开云体育

I installed my Sprog in JMRI 4.14. Everything works except the window at the bottom of the screen that allows you to read and write changes. I have read the directions, but must be missing something. How do I get this part to appear??

I've had some of that discussion with Bob a few months back. My personal leanings for automatic control are to having all multi-path choice track entities as function specific objects which have all the I/O parameters necessary to be utilized in the simplest control terms without adding further control logic. And to that end, yes, I'd expect a esp32 like device to be a suitable and complete Grand Union controller. including setting the co-located road traffic signals in parallel with the separate trolley movement decisions.

Andy

On 5/25/2019 1:34 PM, rphughesnj@... wrote:

Andy,

You would only need to use a "closed/thrown" sensor if you wanted to know the actual state of a turnout, not just what it was told to do.? As I mention in the my reply to Dick,? what I had in mind was messaging the "ONESENSOR" turnout feedback in the JMRI documentation.

As to where you would put them, one PI or esp32 based MQTT node should be able to handle 8-16 or more (depending on the chip) I/O control lines.? Couldn't you could match up the turnout control with the turnout feedback sensor on the same controller?? Or couldn't you have turnout only nodes and sensor only nodes? Whatever works for you.

In you example, wouldn't "routes" be a solution to throwing multiple turnouts at the same time?? No reason a turnout node could not be setup to handle routes.? A bit more complicated, but I think doable. Looking at your photo I see a lot of what I consider crossovers.? Couldn't you just wired the turnout motors to code/throw turnouts with one command?


richard

---
This email has been checked for viruses by AVG.

I installed my Sprog in JMRI 4.14. Everything works except the window at the bottom of the screen that allows you to read and write changes. I have read the directions, but must be missing something. How do I get this part to appear??

Thanks in advance for any help.

Rance Thompson

If your loco mounted decoder is AI engineer smart and can use just basic pattern recognition to "read" signals, obstructions, distances, and other RR objects you don't need the "air" bus for automation either.

805 888 1284

On 5/25/2019 1:18 PM, rphughesnj@... wrote:

Dick,

Frankly, I don't think I need ID sthat is unique all over the world, just unique within my local universe, my local MQTT network.? Even if I was was building modular pieces that hooked with others I could simply use a prefix, like my initials, to make node names unique enough.? I would much rather deal with names line "trains/track/turnout/harris01/platform06" than an LCC events like E1:0B:C7:56:DE:7B:D5:F7.? Just like I would rather deal with "www.google.com <>" rather "172.217.5.4".

As far as associating a sensor on a turnout to the turnout itself I will point you to the JMRI way of doing turnout feedback.? My example would fit the "ONESENSOR" mode described in the documentation.? JMRI, in this example, would make the association.

The reason I have spent so much time looking at buses and messaging is that I want to automate my layout.? I exchanged many messages on the OpenLCB trying to find out how do railcom, an NMRA standard, using LCC.? How do I message {"status" : "entered", "ident" : "2359"} in LCC?? No one knows.? JMRI can handle "reporters", but as far as I can see, LCC cannot, today. I was told that maybe there will be defined method by the end of the year.

As a alternative I looked at Internet messaging using MQTT (partially supported by JMRI) and I have been able to prototype just about everything I need on a model railroad using MQTT and TCP messaging.? Turnouts, sensors, signals, command station control, throttles, etc.? Using off the shelf hardware and a minimal amount of coding it all works.

For my two cents, we all need model railroad specific software (JMRI, RocRail, Railroad and Co, etc) and we all need model railroad specific hardware (occupancy detection, stall motor control, signals, etc).? What we don't need today, in my opinion, is a model railroad specific hardware control bus connecting the two. The size and price points of Internet connecttion hardware has really shrunk in the last few years.? Most of the existing model railroad buses were designed 10-20 years ago.? Yes, they work, but they don't interoperate.? Think about a video doorbell. Can you envision how big they would have been 10 years ago?? They have cameras, speakers, microphones, motion sensors, and run off low voltage door bell current. And they can communicate with my iPhone.? I contend that the currently available hardware allows for a model rail bus built using inexpensive WIFI/Ethernet hardware and Internet messaging.

On the topic of WIFI, I don't think it will be a problem.? First, I plan on using a separate WIFI access point for MQTT traffic. Also, I am planning to run many devices from one WIFI node. Analogous to the RR-CIrcits Tower LCC.? Image a Tower MQTT. Therefor I will not have that many WIFI nodes.? Also WIFI will only be used where I don't have a hard wire connections.? For example, PIs I deploy will all be hard wired including the MQTT broker server.


richard

---
This email has been checked for viruses by AVG.

Andy,

You would only need to use a "closed/thrown" sensor if you wanted to know the actual state of a turnout, not just what it was told to do.? As I mention in the my reply to Dick,? what I had in mind was messaging the "ONESENSOR" turnout feedback in the JMRI documentation.

As to where you would put them, one PI or esp32 based MQTT node should be able to handle 8-16 or more (depending on the chip) I/O control lines.? Couldn't you could match up the turnout control with the turnout feedback sensor on the same controller?? Or couldn't you have turnout only nodes and sensor only nodes? ? Whatever works for you.

In you example, wouldn't "routes" be a solution to throwing multiple turnouts at the same time?? No reason a turnout node could not be setup to handle routes.? A bit more complicated, but I think doable. Looking at your photo I see a lot of what I consider crossovers.? Couldn't you just wired the turnout motors to code/throw turnouts with one command?


richard

Dick,

Frankly, I don't think I need ID sthat is unique all over the world, just unique within my local universe, my local MQTT network.? Even if I was was building modular pieces that hooked with others I could simply use a prefix, like my initials, to make node names unique enough.? I would much rather deal with names line "trains/track/turnout/harris01/platform06" than an LCC? events like E1:0B:C7:56:DE:7B:D5:F7.? Just like I would rather deal with "" rather "172.217.5.4".

As far as associating a sensor on a turnout to the turnout itself I will point you to the JMRI way of doing turnout feedback.? My example would fit the "ONESENSOR" mode described in the documentation.? JMRI, in this example, would make the association.?

The reason I have spent so much time looking at buses and messaging is that I want to automate my layout.? I exchanged many messages on the OpenLCB trying to find out how do railcom, an NMRA standard, using LCC.? How do I message {"status" : "entered", "ident" : "2359"} in LCC?? No one knows.? JMRI can handle "reporters", but as far as I can see, LCC cannot, today. I was told that maybe there will be defined method by the end of the year.?

As a alternative I looked at Internet messaging using MQTT (partially supported by JMRI) and I have been able to prototype just about everything I need on a model railroad using MQTT and TCP messaging.? Turnouts, sensors, signals, command station control, throttles, etc.? Using off the shelf hardware and a minimal amount of coding it all works.

For my two cents, we all need model railroad specific software (JMRI, RocRail, Railroad and Co, etc) and we all need model railroad specific hardware (occupancy detection, stall motor control, signals, etc).? What we don't need today, in my opinion, is a model railroad specific hardware control bus connecting the two. The size and price points of Internet connecttion hardware has really shrunk in the last few years.? Most of the existing model railroad buses were designed 10-20 years ago.? Yes, they work, but they don't interoperate.? Think about a video doorbell.? Can you envision how big they would have been 10 years ago?? They have cameras, speakers, microphones, motion sensors, and run off low voltage door bell current. And they can communicate with my iPhone.? I contend that the currently available hardware allows for a model rail bus built using inexpensive WIFI/Ethernet hardware and Internet messaging.

On the topic of WIFI, I don't think it will be a problem.? First, I plan on using a separate WIFI access point for MQTT traffic.? Also, I am planning to run many devices from one WIFI node.? Analogous to the RR-CIrcits Tower LCC.? Image a Tower MQTT.? Therefor I will not have that many WIFI nodes.? Also WIFI will only be used where I don't have a hard wire connections.? For example, PIs I deploy will all be hard wired including the MQTT broker server.


richard

Scott,

The files are independent, but do address each other via the object id fields.? I would be careful editing the xml files, as there isn't any checking when loading them.?

Dan

Forgot to mention. I have a Fleischmann Twin Center for digital control.

Hi,
I am not just new to computer controlling modeltrains, I’ve only beenin the modeltrainworld for about 3 months.
I am building my H0 track with switches that are not computercontrolled, but they are switched electrically. I do this because I can use my track both analog and digital that way (not at the same time ofcourse).
Can I use JMRI simply to run my trains speed and lights and stuff, while still doing the switches manually?
For now I don’t want to completely make everything digital but it would be nice to control just the trains via computer.

Greets,
Dennis

How should a “MOLE YARD” be treated for staging / switching I have split it into 2 yards on paper. What I want is train “X” to come into the yard cars to track “Y” locomotives and caboose to their respective tracks and terminated. Inbound WB cars would be used on a Outbound EB train.

I don’t have outside of my A/D tracks #1 and #2 in the yard that are long enough to support a train length.

Any suggestions?

Thanks,

Mike

Thanks Dan,

I was looking at a previous post (#89562) about copying just a single XML file from the Backup directory into the current setup.

Now when I complete my testing, I was thinking about just copying the "OperationsCarRoster.xml" file from my backup into the current setting then manually adding all the Custom loads like you suggested.? It would save me adding all the Schedules in the "OperationsLocationRoster.xml"? file. (I don't anticipate changing any Locations or Tracks)

Is this possible, or?do all?6 of the XML backup files needed to be Restored as a set?

-Scott-
(aardtechus)??

My questions are:

1. Where are the Custom Loads stored in the program? Are they in the "OperationsCarRoster.xml" file?

Yes.

1. Where are the Schedules stored in the program?? Are they with the "OperationsLocationRoster.xml" file?

Yes.

1. Will I be stuck with having to manually re-stage all my cars after I am happy with the new setup?

Maybe.

1. Is there another way to achieve what I want to do??

I would create a new profile just for testing.? You would then restore your backup file in the new profile and experiment with the program, making notes about new custom loads and schedules.? Since custom loads and schedules are pretty easy to add and modify, once you're happy, switch back to the original profile and enter your changes.

To restore a backup file from a new profile, in the Setup window of the new profile, use the Tools menu, and select "Restore".? This will allow you to search for your backup file on your PC.

Dan

Hi Paul,

I tried setting the check buffer to no but still get the timeout errors.

I have been using Operation Pro for quite while and really enjoy it, especially the flexibility of the system.

Mostly?I have been using the standard "Load" and "Empty" for my cars.? I have been experimenting with custom loads and schedules but only for single commodity cars (ie. coal hoppers,?Woodchip cars?etc).
Now I want to experiment with Custom Loads and Schedules for the rest of my cars.? Over the next few months I plan to add these little by little and run simulations to test them out.

I have made a Manual Backup of my latest operating session so that I can revert back if needed. (I might have an operating session?using the current setup)??More importantly, the Backup has a?record of where all my cars are on the layout right now After I finish running all the simulations I plan to just reload the "OperationsCarRoster.xml" backup file into the current xml files.? This way I won't have to manually restage all the cars on the layout.

My questions are:

1. Where are the Custom Loads stored in the program? Are they in the "OperationsCarRoster.xml" file?
2. Where are the Schedules stored in the program?? Are they with the "OperationsLocationRoster.xml" file?
3. Will I be stuck with having to manually re-stage all my cars after I am happy with the new setup?
4. Is there another way to achieve what I want to do??

I hate to do a lot of changes to?my setup?just to have it all undone by overwriting it with an older?backup.?

I look forward to hearing your suggestions.

-Scott-
(aardtechus)

James,

On 25 May 2019, at 12:50 PM, James Purdy <drjjpurdy@...> wrote:

I have a working JMRI progra on my layout with good signaling using CMRI hardware interface that I built many years a go and have updated to use a new SUIC on the mother board with DCCOD detectors. I have an operation led signal heads/masts and great panel of the layout developed on Panel Pro. The computer is an aging HP portable. I want to prevent a fatal computer crash that could wipe out my config/panel xml file/program. I would like to have an off site location - possibly a flash drive if I have to reinstall JMRI. Any suggestions?

Dropbox plays well with JMRI to give you an offsite copy. Many JMRI developers and users do this and we have written a help page on the topic. The principles would also apply to Google Drive, OneDrive and other similar, but Dropbox is probably the one most used with JMRI:
<>
This page may also be helpful:
<>

But Dropbox/Google Drive/OneDrive must not be your only backup. Things can go wrong. You will also need to regularly keep a copy of your User Files Location elsewhere, such as on an alternating pair of flash drives. Google search for "3-2-1 backup rule".

Dave in Australia

Bob and Jon,

On 25 May 2019, at 11:30 AM, Robert Schworm <rkschworm@...> wrote:

I have jmri connected serial to NCE Power Pro. The decoder is a tsunami2 - diesel for Alco-PA1. When I attempt to write all sheets on a programming track, jmri stalls out at 1.127 which is in the area of a feature called Digital Dynamic Exhaust. Anyone know why this is happening. When programming this with the NCE throttle, it states "WAIT" for several seconds and then completes. Jmri 14.

This particular problem has no connection with the known older Power Pro programming track circuitry deficiency.

Instead it relates to a Power Pro firmware issue that affects program track operations with both JMRI and the NCE throttle.

If you, by any means, attempt to write a value to any CV>256 in Program Track mode using an NCE Power Pro, the firmware will instead write to a CV<257, possibly corrupting your decoder.

For, example, if you use the NCE throttle to write a value to CV275, that value will be written to CV19 (the consist address) instead, making your decoder appear non-responsive to its normal address!

If you attempt use JMRI to write a value to CV275, JMRI intercepts that request (to protect your decoder) and instead throws an error (unfortunately not gracefully, but the error can be seen in the JMRI system console).

The DDE and other advanced panes in the TSU2 use CVs>256 so are affected.

The only workaround for these decoders is to read the decoder on the Program Track, then save and reopen in Program on Main mode. Attempting to use the NCE throttle as a workaround will simply corrupt other CVs in your decoder.

( The ESU decoders are not affected because ESU provides an in-decoder workaround that doesn't involve writing to CVs>256 so JMRI intercepts the write command and uses that workaround.)

If you have TSU2 decoders, your only choices are to use Program on Main or to get another programmer (such as a SPROG or an NCE Power Cab/NCE USB combination). JMRI can easily be configured to use two connections in the same session, with Program on Main, throttles, turnouts etc. being directed to the Power Pro and Program Track commands being directed to the SPROG/Power Cab.

Don't try other means to cheat around the firmware limitation in the Power Pro, it will inevitably lead to tears.

Dave in Australia

I have a working JMRI progra on my layout with good signaling using CMRI hardware interface that I built many years a go and have updated to use a new SUIC on the mother board with DCCOD detectors. I have an operation led signal heads/masts and great panel of the layout developed on Panel Pro. The computer is an aging HP portable. I want to prevent a fatal computer crash that could wipe out my config/panel xml file/program. I would like to have an off site location - possibly a flash drive if I have to reinstall JMRI. Any suggestions?

Bob,

Power Pro (PP) is presound decoders, PP has trouble programming them.

There is a mod that can be don to increase the current on the PP, involves two
light bulbs I believe.

Or you could get one of the programming boosters (PowerPax ot the Soundtraxx
unit).

Rumour has it NCE is working on a new command station to address this and other
things like USB, etc., I may be too old by the time it hits the market though.

John