you press the CLEAR key, and you MUST know the consist address (which locomotive number is being used to call up the consist). If you don't do this correctly, the consist alias number will remain in the decoders and the command station. After doing this, I sometimes have to program each individual locomotive's CV19 to a value of zero to make sure the consist has been cleared from the decoder. At our club layout, we have over 290 decoder-equipped locomotives entered into the database, and members frequently kill or create consists without going through the proper procedures, which causes the command station consist stack to become overly cluttered with data that is no longer relevant. Something we've learned the hard way: NEVER use a short address of less than 128 for a locomotive decoder, because the command station uses these numbers as the consist aliases. If a locomoitve has a short address of, say, "10" we program the decoder to "1010" to make sure it is a LONG address; otherwise some really strange things can happen when someone tries to run a locomotive with a short address that the command station has assigned to a consist. CRASH! BANG! Runaway trains!
Something we've learned the hard way: NEVER use a short address of less than 128 for a locomotive decoder, because the command station uses these numbers as the consist aliases. If a locomoitve has a short address of, say, "10" we program the decoder to "1010" to make sure it is a LONG address; otherwise some really strange things can happen when someone tries to run a locomotive with a short address that the command station has assigned to a consist. CRASH! BANG! Runaway trains!
I should mention that I never had a problem until I created the roster in Decoder Pro. I am not blaming Decoder Pro, but before I created the roster, I could setup and kill consists without any difficulty. But, now, I have the problem that I described in my initial post. I create the consist, but the consist number is not stored in CV19. I have to do it manually to get the newly created consist to respond.
Regarding addresses, most of my locos have 4-digit addresses. I have a few 3-digit long addresses, but all are greater than 127. I have a few 2-digit cab numbers (38, 46, 47, and 83), but they become 3-digit addresses (381, 382, 461, 462, 471, 472, 831, and 832).
Perhaps you're not performing the consist setup correctly, but most likely it is a decoder problem. I have ran into some decoders that don't allow an advanced consist (namely, Bachmann, MRC and early versions of the LokPilot).
cacole Perhaps you're not performing the consist setup correctly, but most likely it is a decoder problem. I have ran into some decoders that don't allow an advanced consist (namely, Bachmann, MRC and early versions of the LokPilot).
Maybe, but I am not doing anything different than what I did before. For some reason, when I set up the consist, the address ( 1 to 127) is not automatically placed in CV19.
I am going to go down to my layout and try it again.
I too have an NCE Power Pro connected to Decoder Pro. However, I haven't used DP for consisting because it is easier to do with the throttle. (Advanced consisting) After you create a consist, you can address it by either the consist number, the lead loco number, or the rear loco number. Thus you don't NEED to remember the consist number. As long as all the locos remain on the track, you can also clear a consist using either the consist number, the lead loco number, or the rear loco number. You don't need to remove any locos beforehand. Why do I do it this way? Because it simulates what the railroads would do when lasing up locos for a train. The crews would hook them up or brake them down as the need arose. So I let my crews do it as well. I think if you use the old style consisting, then CV 19 gets written to because you are setting up the consist in the decoder. That is most likely the way DP does consisting.
I have also created a roster in DP, but have not experienced the problem that you have.
Elmer.
The above is my opinion, from an active and experienced Model Railroader in N scale and HO since 1961.
(Modeling Freelance, Eastern US, HO scale, in 1962, with NCE DCC for locomotive control and a stand alone LocoNet for block detection and signals.) http://waynes-trains.com/ at home, and N scale at the Club.
Make sure the decoders are not locked in CV15 and 16. Some decoders but not all will not let CV19 change if it is locked.
richhotrainWhen I kill a consist, I begin my deleting the rear loco by long address.
It is not clear why you would want to do this if you are killing the entire consist.
richhotrain I have a few 2-digit cab numbers (38, 46, 47, and 83), but they become 3-digit addresses (381, 382, 461, 462, 471, 472, 831, and 832).
This is one way of doing it, but gets confusing in a club environment because any other operator would look at the loco and not know what your particular convention is. I think it is much easier to just create the long address by placing a zero in front to the two digit number: 038, 046, 047, and 083. A lot easier to explain.
locoi1sa Rich. Make sure the decoders are not locked in CV15 and 16. Some decoders but not all will not let CV19 change if it is locked. Pete
CV15 and CV16 are not locked in any of the non-sound decoders or the Tsunami sound decoders. It appears that the QSI sound decoders do not support CV15 and CV16. Those two CVs don't even show up in the QSI manual or on my loco roster on Decoder Pro.
maxman richhotrain When I kill a consist, I begin my deleting the rear loco by long address. It is not clear why you would want to do this if you are killing the entire consist.
richhotrain When I kill a consist, I begin my deleting the rear loco by long address.
maxman, your question got me to thinking, and then to doing some reading, and now I am totally baffled. You are correct in that the manual makes no mention of deleting the rear loco as the first step in clearing a consist. Now, I am trying to recall why I do that.
If I recall correctly, at one time, I was unable to clear a consist, and I was told that you have to delete the rear loco first. I wonder if that had something to do with the fact that I had the old EPROM chip in the command station. A few years back, I got the most recent EPROM ugrade chip which was released to correct a bunch of software errors in the original chip.
Anyhow, I will spend some time this morning doing it as the manual directs and report the results. Thanks for bringing this to my attention.
richhotrain maxman richhotrain When I kill a consist, I begin my deleting the rear loco by long address. It is not clear why you would want to do this if you are killing the entire consist. maxman, your question got me to thinking, and then to doing some reading, and now I am totally baffled. You are correct in that the manual makes no mention of deleting the rear loco as the first step in clearing a consist. Now, I am trying to recall why I do that. If I recall correctly, at one time, I was unable to clear a consist, and I was told that you have to delete the rear loco first. I wonder if that had something to do with the fact that I had the old EPROM chip in the command station. A few years back, I got the most recent EPROM ugrade chip which was released to correct a bunch of software errors in the original chip. Anyhow, I will spend some time this morning doing it as the manual directs and report the results. Thanks for bringing this to my attention. Rich
I don't think that what you are doing has anything to do with your original issue, but I was curious none the less.
There is a way where you are supposed to be able to separate one loco from the consist, for example where you would have a switcher in the consist that was being hauled from the shop to be set out at the yard where it normally works. However, I don't know exactly what happens in the command station when you do this. When consisting NCE always asks you for the first loco, the last loco, and then any additional locos. If that switcher was the last loco and you set it out, is the command station smart enough to figure out that the next loco in line is the new "last" loco"? (that was a rhetorical question...not looking for you to answer).
In any case, I still say that the best way to kill a consist is to use the alias address to do the killing.
I also think that if you reconsist an already consisted loco with a new consist number, the new number will overwrite whatever was already in CV19. However, if you do this without killing the original consist I think you will end up with that same loco in multiple consists in the brains of the command station and not remember that you did so. We have this problem down at the club, where folks who don't know what they're doing (less than me, anyway) put a loco in one consist, forget, and then re-consist to something else.
Have you ever use the "browse consist" function to scroll through the consists contained in your command station? You might be surprised as to what you find.
maxman Have you ever use the "browse consist" function to scroll through the consists contained in your command station? You might be surprised as to what you find.
Yeah, I have come to rely on the Browse Consists feature. It is very useful, but not too many surprises there.
OK, I went down to the layout this morning and peformed a series of tests. Here are the results.
First, I was able to clear a consist without having to first delete the rear loco. For the longest time, I had been deleting the rear loco as the first step in clearing a consist, and I no longer remember why I first starting doing that other than to recall that, at one time, I could not clear a consist by simply following the procedure in the manual.
Second, using a pair of non-sound locos, I was able to set up a consist and I was able to clear that consist with no problem. Incidentally, the value in CV19 was zero after setting up the consist. I must be misunderstanding CV19, because I thought that the number of the consist became the value in CV19.
Third, operating my wireless NCE PH-Pro, I set up a consist with a pair of sound equipped locos (QSI) with the Pro Cab untethered. It worked fine. Then, I cleared the consist. I could not then get the individual locos to move. So, I went back and cleared the consist again, and then I was able to get the individual locos to move.
Fourth, I tethered the Pro Cab and set up a consist with the same pair of sound equipped locos. It again worked fine. Then, I cleared the consist. The individual locos moved without having to clear the consist a second time.
So, the problem seems to be consisting sound locos with the Pro Cab untethered.
I have to say that various forms of programming with the Pro Cab untethered has been an ongoing problem in other respects. My system started out as a wired system, and I later added the wireless base station and antenna. I have two Pro Cabs and both have been upgraded to wireless by NCE.
I have discussed my issues in this regard on the NCE-DCC forum in the Yahoo group. The following language from the PH-Pro manual was quoted to me. It reads:
"When performing complex tasks involving many messages displayed on theProCab, communications between the cab and command station may take longerthan expected thus slowing down your programming. In these instances you mayfind it more expedient to plug in the cab while doing extensive programmingor system setup."
What can I say?
Are you maybe setting up what NCE calls "Old style" consists? Those don't use CV19, instead the command station remembers everything and sends commands to all locos in the consist.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
rrinker Are you maybe setting up what NCE calls "Old style" consists? Those don't use CV19, instead the command station remembers everything and sends commands to all locos in the consist. --Randy
No, I use Advanced Consisting, but you raise a good point.
One time, on someone's recommendation, I used Old Style Consist to program the dual decoder Blueline locos.
To get them working right with Advanced Consisting, you have to manually add the consist number as the value in CV19.
Maybe that is why I thought that CV19 is automatically programmed in Advanced Consisting.
richhotrain Maybe that is why I thought that CV19 is automatically programmed in Advanced Consisting. Rich
richhotrainI have to say that various forms of programming with the Pro Cab untethered has been an ongoing problem in other respects.
richhotrain"When performing complex tasks involving many messages displayed on theProCab, communications between the cab and command station may take longerthan expected thus slowing down your programming. In these instances you mayfind it more expedient to plug in the cab while doing extensive programmingor system setup."
I believe that there is some truth to the NCE forum response. However, I have another question. When you are attempting to program untethered, where is the base station antenna in relation to where you are doing the programming? I mentioned in another thread that I thought that having the antenna/RB02 hanging from the ceiling was a better idea than putting it on a shelf next to the command station. Others basically blew off that thought. However, if one were to read the NCE manual for the RB-02, it mentions that the human body can block radio waves in the frequency band used by the cab radios. Hanging the antenna from the ceiling is also discussed.
I think it is also possible to have an issue if your cab is too close to the radio base station.
Here is a link to the instruction if you don't have it: http://www.ncedcc.com/images/stories/manuals/rbo2_rpt.pdf
I don't think any of the above has anything to do with your CV 19 issue, other than potentially slowing down the programming process. But I agree with Cacole...if you are actually using advanced consisting, there has to be a value assigned to CV 19, and that is the alias number assigned to the consist by the command station. I believe that you said that you found that there was no number assigned to CV19 after consisting. How did you determine this? Consisting is done on the main, but you can't read back any CV values unless you move the loco to the program track (and you probably won't be able to read back any sound decoder CVs with the PowerPro anyway). Did you move the locos to the programming track to determine what value was in CV 19?
cacole richhotrain Maybe that is why I thought that CV19 is automatically programmed in Advanced Consisting. Rich You're right, CV 19 should automatically be set with the Advanced Consist number when you set up the consist with the NCE system. I have never had any problem with that not happening. You may need to contact Larry at NCE -- your base station or throttle(s) may need to be upgraded. Another possibility I just thought of: A dead battery in the Command Station causing consist data to be lost when the system is turned off. But even if this is happening, the consist number should still be in CV19 in the decoder. You wrote that your last upgrade was "a few years ago" -- It might be advisable to see which EPROM version you have (how to find out should be somewhere in the manual) and then check with NCE to see if you have the latest. Throttles, too.
cacole, thanks for that post. My NCE pH-Pro system does have the current updates to the EPROM, RB02, and the Pro Cabs.
maxman, after I read your recent reply, and cacole's recent reply, I went back and did some additional tests.
This morning, I set up an Advanced Consist of two sound-equipped locos (QSI) with a tethered Pro Cab, and the locos would not move after programming the consist. I then put each loco on the Programming Track and checked CV19. The value was zero in each loco. So I changed the CV19 value to the consist number (50) in each loco and then there was movement.
I then set up an Advanced Consist with a pair of non-sound locos with a tethered Pro Cab and there was movement. Checked each loco on the Programming Track and the CV19 value on each loco was the consist number (111). So, the problem seems to be with sound-equipped locos.
The NCE-DCC forum has suggested adding a repeater. I am going to call Larry at NCE today to discuss this whole issue.
I agree that the position of the wireless base station/antenna is an important issue as well as the proximity of the Pro Cab to the antenna. I just seem to be having all kinds of issues when programming sound decoders.
I will report back with what Larry at NCE has to say about all of this.
richhotrainThe NCE-DCC forum has suggested adding a repeater. I am going to call Larry at NCE today to discuss this whole issue.
One thing to know about adding repeaters is that it is possible to add too many of them. Friend of mine has his railroad in a two-room structure. One room is about 30 X 80, the other is about half-again as big. Right now I think he has the base station and two repeaters set up for the whole thing. He had several more repeaters installed originally and had a bunch of problems. Turned out that they were interfering with each other.
Unless you turned it off, it should be very obvious when QSI decoders get new programming - the verbal response whould speak out the setting, and say "CV 1 9 equals" whatever consist number you have used. If you HAVEN'T turned off verbal respnse - try that, it can sometimes interfere with programming those decoders.
Since this is happening on the main, there should be no issues of insufficient power to program,
maxman richhotrain The NCE-DCC forum has suggested adding a repeater. I am going to call Larry at NCE today to discuss this whole issue. One thing to know about adding repeaters is that it is possible to add too many of them. Friend of mine has his railroad in a two-room structure. One room is about 30 X 80, the other is about half-again as big. Right now I think he has the base station and two repeaters set up for the whole thing. He had several more repeaters installed originally and had a bunch of problems. Turned out that they were interfering with each other.
richhotrain The NCE-DCC forum has suggested adding a repeater. I am going to call Larry at NCE today to discuss this whole issue.
Yeah, I have heard that. When I first went wireless, I asked Larry at NCE about that, and he suggested starting out with no repeaters, so I have none. If I add one now, it would be the one and only one.
rrinker Unless you turned it off, it should be very obvious when QSI decoders get new programming - the verbal response whould speak out the setting, and say "CV 1 9 equals" whatever consist number you have used. If you HAVEN'T turned off verbal respnse - try that, it can sometimes interfere with programming those decoders. Since this is happening on the main, there should be no issues of insufficient power to program, --Randy
When I program a consist, there are no verbal responses throughout the process. However, if I have to program the CV19 value on the main, there is a verbal response, e.g., CV19=50. But, I think that is how all of that is supposed to work.
Well, I just talked to Larry at NCE, and i had an interesting discussion with him.
After going through my issues with him, he thought that the main problem was the QSI decoder, particularly since I have no problems with non-sound decoders.
Apparently, others have experienced similar problems with the older QSI decoders. Some need upgrade chips, others have no upgrade chips available. Mine do not have upgrade chips available. Larry thinks that the QSI decoders "miss" certain commands, as they react to the programming steps. Interestingly, after talking to Larry, I did some Googling on the subject. On another model railroad forum, the suggestion was made to get the locos moving first and making sounds, then program the consist. I haven't tried that yet, but I will. Another forum member said that he automatically programs the consist number into CV19 for each loco.
Larry did not think that I needed a repeater, given my setup. The fact that the LED on each ProCab is blinking so rapidly indicates that I repeater would not be needed.
His only other suggestion was to send in the ProCab and they could check it out. But, for now, I am just going to live with the problem. Once, I re-program my consists, I have no need to clear them. The only reason that I cleared the consists in the first place was to add all of the locos as roster entries with Decoder Pro. Once the consist is programmed, it is not lost when I power down and then power up again.
This is my theory on why QSI sometimes do not get an address set on some systems - even with verbal response off, the decoder still has the delay, so if the system sends the change for CV29 too quickly after it sends CV17 and 18, it would miss it. This happens if I use the console of my Zephyr. if I use a DT40x throttle with my Zephyr, it programs fine - because on the DT40x throttles, after it sends CV17 and 18, ir prompts you to enable 4 digit addressing, so no matter ho fast you are with the buttons, there's a built in delay, and it works.
However - for setting a consist, that's just one CV, dunno how it can 'miss' one CV. Unless since it's Ops Mode, it misses a packet with its address on it - however, you can test that easily enough by making a one loco consist and then checking CV19 - if the system is only sending out a CV19 change to one address, the decoder can;t get confused.