?
Steve_G, I don't believe in anything "dumb luck". Genius.
?
AD, my "virtual" turnout occupancy block is not mired in Logix.
It looks complicated but really simple in either Logix or LogixNG:
- For turnout state closed and both blocks on that path are active then turnout occupied (set virtual block active).
- For turnout state thrown and both blocks on that path are active then turnout occupied (set virtual block active).
- Otherwise, turnout is unoccupied (set virtual block inactive).
Proper timing to satisfy Dispatcher is done in the debounce times of the blocks indicating occupancy:
- Blocks on either side of the turnout are set debounce active 500 ms, inactive 0 ms.
- Virtual block debounce active 0 ms and inactive 500 ms.
The nice part of this is that there is no need for additional detection hardware and the block occupancy doesn't disappear.
And it's as simple as my old pea brain can keep track of.
?
The caveat is that all rolling stock is detected.
It's as close to prototype operating safety as I can achieve.
?
I have one other standard design feature that really doesn't affect any of this.?
I buffer each hardware detector sensor with a virtual sensor set by Logix(NG) - detect active makes virtual active, otherwise inactive.
I do that so I can simulate occupancy with a panel toggle for testing purposes.
?
--
Ken
NYNH&H, Old Colony Division, Cape Cod Branch (1949-1959)
Loconet * JMRI 5.11.1 * OSX,Win10,Ubuntu
Blog: ?
Youtube:
