Locked SML signal mast problem #sml
|
File moved to "Problems being worked on"
/g/jmriusers/files/ProblemsBeingWorkedOn/Ken%20Fordham -- Peter Ulvestad JMRI Users Group Moderator ( ) Tam Valley Group Moderator ( ) Sprog-DCC Group Moderator ( ) Edmonton Model Railroad Association ( ) |
|
Ken, Your last upload was good. ?I have some initial comments. LT1 is a left hand turnout, but you want the diverging leg to be the normal route. ?In Layout Editor this is done by using the Continuing Route Turnout State option. By setting the option to Thrown, Layout Editor will show the diverging leg as the Closed position. ? The Turnout table and related logic do not have left-hand or right-hand concepts. ?They deal with Closed and Thrown, whatever that means. ?My rule is that switch machines are wired so that the JMRI Closed state matches the expected physical turnout state. ? You appear to have two groups of occupancy sensors, a low number range and a 400 number range. ?What are the differences? Signal masts LT01(B) and LT01(C) are reversed on the panel. ?They are attached to block boundaries defined by the turnout legs, not the logical position as above. The LT25 signal mast icons are not on the panel but they are attached to the block boundaries. ?This can be confusing. ?This is easily fixed by selecting Set Signal Masts for LT25 and selecting Right Hand Side for each mast and clicking on Done. A lot of the signal mast logic is incorrect. ?I suspect this is causing the lack of signal response as a locomotive runs around the layout. Whenever the block topology has been changed, the affected signal masts have to have the SML updated. ?Sometimes a small changed requires updates to the SML for several signal masts in either direction. At this point I recommend running the GrandReset.py script. ?The script deletes the transits, sections, SML and block paths. ?After doing a store and quit, restart PanelPro and load the updated file. ?The first step is to go to the Signal Mast Logic table and do the Auto Generate Signaling Pairs. ? At this point, I would not bother with Sections and Transits. ?The first goal is to get consistent signaling based on train movements. Dave Sand ----- Original message ----- From: Ken Fordham <j.kenneth.fordman@...> Subject: Re: [jmriusers] SML signal mast problem #sml Date: Thursday, September 26, 2024 6:24 PM Hi Dave ? There are 3 files on problems being worked on. All are not good? ? I will send another file tonight or tomorrow morning. ? Sorry for all this confusion.? ? Best KenF |
|
Ken, The second try worked. There are a few more changes. I recommend that the turnout speeds be adjusted. ?The default speed for Thrown is Restricted. ?I normally used Medium. This is set using the "Speeds -> Defaults" menu item in the Turnout table. This affects the signal mast logic when routes include thrown turnouts. The "LT1" turnout table entry currently has "Inverted" enabled. ?I suspect that this was an attempt to make the diverging leg of LT1 to be the normal route. ?I don't use inverted. ?My rule is that if the JMRI Closed state is not the preferred route, the switch machine wiring is changed. The image for LT1 has the wrong color (red). ?Right click LT1 and select Edit. ?Click on the "Create/Edit" block button. ?Set "Occupied Sense" to Active. If the problem is with what LS57 is reporting, fix it instead of masking the problem using JMRI.? The signal mast icons for LT01(B) and LT01(C) need to be flipped. LT14 signal masts: ?LT14(C) is missing and LT14(B) is in the wrong location on the panel. ?Enable "Right Hand Side" for all 3 locations and click on Done. ?This will add/move the masts to the proper locations. LT22 signal masts: ?Select LT22(C) for the 815 - 512 block boundary. ?Create SML for LT22(C). ?Delete and create SML for LT25(A). LT23 signal masts: ?Add the masts to the panel. ------------------------------------ The LT26/LT14 and LT23/LT27 turnout pairs look like crossovers. ?If they behave as crossovers then trains will never stop between the turnouts. ?If that is true, then blocks 833 and 816 are not needed. ?And signal masts LT26(C) and LT14(B) are not needed. ------------------------------------ There is an error where occupancy sensor LS400 assigned to blocks 804 and 810. ------------------------------------ Note: ?The middle loop ( P14, etc.) has an interesting problem that will require more research. ?When the turnouts are lined for the loop, a SML loop occurs. ?When a signal mast aspect changes, the previous mast has to determine if its aspect needs to change. ?If so, it changes and when triggers the next previous mast. ?The method for propagating changes can go forever. I think the problem is the number of signal masts in the loop. ?The outer loop does not have the problem. Dave Sand ----- Original message ----- From: Ken Fordham <j.kenneth.fordman@...> Subject: Re: [jmriusers] SML signal mast problem #sml Date: Tuesday, October 01, 2024 6:35 AM Good morning Let me try again. I made all the changes you recommended so I don't know what happened. |
Peter Ulvestad