layout automation for conventional controls

As a super quick experiment, reverse the wires to the track.  Bell and Horn should swap functions and you can see if Horn would work with a different "jump".

I know this suggestion is a money issue, but I found a good deal on a used TIU and remote and found a good deal on a used PS2 Daylight with passenger cars.  Now I just run the train record all the stops and sounds and then play it back and even on the loop playback feature and it works great.  Next layout I will have remote switches and I am looking for a PS2 Diesel to add.  The record/playback feature is one of the best things about my DCS system IMHO.

Are you enabling "just" the ~20k internal pullups on the Arduino digital input pins, or did you wire in a stronger pullup and possibly a small capacitor in parallel to filter any noise?

Yes, a very nice feature.  But I think MTH needs to provide a way to attach and apply position inputs/sensors into the AIU or whatever as John has done with his Arduino setup.  This allows him to stop the engine at the station every 3rd loop as he originally stated.  I realize you can kind-of sort-of do this without feedback with DCS but it's a your-mileage-may-vary proposition.  Perhaps when Lionel's new LCS SensorTrack plays out...

Hi;

Here are some results to share. My last update reported success with overall Arduino software control and the relay/diode approach for speed and bell control. The nagging problems where erratic behavior of the hall sensors and a non functional horn.

The hall sensor problem was traced to noise from nearby O27 switches. This took days and many test scenarios to pin down but the result was that any time a lighted car passed through a switch on the inner loop, there was a good chance a hall sensor on the outer loop would change state. Keep in mind that there are no physical connections between these loops. Testing showed that even though the hall sensor changed state it was not due to a nearby magnetic field or any signal carried by the rails, but rather from noise picked up by the sensor wires and fed back into the sensor changing its state. Lots of wire rerouting and isolation has cleared up all the hall sensor problems.

The horn fix was something I found by accident. So I'm not a big fan of the solution but it was the only approach that worked after testing many different scenarios. Instead of jumping diodes on the horn side of the home made controller, it was necessary to momentarily (1/3 sec) interrupt the bell side completely. This triggers the horn (remember from the video its an older sound commander board) for about a second and can be repeated as needed.

So the end result is that this Christmas layout is fully functional, fun to watch and has taught me a lot that will go into the design of a larger permanent layout. Also, the Arduino C++ code is breadboard tested for the future layout with dynamic routing of multiple trains based on block availability and switch control.

Thanks,
John