PJ,

The daily build, I posted a link earlier today, has the fix for the baud
rate on CMRI issue. 4.15.8 through 4.17.2 have issues with selecting the
baud rate.

-Ken Cameron, Member JMRI Dev Team
www.jmri.org
www.fingerlakeslivesteamers.org
www.cnymod.com
www.syracusemodelrr.org

All;

To remove some of the confusion regarding the adjustable bracket cap, here is a direct link.



Dan

You can have the parts printed by a third party if you don't have a printer.? If I wasn't on vacation I'd print them for you.

You can also alter the print design to oblate the micro switch holes (or just hog them out with a file), although I believe there is enough free travel in the micro switches and limited travel in the servo throw arm that you won't need to worry.

Ken Moordigian
Jackson Livery
818-522-4292

Thank you muchly, Dan!

You have not only saved me from a few hours of design time, but you've even improved my design with micro switch mounts!

How thoughtful!

Ken Moordigian
Jackson Livery
818-522-4292

Hi PJ,
Sounds like the problem addressed here:?/g/jmriusers/message/160582
--SteveT

Jim,

You wrote:

** When i try to identify a loco on the main that has a roster entry in the pgm track profile, it was coming up with "No Roster Entry", but now I get a no loco found error.

Without special hardware, it is only possible to read NMRA CV values when on the "service mode programming track". You _cannot_ read a locomotive CV when using "programming on the main", unless you are:

- trying to read a Digitrax decoder with its "transponding" feature enabled, and
- the Loco is in a "Digitrax transponding-equipped" track section, and
- the transponding mechanisms are connected and configured in a way that actually works
- the transponding hardware is configured to, and capable of, commuincating with your JMRI computer

or you are:

- trying to read a RailCom-equipped decoder with its RailCom feature enabled, and
- the Loco is in a "RailCom-reader-equipped" track section, and
- the RailCom mechanisms are connected and configured in a way that actually works
- the RailCom hardware is configured to, and capable of, commuincating with your JMRI computer.

(There may be other system-specific solutions to this problem; these are the two that I am aware of...)

Regards,
Billybob

I have a free image that does exactly that at

开云体育

I reloaded 4.16. I also have a problem with it. I posed the problem.? ? ? ? ? ? ? Raccoon Valley Model Railroad Web Site: November Open House List: NJ, DE, PA and MD Layouts: ?

I am using JMRI 4.16 until the new 4.17.3 comes out. But when I try to use my iPad after starting the JMRI Server I have a problem. Everything comes up except the panel that are up on the computer. Under the OPEN WINDOWS it says "No Sharable Windows Are Open". If I click on the pull down menu "Panels" the panels are listed. I click on one of the panels and it goes to the panel page and says loading panel. But the panel never comes up. ? PJ

开云体育

There have been a number of threads talking about the different PanelPro panels. ?This is my view.

On 7/29/2019 10:19 AM, dan@... wrote:

Steve,
If you have access to a 3D printer this is the mount I use and i added the adjustable limit switch mount.

That's excellent. I pledged 3D printer on Kickstarter and when it ships early next year I need some files to try on it. This will be perfect.

--

Rick Jones

You have the right to remain silent. Anything you say will be
misquoted and used against you.

5000 lights is quite a few.

IIRC, each one will put about 7msec of traffic on the OpenLCB CAN cable. That means about 30 second of 100% busy traffic as JMRI starts up. OpenLCB is meant to handle that, but It’ll be interesting to see how well the JMRI implementation works!

Bob

On Jul 28, 2019, at 6:11 PM, steve young via Groups.Io <icklesteve@...> wrote:

Hi Tim,

To test how JMRI copes I've created a panel xml file with 5000 x OpenLCB Lights, each with 3 FastClock Light Controllers.
The hardware address increments on the final 2 byte pair on each side of the paired address.

/g/jmriusers/files/ProblemsBeingWorkedOn/Updated%20Lights%20-%20Testing/openlcb-5000-lights-3-fastclock-controllers.xml

Tested with JMRI 4.17.3ish OpenLCB CAN simulation on an i3 Win7 refurb laptop on its 4th hard drive.
Java max 76% CPU mid panel load, approx 0% - 6% CPU while running at 60x FastClock speed, left for an hour or so, no noticeable memory increment, no issues located.

While 20 CAN frames / sec isn't massive by any means, this was more of a test of the JMRI internals to ensure that it was coping well with adding that amount of Lights, initialising the fast clock listeners, time comparisons etc.
5,000 lights with 3 on times, 3 off times, 3 yesterday on times, 3 yesterday off times = 60,000 times checked per fast-clock minute change.

I loaded the panel in v 4.16 and things weren't happy, as expected.

I then tested with 4.17.1 , no errors in console log and ( simulated ) messages seen when expected in the CAN Frame console, with no obvious lag following the clock change, all immediate.

Without hardware to test with, no issues appearing in the simulation, and no intimate knowledge of OpenLCB protocol I may be the wrong person to assist,
although initially the issue was with the Light Control stuff, now that's sorted it may need someone better at OpenLCB to test with real hardware.

It would be worth knowing EXACTLY what console log messages you are getting with 4.17.2

Perhaps with real hardware connected, start by editing the test file down to 50 or so lights in case your console gets overloaded by 5000 ?
Note you need a minimum of 480 lights to get a light change EVERY fast clock minute, otherwise you'll get a gap, see the xml for the pattern.

--
Bob Jacobsen
rgj1927@...

Steve,
When you click on the Thingiverse link, about midway down the page below the thumbnail pictures is a horizontal menu item, to the far right there is a "Remixes" button, click on that? to take you to the modifications people have made for this item.

The limit switches do not monitor the servo position for servo control, one limit switch is for the live frog polarity change and the other for status confirmation and/or indicators.
Dan

PJ,

There was a message yesterday about a daily build that had a fix for the
CMRI baud rate issue. So far, those who have tried that found they can then
pick and use the baud rates they want.

See...
This will be in Packages 3393, building shortly at
It would be
good to get that tested by people who are still having the problem.

It will be part of 4.17.3 when that gets released next month. But this link
to the daily builds will get you a download that should cure this issue.

-Ken Cameron, Member JMRI Dev Team
www.jmri.org
www.fingerlakeslivesteamers.org
www.cnymod.com
www.syracusemodelrr.org

Having the same trouble with the new 4.17.2. This time even when I try to change the baud it still comes back as 19,200 instead od 9.600. I tried several times. No luck.

Don,

For what you are asking, I'd say setting up the laptop and Pi for remote
desktop is the simpler way to go. You would run JMRI in the normal display
mode and use the remote so the keyboard, mouse, and screen of the laptop
becomes your interface. Simple and direct, JMRI doesn't even know you aren't
on the console of the Pi.

-Ken Cameron, Member JMRI Dev Team
www.jmri.org
www.fingerlakeslivesteamers.org
www.cnymod.com
www.syracusemodelrr.org