¿ªÔÆÌåÓý

Date

Locked Re: 4.12 JMRI Web Server Crash on Startup

 

Wouter,

The reason for the reference to Eclipse in the error log is due to the fact that we use Jetty as one of the web server components, which happens to be an Eclipse foundation project (hence the use of their namespace).

Pierre,

The error you're seeing is indicative of old versions of the jetty files still being accessible in your JMRI installation, in the 'lib' folder.

Jetty .jar files contain the version details in their filename meaning that on Linux installations, just copying directly over a pre-existing JMRI installation runs the risk of older versions of these files still being accessible.

As Steve mentioned earlier, the easiest fix for this is to re-extract the new JMRI installation into a completely fresh folder, either by deleting or renaming the existing one.

Best regards,

Matt H


Locked Re: Exception while setting Signal Logic attempting to set invalid aspect

 

Greg,

Based on my experience, double cross-overs are most likely to suffer from the backward looking logic. This can also occur when there are two cross-overs using the same blocks. This is effectively a double cross-over.

When I run into one of these, I delete the bad entries and manually create the missing entries. To do this, turn off the Use Layout Editor options and enter the blocks and turnouts in the manual section.


Dave Sand

On Nov 26, 2018, at 9:42 PM, Greg McCartney via Groups.Io <gregmcty@...> wrote:

I am uploading a fixed NW signal rules set to the files section and to Github. This should fix the icon mapping errors as well as a few mapping errors I found.

I took a look at Jack's panel. There are some problems with the way the signals are setup. First off, by NW rules, you can't get a Diverging Approach on a two head signal except for the Wabash signals with a top lunar marker. I had neglected to remove those mappings from that signal set so JMRI was looking for Diverging Approach and causing him to have an error. The second issue he has with the panel is the crossovers are not allowing signal mast logic discovery to function correctly. Sometimes the mast logic chooses a signal that is behind the source or facing in the opposite direction.

/g/jmriusers/photo/79332/0?p=Name,,,20,1,0,0
In this photo the signal that is circled is mapping to a signal in the opposite direction. Discovery ignores the two signals on the right end of the diagram which should be the destination signals.

/g/jmriusers/photo/79332/1?p=Name,,,20,1,0,0
In this shot, discovery finds three destination signals one of which is facing the opposite direction.

I'm not certain how to fix the discovery issue other than delete them. Perhaps Dave Sands or one of the others can help with that.

Greg McCartney


Locked Re: Exception while setting Signal Logic attempting to set invalid aspect

 

I am uploading a fixed NW signal rules set to the files section and to Github.? This should fix the icon mapping errors as well as a few mapping errors I found.??

I took a look at Jack's panel.? There are some problems with the way the signals are setup. First off, by NW rules, you can't get a Diverging Approach on a two head signal except for the Wabash signals with a top lunar marker.? I had neglected to remove those mappings from that signal set so JMRI was looking for Diverging Approach and causing him to have an error.? The second issue he has with the panel is the crossovers are not allowing signal mast logic discovery to function correctly.? Sometimes the mast logic chooses a signal that is behind the source or facing in the opposite direction.

/g/jmriusers/photo/79332/0?p=Name,,,20,1,0,0
In this photo the signal that is circled is mapping to a signal in the opposite direction.? Discovery ignores the two signals on the right end of the diagram which should be the destination signals.

/g/jmriusers/photo/79332/1?p=Name,,,20,1,0,0
In this shot, discovery finds three destination signals one of which is facing the opposite direction.??

I'm not certain how to fix the discovery issue other than delete them.? Perhaps Dave Sands or one of the others can help with that.

Greg McCartney


Locked Re: Reading and programming new BLI PRR T1 - PROBLEM

 

¿ªÔÆÌåÓý

Check past emails on the groups websight.
Issue is well discussed with multiple solutions
Need to increase current being used.

Tony


On Nov 26, 2018, at 10:19 PM, Michael Hahn <hahn@...> wrote:

I just got the new BLI PRR T1 steamer with Paragon3 "Rolling Thunder" sound decoder installed.? I am using a standalone programming track, with the latest JMRI software on a PC running Windows 10, and an 18 V powersupply.? JMRI will not detect the decoder.? If I manually select Paragon3 Steam, JMRI will not read the decoder, nor can I program the address.? I finally selected NMRA Standard definition, and then the address can be programmed.? Any ideas on what is wrong here?? And how to fix it??

I have cleaned the track and the loco wheels.? The loco does chatter when JMRA tries to read the decoder, but I get the error "No response from decoder" and no programming is possible.

Thanks for any advice or suggestions!

Michael


Locked New file uploaded to [email protected]

[email protected] Notification
 

Hello,

This email message is a notification to let you know that a file has been uploaded to the Files area of the [email protected] group.

File: NW1981 Signal Rules.pdf

Uploaded By: Greg McCartney

Description:
NW 1981 signal rules.

You can access this file at the URL:
/g/jmriusers/files/NW-1981/NW1981%20Signal%20Rules.pdf

Cheers,
The Groups.io Team


Locked New file uploaded to [email protected]

[email protected] Notification
 

Hello,

This email message is a notification to let you know that a file has been uploaded to the Files area of the [email protected] group.

File: NW-1981icons.zip

Uploaded By: Greg McCartney

Description:
Updated icons to go with the updated xml files.

You can access this file at the URL:
/g/jmriusers/files/NW-1981/NW-1981icons.zip

Cheers,
The Groups.io Team


Locked New file uploaded to [email protected]

[email protected] Notification
 

Hello,

This email message is a notification to let you know that a file has been uploaded to the Files area of the [email protected] group.

File: NW1981 updatedxml.zip

Uploaded By: Greg McCartney

Description:
Updated XML files for the NW signal system.

You can access this file at the URL:
/g/jmriusers/files/NW-1981/NW1981%20updatedxml.zip

Cheers,
The Groups.io Team


Locked New file uploaded to [email protected]

[email protected] Notification
 

Hello,

This email message is a notification to let you know that a file has been uploaded to the Files area of the [email protected] group.

File: NW1981 Signal Rules.pdf

Uploaded By: Greg McCartney

Description:
Scan of the NW signal rules.

You can access this file at the URL:
/g/jmriusers/files/NW1981%20Signal%20Rules.pdf

Cheers,
The Groups.io Team


Locked Reading and programming new BLI PRR T1 - PROBLEM

 

I just got the new BLI PRR T1 steamer with Paragon3 "Rolling Thunder" sound decoder installed.? I am using a standalone programming track, with the latest JMRI software on a PC running Windows 10, and an 18 V powersupply.? JMRI will not detect the decoder.? If I manually select Paragon3 Steam, JMRI will not read the decoder, nor can I program the address.? I finally selected NMRA Standard definition, and then the address can be programmed.? Any ideas on what is wrong here?? And how to fix it??

I have cleaned the track and the loco wheels.? The loco does chatter when JMRA tries to read the decoder, but I get the error "No response from decoder" and no programming is possible.

Thanks for any advice or suggestions!

Michael


Locked Re: Entering New Loco's in DEcoder Pro

 

Peter and Mike:

I am having exactly the same problem with reading and programming the BLI F3/F7 units with Paragon3 "Rolling Thunder" sound decoders in them.? I am using a standalone PR3 with a programming track, and an 18 V power supply.? I get exactly the same behavior as Mike described above, in all aspects.? I eventually used NMRA Standard Definition for these locos in order to be able to change their addresses.? Not a satisfactory solution, but at least I can program the address.? Other BLI locos with the Paragon3 "Rolling Thunder" decoder read and program just fine.? What gives???? Anyone have any suggestions??

Thanks!

Michael

P.S.? The new BLI T1 steamer behaves exactly the same way!!


Locked Re: NCE Packet Analyzer Crashes

 

¿ªÔÆÌåÓý


On 27 Nov 2018, at 9:58 AM, jimalbanowski <jimalbanowski@...> wrote:

I've had both of my PA's quit reading ControlBus that's the smaller RJ with two wires to ControlBus and switch 1 change. I power the PA from and external power supply.

According to the PA manual, DIP Switch 4 is the only one that does anything.

Don't even think of closing JP1 and JP2. That will break the optoisolation as it tries to power the PA from the Track DCC power at pins 3 & 4 of the 4 way socket. This will likely create havoc (magic smoke escaping) unless your computer is a fully floating laptop with no mains or other connection (e.g. to Power Pro serial port) and you have nothing connected to the 6 way socket. These old devices are too precious to risk destroying.

Powering the PA via pins 2 & 5 of the 6 way socket by plugging it into the NCE cab bus is safe and reliable.


Before they quit I did use JMRI to read the data... and on both the CabBus and ControlBus with no crash on either a JMRI Raspberry Pi or Mac... Use to with old PC years ago...

If you have a terminal program on your PC you can try that to see if JMRI/Java is the culprit rather than just pumping the data into the PC..?

Mine is working fine, been dumping data flat out for an hour or more with no crash. I've even done multiple "Freeze Screen" and "Info/Resume" to try to trip a buffer overflow, but not fazed.

BSODs are very unlikely to be caused by JMRI/Java. Very likely to be caused by USB-Serial driver.

For a non-JMRI option, I frequently use CoolTerm on Mac and there is a Windows version:

If using the non-JMRI option, call up the help screen by pressing "?"

From the PA Manual:?

Data Communications

We¡¯ve had good luck using Hyperterminal supplied with Win98 and Win XP.

The communications settings are 38.4kbps, 7 data bits, 1 stop bit. Use the ¡®No Flow Control¡¯ option. The analyzer does not echo characters so you may also want to turn echo on. Also turn on ¡®Append line feeds to incoming line ends¡¯ under the Properties->Settings->ASCII Setup menu.

Before connecting the analyzer to the track or command station verify communications by pressing ¡®?¡± to get a menu of commands and the analyzer software version.


Is the port on the PC a "classic" RS232 or are you using a USB dongle?

If a dongle, what make and model.

Dave in Australia


Locked Re: Trottle for ZIMO?

 

Assuming you have the right hardware (MXULF), you would then startup JMRI "WiThrottle"i throttle server.?
An App on Andriod or Apple phones can then connect as throttles.
See here:
http://www.jmri.org/help/en/package/jmri/jmrit/withrottle/UserInterface.shtml


Locked Re: NCE Packet Analyzer Crashes

 

Jim:

I've had both of my PA's quit reading ControlBus that's the smaller RJ with two wires to ControlBus and switch 1 change. I power the PA from and external power supply.

Before they quit I did use JMRI to read the data... and on both the CabBus and ControlBus with no crash on either a JMRI Raspberry Pi or Mac... Use to with old PC years ago...

If you have a terminal program on your PC you can try that to see if JMRI/Java is the culprit rather than just pumping the data into the PC..?

Is the port on the PC a "classic" RS232 or are you using a USB dongle?

Jim Albanowski


Locked Re: Cabooses

 

Dennis and Dan - thanks for the ideas. I would never have thought of that, and it will serve the purpose.

Jamie


Locked Re: NCE Packet Analyzer Crashes

 

The NCE Cab bus is monitored by setting DIP Switch 4 on before power up. The output isn't terribly informative as it consists mainly of cab pings from the command station, LCD display quadrant updates and the occasional key press. I just tried with JMRI V4.13.5 and it's been spewing out data like "80 81 82 95 96 AA" for many minutes now with no signs of a crash.

The NCE Control bus (a low current version of the DCC track bus - the booster input signal) or even jumpers to the track via pins 2 & 3 of the small 4-way Jack is monitored by setting DIP Switch 4 off before power up. You can then use the radio buttons to change the display format between decoded user-friendly output (Verbose) to various hex formats of the actual DCC packets. You need to use the "Show xx" radio buttons to show various packet types otherwise display will be blank. The "+ on" checkbox suppresses repeated packets, replacing them with "+" symbols. The "Info" button interrupts normal communication to ask the analyser for a comfort command help screen (JMRI is using those commands when you click on radio buttons).

Your BSOD has to be a USB-Serial cable/driver issue. What brand are you using?

Make sure you try JMRI 4.13.5 - just in case... I avoid V4.12 as it has many issues.
--
Dave in Australia

On 27 Nov 2018, at 8:55 AM, James Muthig <jmuthig3@...> wrote:

Hi Dave and Jim A., thank you for your replies. I am using JMRI version 4.12. I'm not using any app to capture data, just using the NCE Packet Analyser's serial monitor in JMRI to view the output of the CabBus. After 10 or so seconds the blue screen of death.... I've never tried to monitor to the control bus. Any advice for doing this?


Locked Re: reading LocoID help

 

Norb:

Decoders on the railroad can be read as on a programming track... However...

This isn't easy without some other tech... Digitrax has it's transponding and in Europe there is RailComm where decoders can transmit data. Neither supported by NCE.

An alternative is an RFID tag and reader to identify a locomotive. One source is...



This is a Raspberry Pi and Arduino system.

MRH ezine with their NMRA convention coverage mentioned a company TrainTraxx that has a different system that's cloud based with only a promised connection to JMRI.

If you want to go to a DIY search on Arduino and RFID for a lot of "stuff..."

I've built several RFID projects but none interface to JMRI.

Jim Albanowski


Locked Re: NCE Packet Analyzer Crashes

 

Hi Dave and Jim A.,? thank you for your replies.? ?I am using JMRI version 4.12.? I'm not using any app to capture data, just using the NCE Packet Analyser's serial monitor in JMRI to view the output of the CabBus.? ?After 10 or so seconds the blue screen of death....? ? ?I've never tried to monitor to the control bus.? Any advice for doing this???

Thanks,?
Jim M.??

On Sat, Nov 24, 2018 at 2:07 PM Dave Heap <dgheap@...> wrote:
I had assumed Jim M. (the original poster) was using the NCE Packet Analyser monitor in JMRI, but he didn't say which version of JMRI.

My NCE Packet Analyser is an essential tool that I use frequently and I don't recall ever having had a crash while actually using it.
--
Dave in Australia


> On 24 Nov 2018, at 10:20 PM, jimalbanowski <jimalbanowski@...> wrote:
>
> Also... what app are you using to try to capture the PA output into?
>
> If looking at CabBus that can be a torrent of data and you might be overflowing your apps buffers. If not well behaved that might be the cause a system lockup... Blue screen of death or recoverable with a three finger salute?




Locked Re: reading LocoID help

 

Sorry left out a few additional vitals.? Running JMRI 4.12 on Windows 10 PC with Java 1.8.0_191


Locked Re: Trottle for ZIMO?

 

Here's the Zimo hardware page:


Michael Mosher
Member SFRH&MS
DCC Master PVSMR

On 11/26/2018 4:28 AM, Titoffe wrote:
we use a ZIMO digital control panel and the STP software to manage our rail network.
we are sorely lacking trottles (10 members and 3 trottles !)
can we use the JRMI system to install the use of smartphones?
if so how?


Locked reading LocoID help

 

NCE layout is up and running.? Using NCE SB5 USB interface v7.3.2. I've written a logix to sense the end of track in a 8 track passenger station.? Once the reed switch is sensed, the logix script rings the engine's stop bell and stops the train just before it reaches the end of the track.? Works great. For testing, the Jython script, executed as part of the logix, has a hard coded short address for one of the test locomotives (stored as LocoID).? I want to now generalize the script and combine it with the block sensing script to do three things: get the address of the loco that has entered the track, send a throttle slow down command to slow own the train to about 15% throttle and pass the engine short address via memory variable to the logix script, so that it will stop whatever engine enters the track. Each of the station tracks is a separate block with a separate NCE BD20 sensor that reports the engine entering the block (track), but I can't determine how get the loco address that entered the block. This should be pretty straight forward, but look through some example scripts for ideas.? Can anyone provide some help?

Thx Norb

?