Topics

DispatcherPro Implementation Problems

Nick Brownsberger
 

I recently began implementing DispatcherPro for the first time and I've run into a few problems. Can someone help?

i'm using 4.11. 6, iMac and Digitrax. Sections were created automatically from my signal mast logic.  A couple of transits were successfully created. I've also created two trains and have run each automatically by itself, i.e. one train running on the layout. Current panel file is in Files>Problems Being Worked On>Nick Brownsberger.

Problem 1. Both trains had this problem but only in a couple of sections. They ran through several section ok before problem occurred. The problem is, with 3 blocks allocated beyond the currently occupied block, the train stops. The dispatcher window says train status is RUNNING. System Console wrote this message:

018-06-11 06:26:48,469 dispatcher.AutoTrainsFrame            INFO  - No Throttle[4905/IZ4(Passenger Harrisburg (Trk 1) to Philly)] [AWT-EventQueue-0]
2018-06-11 06:26:48,516 dispatcher.AutoActiveTrain            WARN  - Stopping train [4905/IZ4(Passenger Harrisburg (Trk 1) to Philly)] in section [IY:AUTO:0017(Downingtown e Mast 1:Whitford M1)], as next section [Whitford M1:Paoli w Mast 1] is not allocated [AWT-EventQueue-0]
2018-06-11 06:27:19,302 jmri.Block                            WARN  - count of 2 ACTIVE neightbors with proper direction can't be handled for block Paoli 1 interlocking w but maybe it can be determined when another block becomes free [AWT-EventQueue-0]
2018-06-11 06:27:28,778 jmri.Block                            INFO  - Block Paoli 1 interlocking w gets LATE new value from Whitford 1, direction= East [AWT-EventQueue-0]
 
To get the train started again, I have to terminate and recreate it.
 
Problem 2. (I don’t think it is the same and problem 1.) Train stops one section shy of last section in transit. Train Status says WAITING.
 
I also have one gripe: the Dispatcher Window in all three sub-windows does’t let me expand Allocated or Requested Sections to see the whole name, even if I enlarge the bigger window.  Same problem with Active Train in Allocated Sections.
 
Thanks in advance, Nick

Roger Merritt
 

Nick,

     I've seen that WARN message before with the "next section not allocated".  Come to find out I was selecting only 2 blocks look ahead.  My train was longer than some blocks and overlapping them.  Software not fast enough to see this problem.  Increase the look ahead to 3 or 4.

Roger

Nick Brownsberger
 

Thanks, Roger.

I've got mine set at three and the blocks were allocated. But I don't see how a long train has an effect on this. Can you explain a little more?

Nick

Steve_G
 

On Mon, Jun 11, 2018 at 06:24 am, Nick Brownsberger wrote:
2018-06-11 06:26:48,516 dispatcher.AutoActiveTrain            WARN  - Stopping train [4905/IZ4(Passenger Harrisburg (Trk 1) to Philly)] in section [IY:AUTO:0017(Downingtown e Mast 1:Whitford M1)], as next section [Whitford M1:Paoli w Mast 1] is not allocated [AWT-EventQueue-0]
2018-06-11 06:27:19,302 jmri.Block                            WARN  - count of 2 ACTIVE neightbors with proper direction can't be handled for block Paoli 1 interlocking w but maybe it can be determined when another block becomes free [AWT-EventQueue-0]
2018-06-11 06:27:28,778 jmri.Block                            INFO  - Block Paoli 1 interlocking w gets LATE new value from Whitford 1, direction= East [AWT-EventQueue-0]
Hi Nick

The first error occurs when either the next section really isn't allocated, isn't allocated to the right train, or direction sensors are having a senior moment (as all your directions are consistent in the transit, that is highly unlikely)
The subsequent  errors are consistent with using internal sensors for occupancy, unless they are set up really carefully they can be triggered to set "Active", or "InActive" in the wrong order in relationship to adjoining real sensors. ie sensors A-B-C. A is real, it fires when occupied, C is real and fired when occupied. B gets fired by C firing, is flagged as late and, most likely due to a switch in the section, it cannot work out which train B belongs to, so everything stops. Or if A-B-C are all real, which is not your case, and the train doesnt have resistors for its full length. Then A goes active, B goes active, C goes active, B goes inactive, A goes inactive, B goes active and.... another senior moment. This can be tested by just running an AWD loco through the transit.

Steve G.

Nick Brownsberger
 

Steve, 

Thanks for helping.

I've redone my logix to make sure visual sensor go active in the right order, at least for the area of track I'm testing in.

Here's the current situation. Create a train to run in auto on track 4 which is a one way track. Start in block A (Philly B) with real BOD. Three sections beyond are allocated. Enters block B (52nd St. 4) with real BOD which triggers virtual BOD for block C (52nd St 4 extension) to go active. So 52-4 sets, then 52-4 extension sets. Then -see below- happens: (i've combined the Loconet Monitor output and the Sys Console output (indented) based on time stamps.) And the train stops.

I think I've controlled the sequence of blocks going active, but evidently not. Also, don't understand why Block 52nd St 4 cares about Philly 4 B since Philly 4 B is no longer occupied.

The last statement about LT105 is confusing. LT105 is a Tower Controller-ASD4 combination controlling a signal, one of only a few like this on the layout. Most signals are controlled by Signalmen.


 

06:20:40.454: Sensor LS2055 (BOD 52nd St 4) is High.  (BDL16 # 129, DS7; DS54/DS64 # 257, AuxD/A4).

06:20:41.190: Sensor LS2067 (BOD Philly 4 B) is Low.  (BDL16 # 130, DS3; DS54/DS64 # 259, AuxB/A2).



SYSTEM MONITOR: 2018-06-14 06:20:41,460 jmri.Block                            WARN  - count of 2 ACTIVE neightbors with proper direction can't be handled for block 52nd St 4 but maybe it can be determined when another block becomes free [AWT-EventQueue-0]


06:20:42.278: Send packet immediate: 3 bytes, repeat count 2(50)

DHI=0x01, IM1=0x14, IM2=0x75, IM3=0x03, IM4=0x00, IM5=0x00

packet: 94 75 03 .

06:20:42.279: LONG_ACK: the Send IMM Packet command was accepted.

SYSTEM MONITOR: 2018-06-14 06:20:43,193 jmri.Block                            INFO  - Block 52nd St 4 gets LATE new value from Philly 4 B, direction= North, West [AWT-EventQueue-0]

06:20:44.007: Send packet immediate: 3 bytes, repeat count 2(50)

DHI=0x01, IM1=0x16, IM2=0x75, IM3=0x00, IM4=0x00, IM5=0x00

packet: 96 75 00 .

06:20:44.007: LONG_ACK: the Send IMM Packet command was accepted.

06:20:44.007: Send packet immediate: 3 bytes, repeat count 2(50)

DHI=0x01, IM1=0x0F, IM2=0x77, IM3=0x15, IM4=0x00, IM5=0x00

packet: 8F 77 15 .

06:20:44.007: LONG_ACK: the Send IMM Packet command was accepted.

06:20:44.262: Send packet immediate: 3 bytes, repeat count 2(50)

DHI=0x01, IM1=0x05, IM2=0x71, IM3=0x03, IM4=0x00, IM5=0x00

packet: 85 71 03 .

06:20:44.262: LONG_ACK: the Send IMM Packet command was accepted.

06:20:44.262: Set speed of loco in slot 12 to 33.

06:20:44.507: Send packet immediate: 3 bytes, repeat count 2(50)

DHI=0x01, IM1=0x0F, IM2=0x77, IM3=0x15, IM4=0x00, IM5=0x00

packet: 8F 77 15 .

06:20:44.507: LONG_ACK: the Send IMM Packet command was accepted.

06:20:45.024: Send packet immediate: 3 bytes, repeat count 2(50)

DHI=0x01, IM1=0x05, IM2=0x71, IM3=0x1D, IM4=0x00, IM5=0x00

packet: 85 71 1D .

06:20:45.024: LONG_ACK: the Send IMM Packet command was accepted.

06:20:45.025: Set speed of loco in slot 12 to 77. (SPEED IN THIS BLOCK IS NORMAL. PREVIOUS BLOCK WAS MEDIUM.)

06:21:00.579: Sensor LS2058 (BOD 59th St 4) is High.  (BDL16 # 129, DS10; DS54/DS64 # 258, SwiA/S1).

06:21:00.962: Sensor LS2055 (BOD 52nd St 4) is Low.  (BDL16 # 129, DS7; DS54/DS64 # 257, AuxD/A4).

06:21:02.414: Send packet immediate: 3 bytes, repeat count 2(50)

DHI=0x01, IM1=0x0A, IM2=0x71, IM3=0x03, IM4=0x00, IM5=0x00

packet: 8A 71 03 .

06:21:02.414: LONG_ACK: the Send IMM Packet command was accepted.

06:21:02.414: Send packet immediate: 3 bytes, repeat count 2(50)

DHI=0x01, IM1=0x05, IM2=0x71, IM3=0x03, IM4=0x00, IM5=0x00

packet: 85 71 03 .

06:21:02.414: LONG_ACK: the Send IMM Packet command was accepted.

06:21:02.414: Send packet immediate: 3 bytes, repeat count 2(50)

DHI=0x01, IM1=0x05, IM2=0x73, IM3=0x03, IM4=0x00, IM5=0x00

packet: 85 73 03 .

06:21:02.414: LONG_ACK: the Send IMM Packet command was accepted.

06:21:02.459: Set speed of loco in slot 12 to 0.

06:21:03.651: Send packet immediate: 3 bytes, repeat count 2(50)

DHI=0x01, IM1=0x05, IM2=0x71, IM3=0x03, IM4=0x00, IM5=0x00

packet: 85 71 03 .

06:21:03.651: LONG_ACK: the Send IMM Packet command was accepted.

06:21:03.781: Requesting Switch at LT105 (Signal Whitford 4a) to Closed (Output On).

Nick


 

Nick Brownsberger
 

P.S.- that's "virtual sensors", not "visual sensor".

And, I wanted to mention that when train stops, 6 sections are now allocated even though I have 3 selected in Dispatcher options. Is this a problem?

Nick

Roger Merritt
 

Nick,

     Are you still having problems with "next section not allocated"?   I see you mention seeing 6 sections being allocated.  How many sections are being occupied at that time?  I have 3 sections look ahead set and see on start up 4 sections being allocated with sometimes 5 being allocated as train goes around.  I had problems before with issue you are seeing and solved problem by combining the short blocks into large sections.  This proved helpful while running longer trains.

Roger

Nick Brownsberger
 

Roger,

Yes, but I thought I'd focus on just one problem at a time.  Each section has several blocks, typically 3 or 4, and all of them can be occupied at a time.

When my test loco (12" long) ran beyond the blocks discussed above, it ran through one section and then stopped in the next one with this complaint. The section it stopped in has only one block and the section after that (the one that wasn't allocated, even though Dispatcher said it was) also has only one block. When it stopped 6 sections were allocated. When it had started, Dispatcher allocated, correctly, 3 sections.

Nick

Roger Merritt
 

Nick,

    Are you seeing same problem with short trains?

Roger

Nick Brownsberger
 

Yes. Just running engine.