Festiniog Railway signals

[Trevor Thompson]

I have started another project ( I know I haven't finished lots of other projects ) but it is the result of exploring what my new printer can do.

If you have been following what I have been printing you will know that I am in the middle of printing a set of Birmingham 4 wheel coaches to go behind Palmerston. I have become increasingly concerned about the health implications of the photo sensitive resins - so this is an opportunity to see what a the latest filament printers can do. I am leading up to a filament printed Birmingham coach for a comparison.

So the first thing I tried to print was the Festiniog Disk signal I have been planning to build for a while. This is the CAD model:

Screenshot 2024-11-15 at 10.26.30.png (30.32 KiB) Viewed 2166 times

The plan is to use a 2mm diameter brass tube as a shaft which will have a wire passing through it, and an LED on top of it so that the lamp works. Perhaps a servo in the base to turn the disc, and a single 3032 cell in a switched holder to power the LED.

So I tried printing the main body as one piece to see what would happen. Here is the result:

IMG_4209.jpeg (2.06 MiB) Viewed 2166 times

That was printed in ABS as you see it with the bottom of the base on the printer bed. ABS is difficult to print with because it tends to split as it cools. This print had only one split in it in one leg of the ladder - and that was at the level where I had paused the print to adjust a cover in the printer. The very long supports worked and supported the platform properly but the frame under the platform broke off while I was removing the supports. It withstood some rough handling. It worked but I think there are advantages to making this as components and bonding them together. The glueing is so easy with ABS ( the solvent is acetone) that it is the obvious way forward. This printer is certainly a step forward in that I can now print in ABS without a raft or brim. The textured metal bed does not come out as the last one did and I was concerned about how I would get the components off the bed - without a raft to reduce the adhesion. Particularly as I couldn't get it off the bed as soon as the print had finished. There is a glue which they insist you apply to the bed before first use. Something I have not had to do before. I was amazed when I went back after it had cooled and the print just fell off the bed.

It is also fast. The castle towers which I have also been printing for a while took 4 prints each of 12 hours on my old printer, in this one I can do the whole tower as one piece in 6 hours. So that project will move along swiftly when I get back to it.

I have fine tuned the design and I am now printing 10 sets of components.

The ladders and the discs are printed:

IMG_4210.jpeg (2.75 MiB) Viewed 2166 times

More as it progresses.

Trevor

[philipy]

That all looks pretty good Trevor, although it looks as though you'll have a lot of elephants feet to clean off the ladders?

What is the glue you put on the printer bed?

[Red Star]

It's the little bits of detail!

I'm currently building working semaphore signals using aluminium tube, with the arms cut from laser laminate and PLA mounting brackets (well coated in enamel paint).

Will be keeping an eye on how these progress.

Andrew

[Trevor Thompson]

That all looks pretty good Trevor, although it looks as though you'll have a lot of elephants feet to clean off the ladders?

What is the glue you put on the printer bed?

The elephants feet look much worse in the photo than they really are. They are only 1 layer of filament deep, and were swiftly removed with a needle file. The glue us something which came with the printer, here is a photo of it:

IMG_4220.jpeg (1.53 MiB) Viewed 2107 times

I don't really know what it is. It certainly isn't like "prit stick". The best I can do is to describe my impression after applying it from the slightly "squeezy" bottle with a felt pad (when you remove the lid). The light colourless liquid glue passes through the pad when you apply it. When applying it I can see that there is a liquid on the bed, by the change in colour, but otherwise I wouldn't know I had applied anything. It dries quickly and leaves no stickiness.

I have applied it once since starting to use the printer - but have no idea how often I will have to reapply it yet.

I suppose it will just be another consumable that I will have to replace when it runs out.

I mentioned earlier that I was able to print without a brim or raft - but I have found the limitations on that statement. Small items which have a small area on the bed need a brim or they just fall off:

IMG_4219.jpeg (2.19 MiB) Viewed 2107 times

This brim is very thin and I have been able to cut the parts out with a scalpel.

And the first signal assembled (with the disc temporarily mounted on a 2mm drill):

IMG_4217.jpeg (2.24 MiB) Viewed 2107 times

Trevor

[philipy]

The signal looks really good. Are you going to motorise them or just have them static for show?

The glue sounds a bit like "3D Lac Plus" which is also a clear liquid and can't be seen once dry, it comes in a pump spray bottle. I've been using it since Steve put me on to it several years ago and I think in that time I've re-applied it twice.

[Trevor Thompson]

For now I will be satisfied for them to just be static, just adding to the "Festiniog" feel of the railway. However I would like them to operate, perhaps by radio control. As a retired electrical engineer I dislike the idea of wires, and more importantly plugs and sockets, permanently exposed to our damp environment. So I would prefer a receiver and battery in each signal, and I don't know how to achieve that yet - particularly making it as small as possible.

Part of the challenge is the particular (peculiar) arrangement at each station. There would be two disc signals, one at each end of the passing loop, two "capstans", which in the real thing control the discs, on the platform. There would be a double arm signal in the centre as well. Ideally the capstan should be interlocked with the appropriate disc. So that is 5 servos and 3 channels. I am sure it can be done but lets just make the signals first!

I now have 10 disc signals printed and assembled, and there is the prototype capstan in the foreground:

IMG_4222.jpeg (2.31 MiB) Viewed 2062 times

The capstan was resin printed, it is stable ( as in not warped in 7 months sitting on the bench) but it needs redesigning to be robust enough to go outside permanently. The bases of the disc signals have the brass rod holding the disc passing down through them so a bevel gear can be fitted onto the rod - intended to connect to a servo. The base of the signal is the lid for a box to contain a servo and electronics. The box is just a drawing at the moment - it cant really be finished yet - I don't know big the electronics and battery will be.

The same lid will have to be fitted as a base for the capstan as well.

I think the 2 arm signal is next.

Trevor

[philipy]

I would like them to operate, perhaps by radio control. As a retired electrical engineer I dislike the idea of wires, and more importantly plugs and sockets, permanently exposed to our damp environment. So I would prefer a receiver and battery in each signal, and I don't know how to achieve that yet - particularly making it as small as possible.

I assume you've been looking at Rik's current thread where he has done what appears to be exactly what you are talking about?
https://gardenrails.org/forum/viewtopic ... 20#p182420

[Trevor Thompson]

I would like them to operate, perhaps by radio control. As a retired electrical engineer I dislike the idea of wires, and more importantly plugs and sockets, permanently exposed to our damp environment. So I would prefer a receiver and battery in each signal, and I don't know how to achieve that yet - particularly making it as small as possible.

I assume you've been looking at Rik's current thread where he has done what appears to be exactly what you are talking about?
https://gardenrails.org/forum/viewtopic ... 20#p182420

Yes I have and that is what has inspired me. However he seems to have the wires permanently installed and a central receiver - which is what I don't actually want. Not intended as a criticism of Rik for whom I have a great deal of admiration. Just concerns for reliability and the difficulty of fault finding poor electrical connections. I really want to follow Rik's ideas but with a system where all the electronics goes indoors when I take the signals in.

Trevor

[ge_rik]

Yes I have and that is what has inspired me. However he seems to have the wires permanently installed and a central receiver - which is what I don't actually want. Not intended as a criticism of Rik for whom I have a great deal of admiration. Just concerns for reliability and the difficulty of fault finding poor electrical connections. I really want to follow Rik's ideas but with a system where all the electronics goes indoors when I take the signals in.

Trevor

Hi Trevor
What you describe is my old system from around ten years ago. In my latest version, each signal is self contained with its own battery, receiver, microprocessor and servo

https://riksrailway.blogspot.com/2024/0 ... phore.html

Rik

[Trevor Thompson]

Yes I have and that is what has inspired me. However he seems to have the wires permanently installed and a central receiver - which is what I don't actually want. Not intended as a criticism of Rik for whom I have a great deal of admiration. Just concerns for reliability and the difficulty of fault finding poor electrical connections. I really want to follow Rik's ideas but with a system where all the electronics goes indoors when I take the signals in.

Trevor

Hi Trevor
What you describe is my old system from around ten years ago. In my latest version, each signal is self contained with its own battery, receiver, microprocessor and servo

https://riksrailway.blogspot.com/2024/0 ... phore.html

Rik

Thanks for that Rik - I obviously didn't find the up to date information. I will follow the link.

I have read it. It is exactly what I want to do! and I like your use of Arduino (because I have the software for it on my iMac).
Thanks again

Trevor

[Trevor Thompson]

The brass tubes have arrived and I have fitted one into a signal.

I have some enameled copper wire left over from rewinding a coil, which was the right size to pass through the tube. With the tube cut to the correct length and the ends de-burred, the wire passed through and the end if cleaned and tinned:

IMG_4228.jpeg (793.74 KiB) Viewed 1620 times

The LED has been soldered to the copper and the brass tube. The black collar which forms the base of the lamp fits up from the bottom (it won't fit over the LED, or its leads) followed by the disc. Installed in the signal:

IMG_4230.jpeg (1.05 MiB) Viewed 1620 times

The lamp itself is a firm push fit over the base already fitted to the tube. It is removable, as I will need to add red acetate to two of the windows in the lamp before pushing the lamp onto its base fully.

Perhaps I should add that the lamps and bases were all filament printed. I think they are good enough not to bother printing them in resin:



I am about to print bevel gears one of which will go on the bottom of the brass tube, and the other on the servo when it arrives. Then I can connect the resistor and the battery. I should add that I have tested for continuity, insulation, and that the LED works!

Trevor

IMG_4232.jpeg (1.56 MiB) Viewed 1620 times

[Trevor Thompson]

Underneath the signal I intend to mount a servo, so I created a sketchup drawing of a course ( as in big teeth) bevel gear which I hoped would work as a printed gear. It printed on the filament printer well:

IMG_4236.jpeg (1.59 MiB) Viewed 1032 times

I have temporarily connected a servo to the base with black tack, and wired up the light:

IMG_4239.jpeg (2.26 MiB) Viewed 1032 times

I won't share the video of it, but it rotates through the required 90 degrees, and the gear meshes quite well. The one thing I learnt was that the wires connecting the battery to the shaft and enamelled wire were resisting its turning. But the light did work, and when the coloured acetate is fitted it will change colour, from white to red . So that requires a different solution. I could mount the battery wires on a printed block to press the bare ends onto the enamelled wire and brass tube - like the slip ring in an electric motor Or perhaps coiling the wires like springs might work. To be tried.

I now have the remaining 9 discs, LEDs and shafts assembled:

IMG_4248.jpeg (2.48 MiB) Viewed 1032 times

8 of them work and one does not light. So something to be repaired tomorrow - probably damaged the LED while soldering it.

Trevor

[ge_rik]

I really like the signals and the way you've gone about animating them. Looking forward to seeing them in situ and in action.

Rik

[philipy]

I agree with Rik, can't wait for the video, working, in situ.

[Trevor Thompson]

I really like the signals and the way you've gone about animating them. Looking forward to seeing them in situ and in action.

Rik

I am glad you like the way I have tried to do it (lets see it in action before saying it works!). However I have to admit that without your example to follow I don't think I would have got this far for a while.

Trevor

[Trevor Thompson]

The disc signals are progressing.They are all painted and I have started to assemble the electronics. More on that when the first one is working.

I have moved on to producing the semaphore signal. The final design:

Screenshot 2024-12-08 at 20.23.41.png (102.02 KiB) Viewed 196 times

and the prototype so far:

IMG_4255.jpeg (2.18 MiB) Viewed 196 times

The 10 disc signals are lined up behind it as well. These signals are 3 aspect (all right, caution, danger). One semaphore for each direction. The lights were white, green and red, denoting the meanings above (and in the same order). I initially planed to make these as just something to grace the line side, but I want to be able to control them remotely in due course. Thinking ahead I should work out how to implement those three colours. I have found tri-colour LEDs, Red green and a third colour by turning on both red and green. I wonder what colour that would be! So I think a test purchase is in order. I can see how that might work - with 2 tri-colour LEDs in the lamp one lighting in each direction and following the semaphore indication. Just more complex programming.

There was a time when I wondered if some form of "block working" could be controlled by the signals - but I realise now that would be totally inappropriate for a railway set in the 1880's. Block working wasn't widespread until 1890, when the railway inspectors gained the legal power to make railways change the way they operated. So really it is back to line side ornaments - but more interesting if they actually work!

Trevor

[Trevor Thompson]

Of course all of this has caused me to investigate the "Old company signalling" in more detail. Boyde mentions point indicators. So searching on Festipedia I came across a drawing of such a point indicator - and photos as well:

Old_Company_Point_indicator_main_JMW_sketch.jpg (59.59 KiB) Viewed 189 times

So I couldn't help drawing it. Here is my prototype - as far as I have taken the design so far:

IMG_4252.jpeg (1.83 MiB) Viewed 192 times

So you can see that it is a fairly big item, and I have never seen it modelled. I initially thought it should be something else to grace the line side - but actually it might not be hard to make it actually work. The semaphore arm is hidden within the body, hanging down for straight on and at 45 degrees if the point is set to diverge - in the direction it is set for. My points all have a 1mm diameter rod operating the blades. I drill a hole through the main rails and the blades so that this rod can pass right through, at a level below where the wheel flanges will pass. The rod is soft soldered to each blade. So at worst I would need to replace that rod with a longer rod the extension of which could operate the indicator.

It is just an ornament - but it would be interesting if it actually showed what it was meant to show!

Trevor