Jump to content

Semaphore Signals - 4mm Scale (Mainly)

Recommended Posts

  • RMweb Gold

They were definitely called 'Backing Machines' in Reading S&T Works Steve - they even had a special rig for testing them after reconditioning although as Beast's pic shows they could be a bit tardy in operation at times (not that ever noticed the ones at Moreton Cutting were like that on the few occasions I saw them working from the lineside).

Link to post
Share on other sites

Thanks Das, Mike.

It will be up to the Layout Owner to determine how these are going to operate.


I normally wire the switches for a multi arm signal in such a way that an incorrect combination cannot be displayed.

To achieve the "delay" in the Distant returning to Danger I would introduce an extra Micro-switch into the system.

This would sense the Home going back to Danger before switching the Distant back.

Normally the switching is simultaneous, using extra poles on the controlling switch.


We'll all wait and see................



  • Like 2
Link to post
Share on other sites

The Backing Machines.....


Using the documentation I received from Richard (LNERGE) my friend Les Green produced the necessary 3D drawings for these printed beauties:




This little print from Shapeways contains thirty Westinghouse Signal Machines of two styles.


Style G is for operating the semaphore arm from the Right side of the bottom of the post.

Style K is for operating the semaphore arm from the Left side of the bottom of the post.


They appear to be identical except for the mounting bracket being to the Left or Right respectively.


I'll be using style G on this signal.



You can just see that the mounting bracket is printed with a concave surface of 1mm radius.

This enables mounting directly on a Round Doll.

By one stroke of a fine fine, the mounting bracket acquires a flatter surface, suitable for mounting on a Square Doll.


The operating shaft will be made from my favourite Lace Pin, the bearing hole being printed in place at 0.5mm dia.


More soon....




  • Like 4
Link to post
Share on other sites

Wow they look good. Don't forget the hiss/bounce equation.. More hiss ,the dashpot is working = no bounce. No hiss, no one bothered to adjust the dashpot/threw the screw away = more bounce <G>


Edited by LNERGE
  • Like 1
Link to post
Share on other sites

I've just been having a search through the internet, as you do at 2am and you can't sleep, and I've found these two pictures and information. One is of a very nicely restored finial and the other is of one in not so good condition. But the important bit, i think is that they give the height of the finial in their catalogue.


attachicon.giflswr finial.jpg


attachicon.giflswr finial 1.jpg


attachicon.giflswr finial auction 24in tall.jpg


I did think that you could get some measurements from the first picture using the bricks as a scale guide. If so the base plate appears to be about 10 - 12ins square, assuming a brick and mortar joint is 9ins. Trouble is, the angle the photograph is taken, means it doesn't correlate with the 24in height given in the auction catalogue. You might be able to make more of it than me.


All the best



I forgot to mention about the finials when showing the Westinghouse motors on the previous posting:


On the edge of the printed item from Shapeways, Les has included the bases for the LSWR finials which I was gathering info about.

He also drawn up and had etched these N/S items, which form the upper part of a finial.


They are slotted to assemble into the cruciform shape and will then fit into the small base parts on the 3D print!.

The bases are printed with a mixture of circular and cruciform holes to see which works.

I'll not be tackling these for a little while - they're incredibly small and delicate!!!!!!!



Edited by SteveAtBax
  • Like 4
Link to post
Share on other sites

The finials are looking good Steve. A nice little etch. Are you hoping that the base will hold them in perfect alignment and at 90deg. to each other, or will you have to fix them some how?

Link to post
Share on other sites

More Bits & Pieces prepared............


Whilst all the 3D printing was being dealt with, I prepared most of the other bits and pieces which will go in the final model:


First the foundations.


This is the Foundation Tube, into which the main post will be mounted.

Turned from Brass Bar, the flange allows it to be located squarely in the signal's Base Plate:



The Guide Tubes for the operating wires are soldered into the Base Plate, and a scrap of brass etch holds them steady:


Finally this 1/2in dia tube protects everything and ensures the signal will be mounted securely in the Base Board, and the servo assembly will be accurately aligned (when its contructed later).



More Bits & Pieces to follow..........




Edited to correct spelling.

Edited by SteveAtBax
  • Like 3
Link to post
Share on other sites

MoreBits & Pieces............


The arms were assembled from MSE etches of Spectacle Plate and Arm on a Lace Pin:


Three Distants and one Home required.


The Signal Lamps are MSE castings, drilled through (carefully!) at 0.5mm to take the fibre optics.


Four of these.


The Balance arm is a lost wax cast brass item from MSE, with a white metal weight soldered on.post-3984-0-81405300-1420389275_thumb.jpg

Only one of these.


Time to get it assembled...........


  • Like 3
Link to post
Share on other sites

Putting it together.....

All the items assembled so far have been Brass or Nickel Silver.

To avoid problems with bits falling off, I've tried to complete all the sub-assemblies which use "ordinary" solder, and to assembled these into the main structure.



Here the dolls have been added to the main bracket.



The guide for the operating wires has been added, and the two "rockers" connected up.



The main post is fixed to the base assembly.


Ready for the Backing Machines...........



  • Like 1
Link to post
Share on other sites

The "Backing Machines".........


A few posts ago I showed the 3D printed Westinghouse Signal Machines which Les had made for me.

Now to fit them.


From prototype photos it seems that the machines were normally located a similar distance below the Distant Arm to be moved, regardless of the Doll height.

To achieve this, and to ensure that the bearing hole through the Machine is parallel to the Signal Arm Bearing, I first made a little aid from 0.45mm N/S wire.



Just two right angles, but when used in the two bearing holes it ensures they are correctly spaced and aligned.


The model Machines have their mounting bracket formed with a 1mm radius concave face.

This gives a good fit on the 2mm dia Doll.

The surface of the Doll is abraided with an emery board to give a good key for the Cyano adhesive.



The front



and rear "Cranks" are from modified Alan Gibson etches.



This gives some idea of how they will look in use.

You will notice that the white metal Signal Lamps have also been soldered in place.

I use Carrs 100deg. soldeer these days for all white metal soldering.

The important benefit is not having to tin Brass or N/S before using it.


More soon......


  • Like 4
Link to post
Share on other sites

A Wash and Brush Up ...............


With the main assembly completed, and all the other bits and pieces prepared, it was time for a good scrub and polish then off to the paintshop.


Those processes entail quite a lot of time waiting for things to dry, so I got on with preparing my usual Transport & Test Frame, and getting the servos mounted.



This is the frame, which will be used from now onwards to hold the signal during all subsequent assembly processes.

The top surface represents the location on the layout where the signal will be installed.

There is a space restriction to the right of the signal, so everything is kept clear of this. The line on the Transport Frame indicates the limit.


The layout surface is 10mm Sundeala topped with 2mm foam.

The top of the frame is therefore 12mm thick MDF.



Because I'm not sure if Sundeala will offer a firm enough anchor for the Servos, I've prepared this little sub-frame.

This is intended to be glued securely to the underside of the Sundeala, and the captive 6BA bolts will then hold the Servo assembly securely.



A short length of Brass Tube ensures all is properly aligned.



The sub-frame is screwed to the Transport Frame rather than glued - that would make it difficult to transfer to the layout!



This is the servo assembly.

Its designed to allow easy alignment of the operating wires with the servo "Horns".




This is how the Servos are located.



To make it easy to locate the nuts on the 6BA screws, I made these "extended nuts"



All assembled, using the Brass Tube to ensure alignment.

All it needs now is the signal.............



Edited by SteveAtBax
  • Like 2
Link to post
Share on other sites

Back from the Paintshop........


The whole signal has had a few thin coates of Humbrol's best white primer.

All the black has been added by brush - Precision Dirty Black.



Here's the newly painted signal in the Transport Frame for the first time.




From other angles.


Those who have followed my earlier postings will know that I've been very pleased with the Decals from Chiltern Modelling Services.

These have only been available for Upper Quadrant designs until now.

I'm delighted to say that Keith Wallace of Chiltern has some new Decals available for GWR Lower Quadrant arms.

These decals are made from prototype photographs, but because of discrepancies between prototype and the 4mm scale etches available, some serious Photoshopping has been required!

The two sheets which Keith has sent me are made for specific MSE frets.

I believe similar sheets may be developed for other suppliers etches.


This is what the sheets look like:


And this is what the completed arms come out like.




Before I can put it all together I'll have to install the lighting for the lamps.....



  • Like 9
Link to post
Share on other sites

Great work, as ever.


While you are talking to Keith about transfers, I don't suppose you fancy putting a bid in for sone GNR semphores, too, do you? I seem to spend hours making a very poor job of painting distants...





Link to post
Share on other sites

Great work, as ever.


While you are talking to Keith about transfers, I don't suppose you fancy putting a bid in for sone GNR semphores, too, do you? I seem to spend hours making a very poor job of painting distants...





Hello George,


Thanks for the complement.

I suggest you discuss your requirements with Keith direct, rather than me passing things back and forth.

I'll PM you with his email address, since I'm not sure its fully public.


For anyone interested in these new transfers, they're now available on-line from:


The website has loads of good info. as well.



Link to post
Share on other sites

The Lighting.......

In 4mm scale I always use Fibre Optics to make working signal lamps.

I suppose the system consists of three main components:

1.     The light source

2.     Carrying the light to the lamp

3.     Shining it through the Spectacle.


1. For the last few years I've standardised on "Gaslight" LEDs from Helmsman Electronics as my light source.

The give a gentle light that I think gives a good impression of an oil lamp.

They are long lived, and generate no heat.

Once I've fitted a 1000 ohm resistor and shortish power leads, clearly coloured for polarity, I install the LED in a short length of 1/8th inch ID brass tube.

(Installed is a posh word for "Glued")


This assembly can then be fixed to small piece of plywood or whatever for fixing to the baseboard.

(Hot Glue Gun this time)

The power leads are terminated in a small 2-pin plug.

The matching socket is fitted to another short lead, which I finish with a PP3 connector (9v Battery).

A working voltage from 9 to 12 volt DC is ideal.


2.  I use 0.5mm dia Fibre Optic Cable to carry the light from the LED to the Signal Lamps.

One strand of fibre is required for each lamp.

In this signal there are four lamps, so four fibres.

Starting near a lamp. I've carefully threaded the fibre through a hole into the Doll, and then when it emerges from the bottom, threaded it into the main Signal Post.


Once it comes out of the bottom of the post a length of 200 mm or so is left until all the fibres are installed.

These four fibres are then terminated in a small Brass Plug, which I've turned to fit the LED Tube.

Before threading the fibres into this plug and fixing them with a spot of Cyano, I gently "Plait" them so they support each other.


This Fibre "Rope" is also shrouded in Heat-Shrink tube for further protection. The heat-shrink only being shrunk onto the brass plug.


3.  To shine the light through the Spectacle it is necessary to terminate the Fibre in the model Signal Lamp.

Each White Metal Lamp has been drilled through 0.5mm and the front of this hole counter bored 0.8mm for about 0.5mm.

The lamps were carefully aligned with a spectacle plate when they were soldered to the Lamp Brackets.

The upper end of each fibre is threaded through its Lamp and trimmed square with very sharp cutters.

The end is then expanded with heat from a soldering iron - near but not touching.

This swelling is the Lens of the lamp.

The fibre is eased back so the swelling is inside the cast lamp.

The slack of the fibre is then eased back into the lamp structure.



Connect it all up and this is what you get:



On next to the mechanical assembly….




  • Like 3
Link to post
Share on other sites

They look spot on Steve, and the fibres, although looking a bit odd in the cruel close-ups, blend in rather well at normal viewing distances. I may have to try it out sometime on my 7mm signals.

Those signal motors look good too. I can see me trying 7mm ones on one of mine!

Wonderful stuff. :imsohappy:

Jon F

Link to post
Share on other sites

Thanks Jon,


Yes, the fibres do look quite prominent in photos, but are strangely "invisible" with the naked eye.

Its probably why we always see more faults in photos of our models which we haven't noticed previously.



Link to post
Share on other sites

Connecting the mechanical bits………


Because I ensured the Signal Motors were a uniform distance below their Distant Arms, the operating links were also a uniform length.

I fitted the links between each Arm and its Motor Crank, then secured the crank to the shaft with a tiny spot of Carr’s Solder Paste, using a hot clean Iron and some speed!

To get the crank angle correct I set the Arm below its Clear position and the Crank up to its highest possible point. With the operating wire (on the rear crank) at a similar high position. The operating wire could now pull the rear crank downwards, rotating the front crank with it, which in turn pulled the Arm upwards; first to the proper Clear position, and then on to the Danger position.


The stop Arm was connected to its Weight Bar in a similar way.


All this was done with the signal mounted in the Transport Frame, so now its time to connect up the servos.



Using the GF Controls, I set all the Servos to their “Safe” position, then positioned the Horns on a suitable spline so they were as near horizontal as I could manage.

This “Safe” position is set digitally by the Controller software and is the Servo’s Mid Position.


The lower end of each operating wire is already fixed inside its own 1/32in tube.

This slides smoothly inside its Guide Tube described previously.



A 1/16in tube was then fashioned to fit each Operating Wire to its respective Servo Horn. There is a small hole filed in this outer tube to allow it to be soldered in place.

At this stage, the operating wires are still free to move, sliding inside their connecting tubes.



With the Signal Arms all set to a “mid-position”, and the servos set at “Safe”, they are all in the correct position.

A drop of liquid flux, and a little solder on a Hot Iron is all that’s needed to fix them in place.



You have to look carefully to see the solder…………I didn’t use much.


Next it’s the electrickery …………..





edited for spelling.

Edited by SteveAtBax
  • Like 3
Link to post
Share on other sites

Electrickery …or … Wiring it up…….


Each arm of this signal is moved by its own servo motor.

This in turn is controlled by a GF Controls semaphore controller.

The GF will move its servo when two contacts or connected (move to Danger) or dis-connected (move to Clear).


Because there are four servos, it can all work with four simple On-Off switches.



The user will then have complete control over which signal arms show Danger or Clear at any time, including incorrect displays such as Clear Distants with the Stop arm at Danger, or two or more Distants showing Clear at the same time.



The following colours are used:

White is the common return from all the GF Controllers

Red is the Stop connection

Yellow is the Main Distant connection

Green is the Left Distant connection

Blue is the Right Distant connection.

The Purple and Black wires are for the 12v DC power supply to the GF Controllers.



However, by using multi-pole changeover switches and applying a little logic, it is possible to wire things up so that only allowable combinations of signal arms can be set to Clear.




Starting with the switch for the Stop arm, whenever that is closed to set the stop arm to Danger, it will be necessary to ensure the three Distant arms are also all at Danger.

This is achieved by using 4-pole switch, with each pole connected to a separate servo controller.

This will mean that once the Stop arm has been moved to Clear, all the Distant arms will also be allowed to move to Clear.

They are all prevented from moving by their own switch, which is still closed, and holds the Distant at Danger.



Note the White connection to all the centre switch pins – Common Return

On each switch, the top left pin is connected by its correct coloured cable to its GF controller.

The same colours are used for the interlocking links from each switch.




By now using 2-pole changeover switches for the Distant arms, these can be wired up to ensure that as one Distant switch is opened, the other two Distants are Closed, to hold them at Danger.


By repeating this wiring for each of the three Distant switches, it will be impossible to display any incorrect combination of Clear arms.




I know its bit busy, but hopefully its possible to follow the logic in these photos.


If you've stayed with it so long, you might wish to see this little video of the Signal in operation on my bench.





Edited by SteveAtBax
  • Like 5
Link to post
Share on other sites
  • 4 weeks later...

To answer George's query of 16 January regarding future signal arm transfer plans, there's now a set of Upper Quadrant corrugated arms, with an LNE flavour as it includes discs, shunting signals and spectacle plates.


GCR and LYR signals are the current focus. However I do need to quantify what is on the MSE fret S003, as this is listed as GCR, but looking at it more critically seems to also contain some arms that look suspiciously like GNR somersaults. George, if you could advise which fret you are using, I'll consider GNR for a future sheet.



Link to post
Share on other sites


Why do you not set up the signal controllers so that the open contact sets the signal to stop and the closed contact sets the signal to off.

Not only would it be closer to the prototype 'fail safe' philosophy but it would also greatly simplify the wiring of your switches and avoid the need for multi pole switches.



Link to post
Share on other sites

Hi Keith,


First of all, the choice of "ON" or "OFF" for Danger isn't mine.

Its determined by the maker of the Servo Controllers.


Having said that, some are made with an "Invert" option, making the logic switchable.

The GF 8-way unit has this feature, and the next release of the GF 2-way unit will also have this.


The use of multi-pole switches is necessary to give the interlocking I require.

This cannot be achieved by using simple switches.


If using "On" for Danger, any condition that requires a signal to show Danger uses a switch set to "ON". i.e. Closed.

These switches can all be separate and simply connect the Common Return to the single Switch Feed for the signal concerned. Extra conditions can be added to the logic at any time.


If using "Open" for Danger, it is necessary to link all the switches in series.

That ensures that any switch moved to "OFF" i.e. Open, will break the chain and force the signal to Danger.

In this method, there is only one connection to the Common Return, at one end of each chain, and one connection to the signal's Switch Feed at the other end.


The problem is that chaining all the switches together is more difficult, and adding extra conditions makes this even worse.



Link to post
Share on other sites

To answer George's query of 16 January regarding future signal arm transfer plans, there's now a set of Upper Quadrant corrugated arms, with an LNE flavour as it includes discs, shunting signals and spectacle plates.


GCR and LYR signals are the current focus. However I do need to quantify what is on the MSE fret S003, as this is listed as GCR, but looking at it more critically seems to also contain some arms that look suspiciously like GNR somersaults. George, if you could advise which fret you are using, I'll consider GNR for a future sheet.






Awsome customer service! I am using frets S003/3 and S003/3M (though the latter is perhaps a minority activity..) If you think you could do anything for these you will at least have one customer (if you look at my Loxborough thread you wll see my problem; inability to paint decent distants)





Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.