kvp's layout ideas and attempts

[kvp]

18 hours ago, nxcale said:

The control panel looks very good!

Could you make a video showing its operation or pictures indicating the function of the buttons/switches?

I'll try to shoot a video tomorrow. The red rotary switch is the signal control (departure/stop/arrival), then you select one of the tracks with the white buttons. It sets up the turnouts, the power routing and the signals. The white rotary switches select the power source (between the down main/automatic/up main and the 1st/auto/2nd-3rd), while the black rotaries are the turnouts (straight/automatic/diverging). Unless one wants to do shunting, the red rotary and the white buttons are enough. The 3 on off switches on the right are the main power and the two lighting groups.

[NXCALE]

 

On ‎1‎/‎11‎/‎2017 at 10:45 PM, kvp said:

I was thinking about how to wire up my still not quite ready ttrak terminus station (see above) and decided to try creating a somewhat japanese style control board.

 

The setup is the following:
-a crossover with an entry signal
-three tracks, with one island and one side platform
 (created by using a single turnout to branch off one track from the standard ttrak straight track position)
-three starter signals, one for each track

 

This means there are 6 possible routes, 3 arrival and 3 departure. The mode of operation, arrival/stop/departure should be selectable with a single 3 positon rotary switch (instead of using 4 two positon rotarys), while track selection is done by 3 pushbuttons. The resulting operation is that the station controller selects the direction and then one of the tracks. To cancel the route, the rotary is turned back to the middle to select stop.

 

The implementation (in theory) should have one route holding relay for each route, with the rotary used as an enable/release switch. The buttons should activate the route relays, while the rotary decides which one should be used out of the two assigned for each track, depending on the selected direction. The route holding relays should also cut power to the other two in the same direction and have a route selected output, so with the self holding, 4 circuit double throw relays are needed. The route selected outputs could directly drive the starter signals, while the entry signal could be driven through a diode matrix and an auxiliary relay, very much like the two turnouts, where two single circuit morse relays could drive the two bipolar capacitor discharge drives for the turnout motors. This would allow two turnout override rotarys to be added with center position as route select and left and right as straight and diverging, but this is not really required for such a simple setup.

 

Finally we have the problem of analog traction power. There are two power sources, the inner and outer track of ttrak, having one controller each. Arrival selects the inner, while departure selects the outer track (the buffers are on the right). In this setup, the crossover is isolated on all 4 ends and both loops fed with a single power source. (removing this Y feed splits it into two to serve as a normal crossover between two loops) The station tracks have two feed points, one for track 1 (outer loop) and a shared one for track 2 and 3 (ttrak inner loop), routed by the left handed turnout. The control logic must use the 6 route selected signals to take power from either the arrival or the departure line (2 two circuit morse relays) or none in stop state. This must be routed to track 1 for two routes and the common feed point of track 2 and 3 for the other four routes (another 2 two circuit morse relays), so a small diode matrix is required here too.

 

In overall, 1 three position rotary, 3 momentary pushbuttons and lots (6+3+4) of various relays are required to drive 2 turnouts (crossover, left handed), 4 signals (one entry, 3 starter) and 4 blocks (crossover and 3 station tracks). No ATS is used, trains must stop under driver control at the entry signal and before the buffer stops.

 

This is my idea so far and i'm still looking into making this a bit more simple, while keeping the 1 rotary, 3 pushbutton interface for ease of control. Also thinking about how to add shunting mode so trains could also leave wrong way towards the yard (a two state switch might be enough for this).

 

I have a question too: What kind of starter signals should i use? For entry, a 2 color signal seems ok or a 3 color one with only red and yellow in use (i have a few of this type). I'm thinking about both the type (shape, location, mounting) and the signal aspects that i should use for the starting signals.

 

On ‎28‎/‎04‎/‎2018 at 3:52 AM, kvp said:

I'll try to shoot a video tomorrow. The red rotary switch is the signal control (departure/stop/arrival), then you select one of the tracks with the white buttons. It sets up the turnouts, the power routing and the signals. The white rotary switches select the power source (between the down main/automatic/up main and the 1st/auto/2nd-3rd), while the black rotaries are the turnouts (straight/automatic/diverging). Unless one wants to do shunting, the red rotary and the white buttons are enough. The 3 on off switches on the right are the main power and the two lighting groups.

 

I have been reading previous pots, using route holding relays is a good idea.

 

What type of signal did you choose for the entry signal?

I would go for the 2 aspect signal (red and yellow) instead of a 3 aspect one.

Also I would include 2 junction indicators at the 1st (315 degrees) and 4th position (45 degrees) or a "theatre type indicator" that shows numbers "1", "2" or "3" or the way that diverging routes are shown in JPN signalling practice.

 

 

 

 

 

 

 

[kvp]

VID_20180428_155306.3gp

[NXCALE]

25 minutes ago, kvp said:

VID_20180428_155306.3gp

 

Wickedly nice!

 

Below a diagram that shows some signals with junction indicators (the diagonal line on top of the signal symbol). As the arriving signal doesnt have a green aspect, a theatre type indicator could be also used (all this stuff is BR practice). I understand that JPN signalling uses a second signal to indicate an existing divering route so the junction indicators could be replaced by a second signal (and a third one) as required.

 

Belown the same thing but including shunt signals.

I just noticed that one of the shunt signals could not be required (I think the one on the entry signal...)

Anyway, this is just an idea.

[kvp]

I'm not using route signalling, just speed signalling. This means a two color red/yellow entry signal and two color red/yellow for tracks 2 and 3 and red/green for track 1 exit signals. Shunting will happen towards the entry signal as the future depot will be accessible from the up line through a turnout normally ran through in the trailing direction.

Edited April 29, 2018 by kvp

[NXCALE]

That makes sense. It is gonna be interesting to see modules with speed signalling.

[kvp]

A small z scale idea:

 

 

 

 

[kvp]

Lillafüred State Forest Railway:

 

 

https://hu.m.wikipedia.org/wiki/Lillafüredi_Állami_Erdei_Vasút

 

[marknewton]

I like Mk48s! :)

 

Cheers,

 

Mark.

[kvp]

A bit of progress i managed to do:

a Zj scale mini diorama:

 

and two fremo spax boosters:

and two more for a fellow clubmember, but i forgot to take a picture...

 

[MichiK]

That's pretty much wood for a small layout - what is it's weight so far?

 

And then to electrics: There's the poti for speed controll, and a reverse switch, of course. But what are the other two switches for?

[kvp]

1 hour ago, MichiK said:

That's pretty much wood for a small layout - what is it's weight so far?

 

And then to electrics: There's the poti for speed controll, and a reverse switch, of course. But what are the other two switches for?

It's not too heavy as it's rather small, the idea is that the outer half could be flipped over to form a box for cover and transportation and the front right plate could come off to access the rolling stock storage area under the main board.

 

The controls are:

  -main power on/off

  -lights mode 1/off/lights mode 2

  -direction clockwise/stop/counter clockwise

  -speed control (it will be pwm based with a pic microcontroller and an uln2003a driver)

[MichiK]

Of course, the lid!

Way to often forgotten, but the dust cover should really be one of the first things on a layout..

 

And a main switch - while maybe not essential  - is not a bad idea, too!

[kvp]

The fremo spax boosters i made for my fellow club member. Unlike my own, these are properly boxed:

[kvp]

A small 7cm long Z scale bridge to fit under a piece of Maerklin track added between two Rokuhan ballasted pieces to cross over the back gap on the diorama mentioned earlier:

 

 

[gavino200]

On 2/29/2016 at 12:07 PM, kvp said:

A few detail shots of some my work in progress t-trak modules:

(made a day before the last meeting, some of the catenary was still missing)

       

 

I think i should offically name it Kōgai station (郊外駅), located somewhere in Japan... (what do you think of this name?)

 

 

 

How did you do the grass kv? Is that "Static grass"?

[kvp]

1 hour ago, gavino200 said:

 

How did you do the grass kv? Is that "Static grass"?

Yes, just a bit of white glue diluted with a bit of water and applied with a wide paintbrush, then sawdust and woodchips colored brown with a diluted acrylic paint for the ground. After dry, add another layer of diluted glue, then 2 mm static grass with a noch battery powered applicator. The 3rd layer could be anything on top of the grass like bushes and flowers. For the bushes a dab of white glue, for the flowers and tall grass spray adhesive (or hairspray).

[kvp]

Another day, another project idea. (at least this is somewhat already in progress)

 

A few months ago, i had an idea for a station control bus system, essentially a very simplified variant of the Siemens Simis IS system in use on the real railroads. The bus is very similar to loconet, just uses different voltage levels and has a different protocol. The wire layout is 12V DC, ground, 5.6V DC and bidirectional serial data. The command protocol is 9600,8n2 serial at 5V TTL levels, with (in my case) an Arduino acting as the central and pic12f508/12f510 based boards acting as i/o controllers. The message format is address7 | 128, data7, checksum7, so the 1st byte of the message has the msb bit set. Optional response from any addressed i/o controller is data7, checksum7 if the controller has a bidirectional capability. Output only controllers can share addresses and provide no presence detect or other info. There is an option for using extended address and data fields as long as the messages are backwards compatible.

 

The basic output only controller i designed has a pic12f508 microcontroller as the protocol decoder and an uln2003a as the low active output driver.

The pinout is the following:

-bus side: 12V, ground, 5.6V, serial data

-output side: 12V, output0, output1, output2, output3, output4

 

The board above allows (in theory) an up to 5 lights hungarian/japanese light signal to be controlled (power common, low active types only), at least 2 MTB low active turnout motors. Also usable for any low active lights or relays.

 

I plan to assemble and test a few asap as there are at least 2 hungarian fremo stations in my club that need a working interlocked signal/turnout control system by october and analog cabling using separate wires and the interlocking system construction with relay/diode logic would be too much work and probably cost way more. The control panels are planned to be wired directly to the Arduinos and only the 4 wire control bus going to the modules.

[kvp]

The output only driver circuit in my previous post assembed and being tested with a hungarian 4+1 lamp signal head imitation circuit:

The single side circuit boards are from China and come out pretty cheap, with around 1 usd per board (including shipping) and around 5 usd assembled with parts from the nearest electronics shop. This setup works with usb power only, but the leds are running only on 5V instead of the full 12V and currently powered through the Arduino, which is not really suitable for larger installs. (for that, a separate power supply board is needed that feeds both the bus and the Arduino using a 12V DC power supply)

[kvp]

Test setup with arduino on the left, power supply/bus interface next to it and two output only decoders on the right, one with a moc signal and one with a hungarian 4+1 lights signal (red,yellow,green,yellow+white shunting).

(yes, it's a bit of a mess on the table)

[kvp]

The left 12 is for my hungarian fremo station and the right 8 is for a fellow club member's hungarian fremo station. All assembled, programmed and tested. The labels are the bus addresses as these are fixed and can not be changed without removing the microcontrollers. (the 4 wire ends in the upper left corner are the bus connection used for testing)

[kvp]

The current source code for the output only decoders: (may have to get fixed in the future if i find more bugs)

 

Quote

; output driver Viktor Peter Kovacs (kvp in 2018)
; Licence:
; Creative Commons Attribution Non-Commercial Share Alike (CC-BY-NC-SA-4.0)

; notes:
; 4 Mhz, 1 mips, 9600 baud
; 1 bit = 104.166 inst ~ 104 inst
; 1/3 bit = 34.722 inst ~ 35 inst
;
; packet format:
;  [addr7 | 128], [data7], [chksum7]

; configuration
    include P12F508.inc
    __CONFIG (_CP_OFF & _WDT_OFF & _MCLRE_OFF & _IntRC_OSC)
    radix decimal

; --- defines ---

#define ADDRESS (0x42)

#define PIN_SERIAL (3)

; --- macros ---

nop2 macro
    goto $+1
    endm

nop4 macro
    goto $+1
    goto $+1
    endm
     
wait1_3bit macro
    local wait1_3bit_1
    movlw   10
    movwf   tmp
wait1_3bit_1
    decfsz  tmp, F
    goto    wait1_3bit_1
    nop
    endm
     
wait1bit macro
    local wait1bit_1
    movlw   31
    movwf   tmp
wait1bit_1
    decfsz  tmp, F
    goto    wait1bit_1
    nop2
    endm

; --- data ---
; 0-6 : i/o registers, 7 : scratch pad
tmp      equ 7
recv     equ 8
pkt0     equ 9
pkt1     equ 10
pkt2     equ 11

; --- code ---
    org 0x00
rst                     ; reset vector
    movwf   OSCCAL
    movlw   64+128
    option              ; enable gp2 output mode (+disable timer)
    clrw                
    movwf   GPIO        ; set output pins to low
    movlw   8           ; set pin 3 to input, 0,1,2,4,5 to output
    tris    GPIO        ; set output mode
    clrw
    movwf   GPIO
wait_serial_rdy
    btfss   GPIO, PIN_SERIAL
    goto    wait_serial_rdy
    nop4
    goto    main

recv_bit
    wait1bit
    clrc
    btfsc   GPIO, PIN_SERIAL
    setc
    rrf     recv, F
    retlw   0

recv_byte
    btfsc   GPIO, PIN_SERIAL
    goto    recv_byte
    wait1_3bit
    btfsc   GPIO, PIN_SERIAL
    goto    recv_byte
    clrf    recv
    call    recv_bit
    call    recv_bit
    call    recv_bit
    call    recv_bit
    call    recv_bit
    call    recv_bit
    call    recv_bit
    call    recv_bit
recv_byte_stop
    btfss   GPIO, PIN_SERIAL
    goto    recv_byte_stop
    retlw   0

main
    call    recv_byte
    ; check for start byte
    btfss   recv, 7
    goto    main
main_1
    movfw   recv
    movwf   pkt0
    ; check start bit and address
    xorlw   ADDRESS + 128
    skpz
    goto    main
    call    recv_byte
    ; check for start byte
    btfsc   recv, 7
    goto    main_1
    movfw   recv
    movwf   pkt1
    call    recv_byte
    ; check for start byte
    btfsc   recv, 7
    goto    main_1
    movfw   recv
    movwf   pkt2
    ; check checksum: (addr + data + chks) & 127 == 0
    movfw   pkt0
    addwf   pkt1, W
    addwf   pkt2, W
    andlw   127
    skpz
    goto    main
    ; move bits 01234xxx to 012x54xx
    movfw   pkt1
    andlw   1+2+4
    btfsc   pkt1, 3
    iorlw   32
    btfsc   pkt1, 4
    iorlw   16
    movwf   GPIO
    goto    main

    end

 

[kvp]

Just a quick and simple idea of a 3 platform, 6 track terminus:

The track system is Tomix Finetrack and the track ends are aligned, so the platforms could be easily lengthened with additional straight tracks. The number of turnout motors in this arrangement is only 6 and the number of selectable routes are 12. With full power routing, it's possible to have a separate arrival and departure track speed controller and no further wiring is needed to assign only one track to each and afaik it's not possible to assign the same track to both. Imho even conflicting routes are short free.

[sandiway]

17 hours ago, kvp said:

Just a quick and simple idea of a 3 platform, 6 track terminus:

 

Each curved track coming in can only access 3 out of the 6 terminuses?

[kvp]

12 minutes ago, sandiway said:

Each curved track coming in can only access 3 out of the 6 terminuses?

No, one is the arrival and the other is the departure track. Each track could be accessed by each direction. The first turnout for each direction decides between the top and the bottom half of the station. The double slips allow access either in diverging+diverging or crossover+crossover directions, while the rightmost two turnouts just select between the outer two tracks.

 

The possible routes for the top track, with tracks numbers 1 to 6 from top to bottom:

-track 1: straight, diverging, straight

-track 2: straight, diverging, diverging

-track 3: straight, crossover

-track 4: diverging, (crossing,) diverging

-track 5: diverging, (crossing,), crossover, diverging

-track 6: diverging, (crossing,), crossover, straight

 

The bottom track's access is the mirror of the top. The idea is that trains arrive on one track, reverse and leave on the other. The 3 island platforms serve all tracks, so no side platforms are required. The full power routing in the Tomix turnouts allows a very simple route and power selection system where only one station track is connected to one arrival/departure track at a time and the two controllers can't connect to the same station track at the same time, but are always connected to one station track each. Physically conflicting (crossing) routes could (should) be locked though and this happens when at least one of the leftmost turnouts are set to diverging, so in that case only arrival or departure should be allowed but not both as the trains would cross each other's path either on the middle crossing or through one of the double slips.