Nice video Gary and thanks for the encouraging comments Steve.
Thanks for your comments Don, I'll try and make a few comments on you said.
don9GLC said:
But perhaps I am being a bit ambitious when I aspire to trains of more than one wagon. For realistic operation I think you need to know where the rear of the train is, and this naturally varies with the number of wagons.
Yes I agree that where the end of the train is important information to know. The software I use does potentially know this since one of the attributes that needs to be configured about each piece of rolling stock and loco is their length. The software keeps track of each train set length, and adjusts the length when uncoupling and coupling takes place. I know this works because the software uses it to ensure that train sets only get routed to blocks that are big enough, if the train set is to stop in the block. I've not yet worked out if this information can be accessed and used during coupling and uncoupling operations. I will investigate more.
don9GLC said:
My (simple) challenge is to automate a loco uncoupling from a two wagon train, then putting the leading wagon in another siding, to return and collect the rear wagon and place it in the other siding so that the wagon order is reversed. But without (relatively) high speed couplings.
I'm sure that this wouldn't be a problem to do.
don9GLC said:
And for advanced students, let a computer generate a random reordering of several wagons, and perform that. Then reset initial conditions and let you do it manually. Finally display your inadequacies!
I think getting the computer to generate a random reordering of wagons would be beyond me at the moment!
don9GLC said:
It seems you have opted for track detection. Are you using ZIMO products? If not, which and your experiences, please. And which loco decoders? ZIMO claim 'intelligent braking' so that the stop position is (relatively) consistent irrespective of approaching speed. but I have no practical experience.
I am led to believe that ZIMO make very good quality DCC components (at a price), particularly decoders with good slow speed running characteristics and other features that you point out. I don't have any ZIMO equipment at the moment, The decoders in the locos I used for my uncoupling and coupling demo routines are both by Lenz but the software is not using any special decoder feature and so it should work with any good quality decoders. Although I have not tried this yet I'm going to swap the locos in my demo for two others and see if it works, I think it will but I will let you know.
When I say the decoders don't have to have any special features they do have to have the ability to turn off any internal momentum, which most do via a function, and have good quality and well set up back emf control to keep the loco speed the same at each speed step regardless of gradient, curves, train load etc. All my decoders native LGB, Massoth, Lenz and ESU have these abilities (although the LGB ones don't have a function to turn off momentum, the same can be acheived by setting the CV controlling the amount of momentum to its minimum value)
The computer software does all of the speed and momentum control itself. Each loco needs to be calibrated against the software so that the software knows the loco speed for each speed step, there is a built in routine that runs the loco back and forth between two points at each step to work this out. Using this information the computer can very accuratly stop a train at given fixed locations (such as uncoupling ramps) regardless of initial approach speed, loco or length of stopping distance.
For occupancy detection I'm using the LDT RS8 on my demo layout, however my garden layout uses different occupancy detection hardware. The RS8 seems to work well although I wouldn't use it with high current draw locos.
Sorry if this is a bit techie.