Very cool.

Thank you! =^.^=

wait, a cart that can fit 50 feet? Thought that they are either 40 or 60.

Many domestic inter-modal containers in Japan are non-ISO, and largely shorter than 40ft. These standardized types of rail cars can accommodate 60ft worth of containers. Japanese containers tend to be 30ft (2 per car), 20ft ISO (3 per car, like the JOT tanks & Seino containers I have), and the ubiquitous 12ft (5 per car, the red JR containers). One of the major reasons for the smaller container sizes is the density of many Japanese urban environments and the difficulties larger trucks can have navigating the narrow roads.

Love those EMUs

For all the fancy trains in Japan, I can't help but find JR Central's rather "friendly" looking compared to most.

i like your hobby….. its SMART…… im unintelligent so i cant handle hobbies like yours…… but its impressive……. i just play videogames & draw & sometimes make music…., but u conduct trains like an engineer!!!!!!

Well, you don't need to be "smart" to get into model trains these days. Especially with things like the Kato Unitrack system. It's quite literally all snap-together and unless you're getting too fancy, hooking up the power to the correct places isn't much more difficult. Of course, it can get pretty expensive pretty quick... but that's the thing with most physical hobbies these days...

Was looking back and see I forgot to post on this one.

Even many non-physical hobbies can kill your wallet in a hurry. I've spent far more than I should on figures/props for rendering in DAZ (and just blew another $600 on ram to upgrade the system I run it on!)

I still have a few N-Gauge train bits, but they're in boxes because my eyes are getting bad enough that I have to get my reading glasses to see to re-rail those tiny wheels. (and I use a large magnifying lens to see to place/solder my electronics - and no, I can't manage SMDs! )

Kato has a wonderful plastic re-railing tool. It's a sanity saver.

Learning a bit of electronics has definitely cut the costs down. I've just put together a 4 channel 555 timer PWM board to connect to some Arduino H-Bridge power supplies to allow for up to 4 train control to take the place of up to 3 addition Kato power packs. And the plus side is, the Arduino can act as an alternate PWM input with no other modification.

Which is why I have a few spare Arduino around just to play with! The little ones have 6 8-bit PWM outs, 6 10-bit analog inputs, and 6 more pins for in/out (8 if you don't want/need it to talk to your PC while running.) Or on the larger one I saw in one of your other shots, 12 P, 16 A, 3 more pairs that can be used as comm ports and some interrupts, 70 data pins total.

Not wanting to pay for higher amp H-bridges for G-Scale (the 2 amp ones just won't cut it!), we have a few high-amp MOS-FETs being fed the PWM for speed and a smaller MOS-FET flipping a relay for train direction (no current flowing when flipping the relay means you don't arc/wear out the relay contacts. )

It took my brother a few tries, but he has his now programmed to properly slowly ramp the PWM up/down to the speed he has set, no more trains jumping to speed or skidding to a stop! Someday he does want to have it set up where the Arduino can see which blocks of track have trains on them and run things automatically. That won't be for a while!

I've already got a prototype program for train control which behaves on a "set the speed pointer" system, where every loop runs a subroutine that checks to see if the train(s) is at the speed set elsewhere in the program, and accelerates/decelerates/maintains speed accordingly, and the acceleration.deceleration rate varies depending on the current speed for softer starts and the final stages of stops.

I already have a bunch of IR sensors to mount under the track facing up to detect cars, for train control and to light up position LEDs on the future control panel too. I'm finding there's a bit of an issue getting the trains to stop in the same spot every time using them as triggers though. Gotta work out the source of the unpredictability at some point.

With IR, most likely issue is stray light being seen as in other lights or reflections holding it on. Another is how your program is 'watching' for the trigger, if your 2560 is watching/doing several things at once it can be finishing up one task before it checks the sensor - which might have been just tripped one time - and three cars in the next. One other issue is each of your engines will run just a little faster/slower - and the load they're pulling will matter as well (for added fun, each engine will run different 'cold' than they will after 'warming up'.)

Stray light might be countered by smaller holes/tubes on the sensors and making sure the bottoms of the engines aren't too shiny.

Testing the 2560 being busy at the wrong time would need you to add LEDs you can see that light when the IR sensor says 'now' and have your 2560 light Another LED when it says 'I see it'. (I had a similar problem and used some RGB LEDs. The sensor lit the red, the Arduino the green; so any red before yellow was late Arduino trigger, and any green before blank was late Arduino release. YMMV )

If those show nothing (or you want to try something easier first) the last test/trick is to run up one of your engines by itself a few laps. Once warmed up start testing the stopping several times. If it's 'behaving' then you may want to test just tell the system which engine/load each one is for proper stopping. However if it's all over the place, you may want to add a few more sensors coming up on each stop; the idea being is when the 2560 sees the train at sensor one and starts to slow it it expects it to trip sensor two after X amount of time. If the train is early then more/faster slow is added, if late less slow, and this continues on up to the stop. (Those sensors get closer together the nearer the stop to give the 2560 better feedback.)

Sorry for the long-winded-ness, if the issue was more defined I'd have better aim rather than this scatter-gun guessing.

Have fun!

Better than using LEDs, leaving the Arduino hooked up to the laptop, and using serial output to display all of the event timing/exact PWM levels in real time. Even handier still with the USB oscilloscope, since I can match the serial output to measured signals too.

I'm definitely going to have to use multiple sensors anywhere I need trains to stop in a specific place, like the station platforms. My thinking about the unpredictability is that it's more to do with the trains themselves and PWM than the Arduino/sensor combination. Even using my custom 555 based manual PWM controls, precision speed control at low speeds is rather lacking compared to the Kato power pack.

Quick question/thought, does the Kato power pack use PWM out in its speed control? (If so the mixing of two PWMs can give you some really strange actions!)

No, it uses straight DC that is incompletely filtered to give it the best of both worlds. Wouldn't have more than one power source connected anyway, since that would risk all sorts of problems if both were on at once.

Just checking, my brother had just plugged the toy I made him between the train power pack and the track and cranked it up. Older unit, so no PWM to worry about (he does like that the PWM on the track cleaner toy gives him a better low speed/creeping mode.)