Locked
Re: Best or correct way to add signal masts
Rummy,
Ah, that shows a likely bug in the rule set for that signal. OR it shows
that the official rules would never have had that combination of signals. If
the ruleset doesn't find a match in the logic of 'the next signal is XXX,
I've been requested by logic to show YYY' it will put up a stop instead
since it doesn't know what to pick. Some of the signal rule sets have had
lots of use and some have not. Also it boils down to you might have tried a
combination (without knowing it) that nobody else had tried. Result is get a
stop signal.
With the right notes of what signal was connected to what would be needed
for an expect to read into the rules as programmed and see if there is a
bug. But you have a working solution now, so it is up to you. If you can
document the case clearly, best with a panel file showing the fault, then
somebody can look into it.
-Ken Cameron, Member JMRI Dev Team
www.jmri.org
www.fingerlakeslivesteamers.org
www.cnymod.com
www.syracusemodelrr.org
|
Locked
Re: Best or correct way to add signal masts
Steve,
You got the order pretty good. With the logic, I find I do a save of the
panel after the discover, then exit JMRI and restart, and load the file.
What I think is happening when the discover runs is the order of the logic
getting built and running, some parts aren't being run yet. On a panel load
at the start, everything gets started together and it settles down right.
The other thing I do most of the time is run the discover twice. The thought
was that first time through, it processes one part of the layout and doesn't
see what hasn't been created in the other part yet. So the second time
around insures everything is seen. But I have no logical proof this matters,
just habit.
Any track you want protected needs to be in a block. But not all blocks have
to be real. Here is an example. At one layout, we didn't detect the turnouts
as being part of a block. I used internal sensors (all set to INACTIVE) and
made them for the block of just the turnout. This let me place the masts
around the turnout. I'm trying to convince the owner that we should detect
the turnouts and use that to prevent changing the turnout while it is
occupied, but that might be something we wire up later.
-Ken Cameron, Member JMRI Dev Team
www.jmri.org
www.fingerlakeslivesteamers.org
www.cnymod.com
www.syracusemodelrr.org
|
Locked
Re: Best or correct way to add signal masts
I understand there may be differences due to a railroads interpretation and implementation of the AAR rules, but the position light dwarf signals indicated stop when the next signal indicated approach. ?I had expected them to at least indicate approach or slow, and probably clear. ?Anything less than a clear indication on the next signal resulted in a stop on the dwarf signals.
But again, thanks for the help!
Rummy
|
Locked
Re: Transferring Operations Files to New Laptop
Thanks very much to Dave Heap for helping me transfer ops files from my old to my new laptop - easy peezy when you get good advice!!
Peter
EVNGR
|
Locked
Re: ESU ECOS and JMRI Roster problem
Hi Peter,
Yes, there is a setting for that: go to Edit then Preferences, scroll down to "ECoS preferences" and click on that item. Then go to the roster tab. There you can change the default behaviour to suit your own preference.
Alain in Paris
|
Locked
Re: How do I obtain and apply the fix for "CMRI baud rates mismatching in 4.16 #7207"
#cmri
#arduino
Ryan,
At this time, you have to use the 4.17.3 test release to get the CMRI fix.
Dave Sand
toggle quoted message
Show quoted text
On Aug 25, 2019, at 11:50 PM, Ryan Blundon <ryansrailroad@...> wrote:
I have run in to the same issue as reported and fixed in issue #7207. Glad I wasn't the only one that ran in to it, but I did spend hours troubleshooting it. I see that there is a fix for the issue. Is there a release that it is planned to be fixed in? How do I apply the fix to my installation running 4.16? I have temporarily reverted my layout to 4.14, but have made many changes to my panels on 4.16 that I would like to take advantage of.
Thanks,
Ryan
|
Locked
How do I obtain and apply the fix for "CMRI baud rates mismatching in 4.16 #7207"
#cmri
#arduino
I have run in to the same issue as reported and fixed in issue #7207.? Glad I wasn't the only one that ran in to it, but I did spend hours troubleshooting it.? I see that there is a fix for the issue.? Is there a release that it is planned to be fixed in?? How do I apply the fix to my installation running 4.16?? I have temporarily reverted my layout to 4.14, but have made many changes to my panels on 4.16 that I would like to take advantage of.
Thanks,
Ryan
|
Locked
Re: Best or correct way to add signal masts
Steve,
Signal mast logic responds to changes to block occupancy, turnout positions and the aspect of the next mast (the pair¡¯s destination mast). Assuming that you an active layout connection or have made provisions to simulate layout activity, the SML for a new signal mast pair may need a little help to get started. When I do a full layout discover, I save the panel xml file and stop/start PanelPro and load the new panel xml file.
I am not sure what you mean by ¡°do turnouts need to be assigned to a block¡±. All track components between a pair of masts are assigned blocks. If you are asking ¡°does a turnout need its own block?¡±, my answer is yes. Others say not necessarily. Signaling projects tend to evolve over time. Since signaling is very dependent on the block design, wiring for full block capability during construction is easier than going back and changing it.
Dave Sand
toggle quoted message
Show quoted text
On Aug 25, 2019, at 10:28 PM, steve <steve@...> wrote:
Ken,
It would be hard to backup to an identical state. What I know is that removing the virtual mast icon from
the layout panel stopped the oscillation. Then adding the real mast made things work by going to the
mast logic panel for the lead mast, deleting the virtual mast from its list, then running the "discover" process.
When I removed the virtual mast icon a popup informed me it depended on and or several items it was associated
with depended on it,, but I couldn't guess what they were at this point. Perhaps one of them was the offender.
On a similar matter, it *seems* like I need to "exercise" the circuit after I build it and run discover. Specifically
the masts don't all change aspects as expected until I've run all the masts, turnouts, and block occupancy
sensors thru several cycles. Is this possible, or do you think some other issue is at play here?
I *think* I've observed this on several different test panels I've set up this past week...
Am I correct in believing the correct procedure is:
Create the panel, add all the track and turnouts.
Assign blocks to all the track areas needing them (do turnouts need to be assigned to a block?)
Set masts at the appropriate block boundaries.
Run the mast logic on each mast, setting all the fields (Use layout panel paths ..., everything except lock turnouts),
then run the "Discover" process.
Steve
On 08/25/2019 07:28 PM, Ken Cameron wrote:
Steve,
To oscillate, there would have to be some sort of loop between the signals.
Namely setting something on the triple head mast was causing something to
change at the far end signal, and that would cause the triple head to change
again. Usually turns out to be a sensor or turnout named wrong so it got
into the wrong part of the logic.
|
Locked
Re: JMRI overloads a spur
Thanks Dan. I did reread the aggressive mode explanation and see now the warning re:overloading. I think I switched to aggressive long ago because normal wasn't creating enough movement of cars. But I just switched back to normal and the builds look pretty good - and without overloading. Perhaps due to the fine tuning I've done over a couple of years. Anyway I will try normal mode for a while and hope to get good traffic.
However, I don't see a practical use of aggressive mode now. The suggestion is to build the trains in the "right" order (to avoid overloading), but there is no way to know beforehand which order will be "right". I usually build a set of four trains to start.
Anyway, thanks for the help to all of you.
Ray
|
Locked
Re: Best or correct way to add signal masts
Ken,
It would be hard to backup to an identical state.? What I know is that removing the virtual mast icon from
the layout panel stopped the oscillation.? Then adding the real mast made things work by going to the
mast logic panel for the lead mast, deleting the virtual mast from its list, then running the "discover" process.
When I removed the virtual mast icon a popup informed me it depended on and or several items it was associated
with depended on it,, but I couldn't guess what they were at this point.? Perhaps one of them was the offender.
On a similar matter, it *seems* like I need to "exercise" the circuit after I build it and run discover.? Specifically
the masts don't all change aspects as expected until I've run all the masts, turnouts, and block occupancy
sensors thru several cycles.? Is this possible, or do you think some other issue is at play here?
I *think* I've observed this on several different test panels I've set up this past week...
Am I correct in believing the correct procedure is:
? Create the panel, add all the track and turnouts.
? Assign blocks to all the track areas needing them (do turnouts need to be assigned to a block?)
? Set masts at the appropriate block boundaries.
? Run the mast logic on each mast, setting all the fields (Use layout panel paths ..., everything except lock turnouts),
? then run the "Discover" process.
Steve
toggle quoted message
Show quoted text
On 08/25/2019 07:28 PM, Ken Cameron wrote: Steve,
To oscillate, there would have to be some sort of loop between the signals.
Namely setting something on the triple head mast was causing something to
change at the far end signal, and that would cause the triple head to change
again. Usually turns out to be a sensor or turnout named wrong so it got
into the wrong part of the logic.
|
Locked
Re: Best or correct way to add signal masts
Steve,
To oscillate, there would have to be some sort of loop between the signals.
Namely setting something on the triple head mast was causing something to
change at the far end signal, and that would cause the triple head to change
again. Usually turns out to be a sensor or turnout named wrong so it got
into the wrong part of the logic.
-Ken Cameron, Member JMRI Dev Team
www.jmri.org
www.fingerlakeslivesteamers.org
www.cnymod.com
www.syracusemodelrr.org
|
Locked
Operations = Car Float Operation - restricted moves
I have a car float with an 8 car capacity.? My local trains have a 4 car limit so it takes two tricks to fully unload the car float.??
I also have a "float sails' job that takes the car float cars to location "off layout storage" and brings 8 new cars back.
The problem is I don't want the cars taken from the industries and spotted on the car float on the first trick to be sent out on the second trick.? I want to fill the float with 8 cars from industries and have it 'sail' to storage then bring back the next 8 cars.
One way I thought of doing this is have two trains "float job 1 and float job 2", each using the same route (the route includes all 11 of my local industries).? Then I would build and operate Float Job 1 but NOT TERMINATE it.? Then build Float Job 2 for the rest of the cars on the float.? This way the cars spotted on the float by Float Job 1 would not be available to build Float Job 2.? I would then terminate both jobs and build the "Float Sails" job with a full float going to storage.
Is there a simpler way to accomplish this?
Thanks
Mike Stewart
Oakley, CA
|
Locked
Quick, East and SAFE way to shutdown a HEADLESS Raspberry PI
#rpi
There is a SIMPLE way to gracefully shut down a Raspberry PI that will work just fine when using the RPI as a JMRI host.
All it needs is to connect 2 wires to the GPIO header('DuPont' connectors to the GPIO pins are OK, so soldering isn't required), and attach those to a SPST pushbutton.
It only requires adding one line to a file, and will shutdown the RPI with the equivalent of the "sudo shutdown -h now".
Here's the reference.......https://www.stderr.nl/Blog/Hardware/RaspberryPi/PowerButton.html
As stated, if the RPI is running, it will perform an orderly shutdown, and after it's been shutdown(to the a low power state), pressing the button again will re-start the OS. So, if you've shut down the RPI and then realized that you didn't want to, it can be re-started without a power recycle. Just make sure the activity on the RPI has stopped after pressing the button, which ususally takes about 20 seconds.
I've tested this on the model 3B and 3B+ and the Zero W, all running pre-buster Raspbian(Buster is needed for the RPI 4), but since the functionality is built-in the OS, it should also work on the RPI 4.
|
Locked
Re: LCD recommendations for JMRI on Raspberry Pi
#rpi
On youtube, check out drone bot workshop for programming these displays.
Bob s
|
Locked
Re: Best or correct way to add signal masts
Hi,
A stumbling block I encountered was that my Triple Searchlight mast would oscillate rapidly when paired with a virtual mast at
the end of a spur as suggested in a tutorial I watched.? I replaced the virtual with another Triple and it all started to work
as expected.? This is with AAR 1946 rules.? I have my openLCB gateway to the point that my DIY mast follows the JMRI icon
exactly!
Steve
toggle quoted message
Show quoted text
On 08/25/2019 03:45 PM, Dave Sand wrote: Rummy,
There are quite a few differences between the "Single head dwarf¡± mast type and the "Position light dwarf¡± mast type. That explains why one works and one does not.
I don¡¯t have access to a copy of the rule book for AAR 1946 so I don¡¯t know if the differences were intentional.
Making the mast types consistent is not difficult. JMRI provides a mechanism for a local modification of the rules. That approach prevents changes to the installed rules that other people may have relied on.
Dave Sand
On Aug 25, 2019, at 3:30 PM, rummy@... wrote:
Dave,
I was using the "AAR 1946" signal system, and the mast type that was not cooperating is the "Position light dwarf signal".
Right now I'm using the "Single head dwarf signal" in the staging yard, which are also in the "AAR 1946" signal system, and they are working great. Ultimately, I prefer to have dwarf position light signals.
It's really a great feeling to get this all working correctly. Thanks again guys!
Ruimmy
|
Locked
Re: Best or correct way to add signal masts
Rummy,
Now about track speeds. Turnouts default to a low speed for the reversed
route. You can change that in the turnout table. The logic will build
knowing this and 'some' signal system rules use this a lot and some ignore
it a lot. I suspect that was your problem with the original dwarf issues.
Granted, which type of mast you use on the panel will effect the rules it
uses to compute the aspect, but the physical can be anything, the panel
doesn't know.
-Ken Cameron, Member JMRI Dev Team
www.jmri.org
www.fingerlakeslivesteamers.org
www.cnymod.com
www.syracusemodelrr.org
|
Locked
Re: Best or correct way to add signal masts
Rummy,
There are quite a few differences between the "Single head dwarf¡± mast type and the "Position light dwarf¡± mast type. That explains why one works and one does not.
I don¡¯t have access to a copy of the rule book for AAR 1946 so I don¡¯t know if the differences were intentional.
Making the mast types consistent is not difficult. JMRI provides a mechanism for a local modification of the rules. That approach prevents changes to the installed rules that other people may have relied on.
Dave Sand
toggle quoted message
Show quoted text
On Aug 25, 2019, at 3:30 PM, rummy@... wrote:
Dave,
I was using the "AAR 1946" signal system, and the mast type that was not cooperating is the "Position light dwarf signal".
Right now I'm using the "Single head dwarf signal" in the staging yard, which are also in the "AAR 1946" signal system, and they are working great. Ultimately, I prefer to have dwarf position light signals.
It's really a great feeling to get this all working correctly. Thanks again guys!
Ruimmy
|
Locked
Re: Best or correct way to add signal masts
Dave,
I was using the "AAR 1946" signal system, and the mast type that was not cooperating is the "Position light dwarf signal". ?
Right now I'm using the "Single head dwarf signal" in the staging yard, which are also in the "AAR 1946" signal system, and they are working great. ?Ultimately, I prefer to have dwarf position light signals.
It's really a great feeling to get this all working correctly. ?Thanks again guys!
Ruimmy
|
Locked
Re: LCD recommendations for JMRI on Raspberry Pi
#rpi
I use the shutdown from the main Pi menu. I have run panels from it, for controlling our modular group's storage yards via a tablet, at present it is used for withrottle and has a configuration file that starts on start up with all our turnouts and storage yard routes so the throttles can see the turnout and route names. I have my roster installed so throttles can select those locos. Occasionally I have edited the locos and routes from the Pi using the screen, not as easy as with a larger screen.
Mike Ruby
|
Locked
Re: Best or correct way to add signal masts
Rummy,
Which signal system are you using? There might be an issue in the appearance xml file.
Dave Sand
toggle quoted message
Show quoted text
On Aug 25, 2019, at 2:43 PM, rummy@... wrote:
Wow!
I guess I had that "ah-ha" moment. I got all the signal masts added, the logic automatically created, and most importantly, the signals all work as I intended! And, just as importantly, I understand what is going on!
But, I did have an issue with the dwarf position light signal heads. I had high position light signals and dwarf position light signals in the staging yard. I could not get the dwarf position light signals to show the clear aspect no matter where they were placed. But, replace them with high position light signals, or dwarf single light signals and everything works great. Same signal mast logic in each case. And just to be sure, I changed those high position light and dwarf single light signals back to the dwarf position light signals and they still would not work right. Any ideas why?
Thanks to all, especially Dave and Ken, for having patience with me and providing some good information.
Rummy
|
Ken Cameron
alain c