Large Scale Central

JMRI Wifi throttle

Coming at it from a different direction, Geoff Bunza built a Wifi throttle for JMRI based on the ESP32 boards. Just ordered a couple of the boards to play with. Lots of possibilities.

https://model-railroad-hobbyist.com/node/35652

Nice idea, I read the thread but did not see anyone total up the parts cost… wonder how much?

Also, it would be nice to have a much larger display.

Another question would be how long does a 9v battery run this? I would think that is a weak point, especially if you are running an OLED display, and for us, in sunlight, plus wi-fi…

But definitely cool and shows how simple the blocks can be assembled.

Greg

The ESP boards are ten bucks

https://www.ezsbc.com/index.php/featured-products-list-home-page/wifi01-35.html#.XEJcaM9Ki1s

I’m going to replace his buttons and switches with a keypad like my other throttles, and use the same 4 line display. And probably a flat-pack rechargeable battery

ESP32 boards arrived today. Here’s a quick size comparison. Left-to-right, the ESP32, Nucleo-32 L432KC, and an Arduino Uno clone

A bit larger than the L432KC, which is the same size as an Arduino Nano. Nice tidy package considering what it is capable of.

Now to build out a throttle.

It freaks me out how much power we have in such small packages these days. For ten bucks. Pretty cool.

Part of the niftiness of these things is the ease of programming. It seems like a couple times a year I find some new board that has more features, costs ~$10, and is Arduino IDE compatible.

Is this going to be like how we end up with so many engines/cars etc…(http://www.largescalecentral.com/externals/tinymce/plugins/emoticons/img/smiley-wink.gif)

A quick followup on the ESP32 boards (Ill probably make a new thread for this eventually).

I found the painlessMesh library, which creates a self-assembling mesh network of nodes.

https://gitlab.com/painlessMesh/painlessMesh

It took me about fifteen minutes to get the Arduino IDE set up with the appropriate libraries, and within five minutes after that, I had a three-node wifi network running the StartHere sample program.

Thats…amazing.

Will it check the wheel set back to back spacing and proper coupler height as well?

That’s pretty cool. I find the mesh concept very interesting but I have never looked at a source implementation.

Yea, I like the zero-configuration of this. I’ve been playing with trying to hack together a mesh network using HC12 modules, but they’re incredibly simple, and I havent got it working like I want it for my signaling. I think the throughput and ease of setup for these will let me concentrate on getting the signaling right without spending most of my time figuring out the networking bits.

So, clearly the advantages of a mesh network is increased coverage / effective range, and also you are not forced to have a central locus of control.

The idea of signals talking to each other, and effectively telling their “neighbors” what the status is appealing from a pure computer science viewpoint.

But, this also has some drawbacks… clearly the signals need their own local intelligence (but it’s getting cheaper), and also, the lack of a central control means if you want some overall status, you might have to poll the signals from another place.

Interesting and fun to consider.

Greg