From: David D. <djd...@ma...> - 2008-07-03 11:14:53
|
On Jul 1, 2008, at 9:49 AM, Dick Bronson wrote: > David Duchamp wrote: >> >> On Jun 30, 2008, at 3:05 PM, Dick Bronson wrote: >> >>> Dave, >>> I think that all is required for APB is the 'traffic direction' >>> information like I generate with Logix in the fully operational CTC >>> panel. Any traffic direction 'active' could automatically put a >>> 'hold' on all opposing signals, or show them 'occupied' with an >>> added >>> sensor input to SSL. Granted, this does not 'tumble down' the >>> signals >>> as the prototype did. I have not yet gotten my head around how that >>> actually worked, or how rapid the changes were from one to the next. >>> Once a traffic direction has been 'set' it may only be 'cleared' by >>> all intermediate blocks (including both OS sections) becoming >>> unoccupied. >> >> That's consistent with what I'm thinking. The tumble down is >> automatic >> in what I'm proposing, but may not be prototypical. When the last >> train leaves the section, is when all blocks in the Section become >> unoccupied. This should only require two internal sensors--one for >> each direction through a Section. > Correct. There is just one sensor needed per direction. The follow on > signals returning to clear would be automatic because only the > opposing > signals are forced to stop. > > I suppose that there would be a hitch in the get along when the train > was leaving the single track because the siding would become > occupied at > the exit end while the turnout route to it was set. I guess the > solution > would be to only trigger the traffic direction sensor if the siding is > occupied with the turnout set for the siding, but not if the OS is > occupied first. (I can visualize this in Logix by including NOT OS > as a > conditional, but not a trigger.) The problem situation would be when a > train reversed out of a siding back into the single track, or a > train it > was meeting was going to enter the single track. That would say that > not > only would the single track and OS need to be cleared, but also the > route would need to change. > > I guess it is not quite as simple as it first appears to get it 100% > right. (so it works for both meeting trains and following trains) > > Dick :) I'm still working on ways to handle meeting trains and following trains, that's going well. (For meeting trains my current plan is for the "parking" siding to be a separate Section, with either two blocks or "stop now sensors" at each end.) You bring up reversing directions, which currently has me stumped. Also I've decided it's impossible to avoid certain gridlock conditions--which simplifies things considerably. Maybe I'll get a chance to talk with you about these questions in Anaheim. Dave > |