Thanks, Clint. As with most of my videos, I'm sharing as I go along so others can hopefully learn a little and not repeat any of the same mistakes. It's not as overly complicated as it maybe comes across if you strip out the first 20 mins of the video or so! Very pleased with the end result though, looking forward to getting the rest of the signals up and running now. Iain

Thanks, glad the GitHub page makes sense. KZfaq videos are great for showing a visual representation of what's going on, but I still also then like to see things written out or in a diagram to understand what's going on. Hopefully you're able to try it out if you'd like to build something similar. I'm toying with the idea of designing a circuit that I can share to get printed to make it easier to solder up the registers since I'll still need more for the north end of the station, and elsewhere on the layout when I get to them. Iain

Thanks! I haven't seen many other folks going this route, but it's actually a pretty good solution for what can really complex if you overthink it (like I did originally!). The end result is pretty cool. Iain

Thanks! I'm happy with how this turned out. Bit of a winding way to get there, but I learned a lot along the way, and it's actually pretty easy now I understand what's happening to quickly add more signals. Iain

Thanks, glad you like it. Feel free to share! I'm starting to dig more into JMRI and controls like this, so hopefully I can record more along the same lines soon. Iain

Thanks, John. Sounds like an interesting, and similar project. I have everything working, just have a few more lights to install. I don't have sensors installed yet, but I have the logic figured out. It's trickier to do that JRMI logic than the electronics for the sensors and lights! I'm slowly working through other projects, but getting there. Iain

Glad you enjoyed it, Paul. More info on the LED driver board is at github.com/fouldsy/arduino-jmri-cmri-signal-control. The link is kinda lost in the video description. In a later video around Christmas I did all the north side signals using the commercial LED driver boards, just had to wire the signals differently. Both work great. All the best. Iain

Thanks. There is a lot to it, but now that I understand more what's needed, the complicated part is soldering all the resistors for the shift register. There's no code involved or programming once it's set the first time. It's still not plug + play, but gives a lot of flexibility. Thanks for following along. Iain

Thanks for watching, Keith. Yes, I'm using the Auto485 library with the Arduino - github.com/madleech/Auto485. This library simplifies the connection part. It may help to look at github.com/fouldsy/arduino-jmri-cmri-signal-control where I wrote up how I'm connecting this, and then the Arduino code for communication. If I recall correctly, on line 10 where the DE pin is defined, that's using an internal pin on the Arduino for the serial communication. Line 11 then creates the Auto485 connection. I didn't change anything for that, and there's nothing physically wired up for it. Hopefully this helps! Iain

Newcastle Central (OO Gauge) Hi Iain, Many thanks for your swift response. Would I be correct in thinking that you’re using a bog-standard USB cable between Pi and Arduino? I couldn’t spot any RS485 hardware as such. Regards Keith

Correct - just the regular USB cable for the Arduino (USB A-B, I believe). The Auto485 library handles the basic communication that's required here. I think in more advanced scenarios you may need the hardware adapters, but for basic communication like I'm doing here, it's not needed. It's all done in software. The CMRI standard seems to pre-date pretty much all the JMRI and modern DCC equipment, so a lot of the CMRI docs are overly complicated for what's available now. At the time, it was rather impressive, I'm sure. Iain

Newcastle Central (OO Gauge) Hi Iain, I was hoping you’d say that - I’ve already got points and sensors sorted with Arduino DCC decoders and was looking for a straightforward expandable solution to signalling - you’re project is exactly what I’m looking for. Regards, Keith

Yes, signaling seemed overly complex with most of the solutions I looked at it. When I discovered CMRI, it seems like a perfect fit. All you want to do is monitor DCC state and shift out light power state. Works great. Good luck with your project! Iain

I got them from Adafruit, though I imagine other vendors are available - TLC5947 24-channel (www.adafruit.com/product/1429) or TLC59711 12-channel PWM LED (www.adafruit.com/product/1455) driver boards, depending on how many LEDs you want to use. Good luck with your projects! Iain

I haven't used the Digikeijs systems, but it should work just fine. I know C/MRI is rather old at this point, but it's also really light-weight, and perfect for controlling a large number of signals like this. There's just not a lot of talk about it. I think the prevalence of commercial signal lights that include DCC give the illusion that's the way to go. The approach in this video, other than having to solder your own shift register boards, is so much simpler and cheaper. I have 15 signals just on the south side of the station here, with another 14 required on the north side of the station. That would start to get really expensive without the approach I'm using here! All the best with your build. Iain

Oh, i guess you did :)

I uploaded the files to www.thingiverse.com/thing:3635986 I also then had a separate video just on 3D designing, printing, and building the signals and gantry at kzfaq.info/get/bejne/idGdZJRm1eC5fac.html Good luck! Let me know how they work out. Iain

Yes, I did some early prototyping back in September/October that used photoresistor sensors, and I have seen setups using current detection for real block detection. On the 1/8th scale railroad I'm a part of, the track signals are controlled like that. I'd still need this kind of set up with C/MRI and JMRI routes to initially set the lights based on the turnouts set for a given route, but I could use something like you suggest to sense when a train enters a section and flip the light from green back to red. All I need some kind of sensor for is to set a signal light back to red when a train passes it. Ideally I'd still want JRMI to know about that sensor state so I can then track the state as a loco would move between two or three signals, for example, to then be able to update the downstream lights as needed. Wouldn't a current sensor require each section of track to be isolated? I have dropper wires for the DCC bus on each section of track, but I didn't use insulated rail joiners other than on the electrofrogs at turnouts. Would I still be able to sent the current when a loco passes the sensor okay? Iain

@@NewcastleCentralOOGauge The module that I've developed (with help from a few other sources) would require the block to be isolated on one side only (the red or nearestside in my case) so that it behaved as per the full size track circuiting system. It would only require one isolated track joiner to be placed adjacent to the signal so as to isolate the sections between signals. It can also tell the signal in the rear that it needs to show an amber aspect and can be told by the next signal in advance the same. I've not moved onto four aspect yet but it's only a matter of time. The module (per signal) comprises of a bridge rectifier, a bi-directional optocoupler, a 555 timer, a 74HC00, plus a couple of capacitors and four resistors. If you price those up, I'm sure you'll agree that it's a very cost effective way of running autonomous signals. I'ts also possible that the sensing circuitry (basically without the 74HC00) can be connected to a C/MRI or JMRI system as the 555 is used as an analog to digital converter. when there's a train in the section, the output from the 555 is high, when there's no train the output is low. I've also got a sensor that can send out a filtered analog signal that can be used with an Arduino when connected to an analog pin. It is so sensitive that it can detect a 10K resister placed across the track yet give no false output at any other time, even with DCC. If you would like any more details on this I'd be happy to send them to you. Best regards, Phil

Thanks for all the detailed information, Phil. Definitely a few different ways to approach this. It would be hard to split all the track segments to isolate one of the rails now, but doable. I know Canal Sidings has done quite a bit with exploring signaling like this and I think went with the IR sensor route. I'll keep looking around at options, wanting to wrap up installing the remainder of the signals on the south of the station first. Iain

@@NewcastleCentralOOGauge No problem Iain. If you decide that you'd like any more detail then let me know and I'll send the schematics over to you. Love the videos. All the best, Phil

Much appreciated, Phil. Seriously. Like I said in the video, I'm by the no means the expert here, just sharing what I've come across and my understandings of how things work. I think input sensors are going to be somewhat similar - I'll try different approaches until I find one that works. At that point, it probably isn't that complicated to get going! Aside, I ordered some IR sensors to compare against photoresistors (which I'm leaning away from), so can explore those options soon. Thanks again. Iain

Cheers, Mark. Sent a follow-up; let's get together soon! Iain