As previously explained, I’m working on the use of an Arduino and an old laptop to control the points on my very small N Gauge model railway. Well, I’m also going to control the signals too.

Working on this track plan:

I consulted my tame signalling engineer (some say that he once completly ruined the brakes on my MG Montego, and that he spends his spare time writing cod-1950s hard-boiled detective story versions of his work, but all we know is that he’s called The Sig), explaining to him that all the lines were bi-directional. He put on his thinking-trilby, sharpened his crayons, and came up with this idea:

(Only the signals inside the box are to be modelled. The ones outside it are off-scene, and therefore just implied to be there and functioning.)

So, 6 signals. As I only had six remaining outputs left on the Arduino, I decided on 2-Aspect light signals.

And, lo, these were on eBay:

Six of them for £19 incl shipping, actually from http://hezhiqing.com/NSIGNAL.html (yeah, it says 5 on the website, but all the eBay listings show 6 of them).

Nominally they run off 12v but are plenty bright enough with 5v from an Arduino supplied to them. Plus, if you tie the Green line to 5v and the Black to GND, putting 5v up the Red line from the Arduino will switch the signal from green to red. This is counter-intuitive to me, as surely the signals should fail-to-Red, not fail-to-Green?

No, matter, it still make the wiring easier.

Eagle-eyes will spot the the lights are the wrong way up.  Ah, well, it’s too late now.

Next… build up the control board for one pair of points, and one signal as a test.

As stated elsewhere, I’ve mothballed the Arduino Dalek Project, and started to cannibalize bits for a new project: Arduino control of the turnout points on an N Gauge model railway layout.

For a start, heres some information about the point motors I’m using: The PECO PL-11 side mounted motors, which attach to the side of the points, like this:

These motors operate on 16v AC and require only (and indeed cannot tolerate more than) momentary current, and are currently controlled from “passing contact switches” drawing power from a Capacitor Discharge Unit.

There are currently 6 turnouts/points, arranged in pairs as below. As it would be reckless (and in real life impossible, due to interlocking) to operate either of the pair independently, I’ve decided to activate them together.

Thus, at all times both A and B will either both lead straight ahead, of both lead off to the left (for sets C&D and E&F it will be right not left).

As the Arduino can only throw out 5v DC, each pair of motors will need to be driven via a 5v DPDT replay. But, as we can’t leave the current on without melting the motors, each pair will actually need two relays (one for each direction of change), driven independently from different output pins. Another complication is that the relays will have to energised and then de-energised within a specific time-frame to avoid motor damage, but this is a programming issue to be dealt with later.

The above diagram shows only the 16vAC side of the wiring. The Arduino 5vDC is shown below.

Obviously I could have chosen to just use SPDT relays, and connect the outputs together, but due to the vagaries of how the motors work and the orientation and placement of them on the layout, I thought it was safer to design and wire it this way at the outset, rather than back-fix it later when it went wrong. Plus, I had all the wiring in place, so thought I might as well use it.

Note the lack of common Ground between the two halves of the circuit. Grounding AC and DC together isn’t a good idea. Just say no.

Anyway, here is all is again on one convenient image.

So now the next stage to to take that abstract squiggle and build it as a circuit.

Note – for myself, mainly: Each “pair” takes up 2 outputs, so for the three pairs of points we need 6 lines from the Arduino, out of a total of 12 (not 14, as I’d rather not use the 0 and 1 lines (RX & TX) as this could cause issues with data transfers activating the relays – something that the Dalek had problems with at first.

To be continued… with a start on Signalling!