Great Southern Railway (Fictitious) - Signalling the changes...

[Compound2632]

I think that working out the layout of the physical locking tray will at least provide an explanation for why levers are arranged in a certain order and perhaps result in changes to your initial assignments, to optimise (minimise) the length of the locking bars.

[Skinnylinny]

I've tried to arrange levers that mutually lock/unlock each other fairly close together, while still having them in a logical order for the signaller. I'm still not finished yet, but I think I'm getting pretty close. You can see there are quite a lot of very short locking bars. 

 

The orange locking bar is annoying me because it's blocking an entire row. It might end up being placed "upside down" atop the pink top bar, freeing up the row for the locking of lever 20. There may well be errors, but the plan at present is to print the whole lot, test it, and then edit as required, as the brain fog has been hitting lately.

[Nick C]

18 minutes ago, Skinnylinny said:

I've tried to arrange levers that mutually lock/unlock each other fairly close together, while still having them in a logical order for the signaller. I'm still not finished yet, but I think I'm getting pretty close. You can see there are quite a lot of very short locking bars. 

 

The orange locking bar is annoying me because it's blocking an entire row. It might end up being placed "upside down" atop the pink top bar, freeing up the row for the locking of lever 20. There may well be errors, but the plan at present is to print the whole lot, test it, and then edit as required, as the brain fog has been hitting lately.

If you moved 10 onto 6 or 7, the lighter-purple bar would be shorter and so the orange one would fit next to it?

[Skinnylinny]

Now there's some lateral (quite literally!) thinking! Thank you! It would still fit in with the general layout of the diagram too - I shall do that.

Thank you so much!

[Skinnylinny]

Much better! 

 

Just lever 20 to do now... 

Edited March 5 by Skinnylinny

[Nick C]

For 20 there's an easy way and a hard way. If you're not worried about allowing the concurrent move from #2 to the down platform, you can simplify it to locking 3,5&14 both ways, and 12 & all the other signals normal. Otherwise you start getting into conditional locking...

[Skinnylinny]

Hmmmmm, I was *hoping* to allow simultaneous arrivals and departures, but I would definitely rather avoid conditional locking! 

So something like this? 


(I realise I've done the lilac bar in sort-of the opposite way to usual, but that's to avoid clashes with other ports in the same tappets).

The only niggling thought in the back of my head now is "Should 5 be locked normal unless 3 is reversed?"

Thank you so much for your help, @Nick C!

Edited March 5 by Skinnylinny

[Nick C]

42 minutes ago, Skinnylinny said:

Hmmmmm, I was *hoping* to allow simultaneous arrivals and departures, but I would definitely rather avoid conditional locking! 

So something like this? 


(I realise I've done the lilac bar in sort-of the opposite way to usual, but that's to avoid clashes with other ports in the same tappets).

The only niggling thought in the back of my head now is "Should 5 be locked normal unless 3 is reversed?"

Thank you so much for your help, @Nick C!

I've just realised that it is possible - 20 doesn't need to lock 2, because 3 being reversed would do so - so you can have inbound movements on both lines simultaneously. 

 

Yes, 5 should be locked normal unless 3 is reversed, in order to trap the down yard. As you don't have a facing move over 3 when they're normal, you could remove the notch there on the green bar and extend that - so 3N locks 4N and 5N, either of those reversed locks 3R?

[Skinnylinny]

5 minutes ago, Nick C said:

Yes, 5 should be locked normal unless 3 is reversed, in order to trap the down yard. As you don't have a facing move over 3 when they're normal, you could remove the notch there on the green bar and extend that - so 3N locks 4N and 5N, either of those reversed locks 3R?

Oh, I think I see what you mean - there wouldn't be any movements from the down platform across 3 unless it were reversed, so the FPL on 4 only needs to lock 3 in the reverse position. Very smart! 



I am a little worried that one port in tappet 5 (that's on row B when normal) will be rather too close to the dog on the green bar - this could allow the following sequence of lever movements (that I don't want!)

Starting all N
3R
5R
3N (while 5R!)

 

I must say, I'm glad I'm not doing this for a living!

[Nick C]

9 minutes ago, Skinnylinny said:

Oh, I think I see what you mean - there wouldn't be any movements from the down platform across 3 unless it were reversed, so the FPL on 4 only needs to lock 3 in the reverse position. Very smart! 



I am a little worried that one port in tappet 5 (that's on row B when normal) will be rather too close to the dog on the green bar - this could allow the following sequence of lever movements (that I don't want!)

Starting all N
3R
5R
3N (while 5R!)

 

I must say, I'm glad I'm not doing this for a living!

I believe in the prototype they're arranged so that the pitch of the locking bars is sufficiently different to the throw, in order to prevent that.

[Caley Jim]

I had a similar issue with the locking on Kirkallanmuir. I got round it by increasing the spacing between the locking bars. 

 

Jim 

[Skinnylinny]

Well, a fair bit of finagling later, and now things are stretched such that the tappets' throw is approximately 1.5 rows, thus stopping the problem. Whew! Now to try printing the bloomin' thing! 

[Skinnylinny]

Well, the printer's been on the blink for a little while now - it suddenly started producing plastic spaghetti instead of lever frame parts. About 8 hours of tinkering later (spread over about a week!) I've managed to get it working, and I now have an idea of what to look for if it fails again. My first successful print was the locking bars, and there are now the first ten tappets on the print bed. 

I've also been playing around with how to mount microswitches, and I think I've got a nice neat system which should do the trick, having stolen the basic concept from @Lacathedrale's suggestion a few pages back. Every third quadrant will have this tab with two holes in, to fit more threaded rods (M2 this time, rather than M3) to support microswitches. It *should* be possible, with care, to route the wires neatly downwards, although I may draw up some kind of cable trough for them. 

 

While the interlocking may be mechanical, the levers themselves will operate the microswitches and thus, indirectly, the signals and points through servos. I'm also hoping to add coloured strengtheners to the lever catch rod guides at some point, but for the foreseeable future I'll be printing lots and lots of black parts (interlocking and more quadrants!)

[Lacathedrale]

That looks fantastic @Skinnylinny! I await your progress with eager anticipation!

[Michael Hodgson]

Looks great.  Robustness/wear again - I'm intrigued by the use of plastic for the catch handles/rods/catches.  I hope if proves strong enough and that the catch rods don't bend unduly. 

I assume the pivots on the catch handles will be simple metal pins, or are you planning to print these in plastic too ?

Are you going to fit springs on the catch rods so that they do tend to drop and lock when the operator removes his finger?

 

One of the potential issues is whether or not you have many guest operators.  They can struggle with model lever frames especiially if fitted with interlocking, and ham-fisted operators can apply too much force to try and move a lever.  It's OK with operators who understand and who are used to the equipment, but newcomers can tend to be a bit heavy with them. 

[Skinnylinny]

Hi Michael,

So the pivots on the catch handles are M2 bolts, which pass through one side of the catch handle, and self-tap into the other side of the catch handle. The catch rods do have springs attached to the bottom where they pass through the quadrant. The springs hook over the hooked "leg" to the left of the lever in the picture above. They're fairly softly sprung, but enough that the catch falls reliably.

I am not expecting many guest operators, and the long-term plan is to have a mimic panel as an alternative method of control, that can be plugged into the same electrical connector that will attach the levers to the layout - the levers are mainly for my own enjoyment. That said, there's a *little* flex in the catch rods, but only enough to allow maybe 1mm of movement at the very tip of the lever handle. 

[Skinnylinny]

Getting there... 

 

4 more quadrants, 6 more levers and one more locking bar... and hopefully that should be all the bits I need to assemble the whole thing!

[Caley Jim]

How are you preventing the rods from jumping up out of the tappets on the locking bars?

 

Jim

[Skinnylinny]

Once everything is in place and tested, I intend to have a clear piece of acrylic over the top.

[Skinnylinny]

It lives! Well, all the levers, microswitches and locking are printed, assembled and working. Unfortunately, the 2mm threaded rod I ordered for attaching the microswitches is... subpar... The M2 nuts just slide along it rather than "biting" the thread! I may have to look for another source. 

 

There are still a couple of minor bits that need tidying (like painting the handles on some of the levers) or that may get changed - the strengtheners to the catch rod guides, for example. There's also a very slight occasional issue with the tappets binding slightly between the quadrants - this may settle down as bits bed in, or I may have to adjust the tappets slightly and reprint, but all in all I'm very very pleased with how the project has turned out! 

In the meantime, I've been working on something more modern - the ScotRail class 334 for the club's forthcoming Haymarket layout. Awaiting varnishing before I grubby up the roof and add the cant rail stripe, then do the glazing.

 

 

[Skinnylinny]

So i was sitting playing with the lever frame, and got to thinking. It's all very well having a frame, but what use is it if the next box down the line can't communicate to send me trains? 

 

 

Now, I could just rely on talking to the fiddle yard operator, but where's the fun in that? So, I've been playing about a bit more in CAD.

 

 

I've not yet decided how to drive the arms, but I'm leaning towards servos unless I can find a way to hook this up to some solenoids in such a way that I can drive it on 5V or so. It's to a larger scale than the levers in order to be visible to a 1:1 operator, and I've yet to decide how to sort out a pegging handle. The outer finish, however, I've decided on. I have some rather nice self-adhesive veneer, which should look rather good over a 3D printed case, perhaps with some suitable decorative wooden mouldings. 

 

Edited March 25 by Skinnylinny

[Annie]

Brilliant Linny, - absolutely brilliant.

[Caley Jim]

I can almost hear the 'ping' of the block bells from here!  takes me back to working the block bells in Greenfoot box in my teens.

 

Jim

[Skinnylinny]

Bells-wise, I'm leaning towards a speaker and recordings, to be honest, rather than physical bells. I love the idea of operating with them, but I daresay my neighbours would appreciate the idea of a volume control. They've enough to deal with now that I'm getting back into organ practice too (well, as best I can with a MIDI keyboard, and an old church organ pedalboard that I've managed to acquire and am planning to MIDI-ify...) but at least that I can do on headphones!

[Dana Ashdown]

On 25/03/2024 at 07:20, Skinnylinny said:

The outer finish, however, I've decided on. I have some rather nice self-adhesive veneer, which should look rather good over a 3D printed case, perhaps with some suitable decorative wooden mouldings. 

A real wooden case would be even better!