Finally, after yet more distractions and life in general getting in the way I have got the major part of the wiring of the layout done. Last night saw the first train movements across all lines and points on this rather small layout (5ft x 2ft). There is still work to be done, the uncoupling magnets are not wired up, the wiring needs to be tied back and labelled and I need to build a control panel.

Although this is a small layout I wanted to use it to try out a number of ideas, so there is rather a lot of electronics in a small space. The control system is MERG CBus and the trains are driven by DCC. There is block occupancy detection, with 6 blocks, one for the point ladder and one each for the 3 sidings, the headshunt and the feeder line. The picture below shows a general view of the underside of the layout and the "spaghetti" that I have just finished installing.

The grey box in front houses the transformer and thermal cutouts that provide two 16V AC feeds to the layout via the thick black wire that plugs into the lefthand side of the board. These two 16V AC circuits feed out from the power socket onto some tag strip from which the 16v circuits go out to the general boards around the layout. Both the grey box and the tag strip have indicator LEDs for each 16V AC circuit, so that I can tell if there is an AC fault in the transformer box or the feeder cable. Also on the shelf with the power socket is a board that produces 5Volt and 12Volt supplies, each at one amp, again each of these supplies also has an indicator LED. From here the 12Volt and 5Volt supplies go to a number of supply distribution boards, again with LEDs. It is to these distribution boards that each circuit board is connected. The idea being that the supply can be checked at various points and that each board has its own connector on the block it get its power from. In the lefthand section there is also a track occupancy detector, a couple of electro-magnets, a servo for the point in this section and a DCC distribution board.

Colour coding of the wiring is important in my scheme, any pair of wires that are made up of red and black wires are always DCC bus or dropper wires. A single green wire is connected to a frog. Blue and green twisted pairs are the CBus, Orange and Grey pairs are 5Volt supplies, White and Violet pairs - 12 volt supply wires. Green and yellow pairs are for occupancy detectors. A pair of red wires indicates a 16V AC feed. Servos are always connected by 3 wire cables, unfortunately the extension servo cables I have are a different colour coding to the cables attached to the servos, but the servo cables are very obvious.

The middle section of the board is the really busy one, this contains the shelf with all the CBUS boards on it, including the DCC command station, the USB computer interface, the board that will control the electro-magnets (when I finally wire them up), the board that connects the occupancy sensors to the CBus and the servo control board. The picture below shows a closer view of this section.

As well as the shelf with the CBus boards on it, there are also a number of boards mounted on blocks of wood that I have glued to the underside of the 4mm ply baseboard. Immediately below the CBus shelf are a pair of power distribution boards, a 5 volt and a 12 volt. It is possible to see that the LEDs are lit on both, so we know power is getting this far up the baseboard. There are three MERG occupancy detectors (DTC8s) that detect the DCC current flowing by using small transformers. This means there is no voltage drop in the detector circuit and yo get very sensitive detection for no lose of traction power. One of these detectors has a small stripboard next to it that is used to join the stock rails on each of the points. The point ladder acts as a single section, but detectors are fitted to each stock rail. This means that a single conducting wheel on any part of the points should trigger the detector, but it also means that the frog is not used for detection. This is important as the frog are switched using frog juicers, so current detection would not work currently across the frog juicer.

The other boards that can be seen, one of them attached to the side beam, are frog juicers - a dual juicer and a mono juicer. A lot of dropper wires also come through the baseboard in this section, and each pair is feed through a cable tie base to act as strain relief and hopefully prevent the wires being pulled off the rails if snagged. The three droppers from one of the points can be seen just to the right of the servo.

The DCC bus can be seen running down the left side rail, this is made up of heavy duty wire and is feed to a set of DCC distribution boards that are mounted on the opposite side of the board to the CBus shelf. It was done this way to keep the DCC as far away as possible from all the signal electronics (and I don't mean railway signals, I mean low voltage electrical signals). This means all the red and black dropper wires run down the picture to the DCC distribution board that is just out of shot. Again the DCC distribution boards have LEDs so I can check the status of the DCC bus at each board.

The final section, on the right of the top picture, contains just a single servo, uncoupling magnet, and DCC distribution board. I have however included a close-up of the servo to show the way it is attached.

I bought some very cheap mounts on ebay that allowed me to fix the motors side on the the baseboard below the boards. I then ran a thin piece of piano wire through the tie-bar, looped it around the servo horn, across the other side of the horn and bent it at right angles. This means the piano wire pivots around the centre of the servo horn, so a very small movement of the servo is needed to change the points. It gives a very direct and simple to install mechanism, but due to the small degree of movement int he servo, less than 10% of the entire movement, it becomes hard to accurately set the end stops on the servo. I am considering replacing this simple mechanism with something that uses more of the servo's throw and should be easier to set up correctly. As it stands it does work, but the servos sometimes chatter, and this can not be good for them.

So having wired this little lot up, I connected a computer to the USB interface, configured all the CBus modules (I'm using FLiM mode for those that know about CBus) and built a JMRI panel to drive the points and display the track occupancy. To my great surprise I only had two real problems; a bad connection to one of the frog juicers which was resolved by reseating a wire, and a dead sport caused by paint on the track. I was able to run a loco across all the points and siding within 5 minutes of first completing the panel design. There are a few rough spots on the hand-built track that will need a little tweak; one joint has somehow managed to loose it;s fishplate and the ends of on of the switch blades is causing a wheel to ride up. Otherwise I am pretty please with my first hand-built track-work and more than happy with the CBus control system.

It was not long before I had a sound loco running and switched to using an iPhone to drive the loco. I have since added route controls to the JMRI panel - which makes it really easy to also set the points via the phone, and made it so the point control gets automatically disabled if a loco is in the point ladder at the time you try to change them.

So is it worth all this complication, especially on such a small layout?

Well, yes in my mind. I would probably not add so much if I was just building a shunting puzzle, I might do away with the block detection, but for anything bigger I would not. As a vehicle for testing out things it is really paying dividends. I would rather learn on this small layout then try things out and have my dream layout go badly wrong.

It will soon be time to start on the scenic side of things, but some housekeeping jobs need to be done with the wiring first - like mounting the CBus RJ22 connectors, wiring the uncoupling magnets, building the DCC handset and maybe even a hardware panel. Another thing I want to try is to connect my RFID stuff up to the CBus, again not really needed on a shunting puzzle, but it gives me enjoyment and a place to try out the ideas.

Now, where did I leave that labelling machine....