don9f
Statesman
Les Warnett 9F, Martin Evans “Jinty”, a part built “Austin 7” and now a part built Springbok B1.
Posts: 960
|
Post by don9f on Sept 25, 2018 20:48:07 GMT
Hi everyone, due to various circumstances that have prevailed over the last few months, the overhaul of my 9F locomotive is “on hold”....probably until winter when I fully intend to carry on with it. In the mean time I have started to build a 5 inch gauge, semi-scale, fully functioning model of an LMS/BR upper quadrant semaphore signal. This is to replace one of the colour light signals on our Miniature Railway, to give a more authentic “look” to the signalling in the station area and includes Stop over Distant arms, plus a miniature arm for the Branch to the Steaming Bay etc. The photo is a cropped/enlargement of something I found that shows the type of signal I am making. I am using a few castings from Doug Hewson’s range for the Spectacle Plates and the Lamps, but the rest will be scratch built. I will be using three RC model servos for the movements, with a simple homebuilt servo controller and relays to interface with the existing signal controls that our track uses. If anyone is interested in following a “build thread”, I will be happy to oblige....anyone like to comment? Cheers Don
|
|
|
Post by Roger on Sept 25, 2018 21:41:38 GMT
Hi Don, That's an interesting project and I'm certainly interested to see the electronics side of that. I presume you will have a white LED to shine through the coloured filters too?
Personally I dislike relays, I'd prefer to optically isolate the two parts of the system. As I understand it, RC servos require a repeating series of pulses, the length of which defines the position of the servo. The circuit presumably just gets a binary input which translates to the two different lengths of pulses.
One thing that occurs to me is that if you instantly switch between the two, the signal will respond very quickly which would spoil the scale appearance. Doubtless you have a solution to these things, but in case you haven't, I'd be inclined to do it all in an 8-pin PIC in software. The only components required would be an opto-isolator for the input, the PIC and a voltage regulator with a couple of smoothing capacitors.
Would you power it separately or use the power from the track electrics?
|
|
61962
Seasoned Member
Posts: 129
|
Post by 61962 on Sept 25, 2018 22:59:20 GMT
Hi Don, May I suggest you take a look at the forum on the GL5 website. We built a number of BR signals, initially for use at our shunting demonstration at Shildon and later for use at varioud GL5 rallies around the country forum.gl5.org/viewtopic.php?f=5&t=311 . You will find a lot of useful information on the signals and their construction. On the lamps (for Roger's interest) a number of experiments were carried out to get the lamps right. The green spectical is actually blue and relies in full size on the yellow light from the oil lamp to provide a green aspect. This was an admiralty invention in the nineteen century for navigation lamps on ships. On our signals we ended up using a grain of wheat bulb on a reduced voltage to give the right colour with the blue spectical. Very effective too as you will see in the forum pictures. Eddie
|
|
don9f
Statesman
Les Warnett 9F, Martin Evans “Jinty”, a part built “Austin 7” and now a part built Springbok B1.
Posts: 960
|
Post by don9f on Sept 25, 2018 23:21:44 GMT
Hi Roger, thanks for your interest. I haven’t got them yet but I’m going to try some “Gaslight” LEDs which are available and are used by modellers of signals in smaller gauges....eg O & OO. These LEDs have a yellowish colour, similar to a flame but I think they are only available in 3mm diameter, so not sure if they will work....depends how good my homemade lamp lenses turn out!
Yes the RC servos I will use require a pulse nominally between 1ms & 2ms long repeating about every 20ms, to move them from one end of their travel to the other. I have already built the servo controller which is just four simple timers contained in two 556 ICs, with switchable resistances to vary the pulses. The servo speeds will be “slowed” by some add-on units in each output. I have one already and two more are on the slow boat! These things can be used in a model aircraft to slow down the speed of undercarriage retracts or flaps etc. Nowadays though it can all be done at the transmitter so add-on things are not required.
Our signalling system has a nominal 12v dc supply at each signal and all the controls/interlocking is old fashioned relay logic. You are quite right about using a PIC but the problem there is....I don’t know how to program one....wouldn’t know where to start! On one of the model railway forums in its signalling section, I’ve read about various electronic signal controllers that can control servo operated signals, allowing adjustment of travel, scale like movement of the signal to its “off” (clear) position and scale like bounce as it returns to danger....but this is all a bit pricy and unnecessary for my project. I’m hoping my simple controller will work and it and the servos will be powered by a 12v to 5v regulator off the aforementioned supply.
I’ve actually got most of the signal parts made but need to make the base and figure out the servo layout etc.....Oh and make the ladder which is going to be a bit tedious methinks!
Cheers Don
|
|
don9f
Statesman
Les Warnett 9F, Martin Evans “Jinty”, a part built “Austin 7” and now a part built Springbok B1.
Posts: 960
|
Post by don9f on Sept 25, 2018 23:34:14 GMT
Hi Eddie, thanks very much for your info....I did a lot of research before starting this project and got loads of info from two threads in particular on the “rmweb” forum and bought a book on LMS signals by a certain G Warburton....didn’t see any mention about bread though.
I know what you are saying about bulbs and the lighting, but I haven’t done any testing on that yet. I do have some coloured film, again from a model railway supplier and included is blue, like you describe for use with “yellowish” light, or green for use with “white” light. We only ever run in daylight, so the lighting isn’t critical really.
Cheers Don
|
|
Lisa
Statesman
Posts: 806
|
Post by Lisa on Sept 26, 2018 1:01:02 GMT
You could probably get it to bounce by mounting the relays at the bottom of the post, connected to the arms with flexible wire; if the relay moves quick the arms will naturally bounce a bit, exactly as they do in full size.
Also yes, pic's of your progress would be lovely.
|
|
|
Post by Roger on Sept 26, 2018 5:28:38 GMT
Hi Roger, thanks for your interest. I haven’t got them yet but I’m going to try some “Gaslight” LEDs which are available and are used by modellers of signals in smaller gauges....eg O & OO. These LEDs have a yellowish colour, similar to a flame but I think they are only available in 3mm diameter, so not sure if they will work....depends how good my homemade lamp lenses turn out! Yes the RC servos I will use require a pulse nominally between 1ms & 2ms long repeating about every 20ms, to move them from one end of their travel to the other. I have already built the servo controller which is just four simple timers contained in two 556 ICs, with switchable resistances to vary the pulses. The servo speeds will be “slowed” by some add-on units in each output. I have one already and two more are on the slow boat! These things can be used in a model aircraft to slow down the speed of undercarriage retracts or flaps etc. Nowadays though it can all be done at the transmitter so add-on things are not required. Our signalling system has a nominal 12v dc supply at each signal and all the controls/interlocking is old fashioned relay logic. You are quite right about using a PIC but the problem there is....I don’t know how to program one....wouldn’t know where to start! On one of the model railway forums in its signalling section, I’ve read about various electronic signal controllers that can control servo operated signals, allowing adjustment of travel, scale like movement of the signal to its “off” (clear) position and scale like bounce as it returns to danger....but this is all a bit pricy and unnecessary for my project. I’m hoping my simple controller will work and it and the servos will be powered by a 12v to 5v regulator off the aforementioned supply. I’ve actually got most of the signal parts made but need to make the base and figure out the servo layout etc.....Oh and make the ladder which is going to be a bit tedious methinks! Cheers Don Hi Don, I thought you probably would have this in hand, and there's absolutely nothing wrong with good old 555/6 timers. PICs are great but you need to have the IDC programming module to program and debug them. They're not hard to program but with anything new they look like a nightmare when you first have a go. Things like getting the bounce is easy, but making the end stop positions adjustable takes a little more imagination and a further input button so you can detect the user's input and interpret that as an instruction to change something. All very interesting and needs to be designed just the once. One day I might play and see what can be done.
|
|
chrisb
Part of the e-furniture
Posts: 345
|
Post by chrisb on Sept 26, 2018 9:37:30 GMT
The GL5 signals mentioned above look very good, I saw them at the GL5 rally held at my former club (Lincoln) earlier this year.
|
|
|
Post by chris vine on Sept 26, 2018 12:28:18 GMT
Hi Don,
The BBC Microbit computers are great at controlling servos. You program them in a super simple block language, a bit like jigsaw pieces with options that you type into them. A microbit costs around £15 on ebay.
Then you make the program move the servo at whatever speed/steps you like. If you want it to bounce, you just make the program move it back a bit!!
Chris.
|
|
don9f
Statesman
Les Warnett 9F, Martin Evans “Jinty”, a part built “Austin 7” and now a part built Springbok B1.
Posts: 960
|
Post by don9f on Sept 26, 2018 18:46:26 GMT
Hi Chris, I had never heard of the microbit before....will definitely look into your suggestion.
Thanks Don
|
|
don9f
Statesman
Les Warnett 9F, Martin Evans “Jinty”, a part built “Austin 7” and now a part built Springbok B1.
Posts: 960
|
Post by don9f on Sept 26, 2018 20:02:31 GMT
Ok well my latest project has generated a bit of interest, so here goes with a bit of background and the thinking behind the idea.
I have always had a lot of interest in Railway Signalling but never any professional involvement with this side of things during my Railway career. As a Model Engineer, I always wanted to build some semaphore signals and install and operate them on a miniature railway but as with lots of other things, never got round to it!
The miniature railway I’m involved with these days is a 5 + 3.5 inch gauge raised track that normally operates with two trains running and has a total of eleven signals which are mostly 3 aspect colour light signals that are representative of the real thing. There is a signal box but the system normally operates automatically and trains are “detected” as having crossed from one side of an insulated track joint, to the other near each signal location. It’s not proper track circuiting or axle counting, but it works....except when the rails are a bit rusty, when vigorous wire brushing is called for! The fellow who designed and was closely involved with all this some years ago, is no longer a member and I have offered to become involved with the signalling system’s maintenance etc. and fortunately some very detailed schematic drawings are to hand.
Some of us agreed that to create a more authentic steam era “look” and to create a bit more interest for visitors, the starting signal from the station would be better as a semaphore, rather than the existing colour light. Ahead of this signal is the traverser, which gives access to the “Branch” that leads to/from the steaming bay and workshop area, some distance away. The signal therefore needed a Stop over Distant arrangement, to take the place of the 3 aspect colour light, plus a separate subsidiary signal for the LH diverging route onto the Branch. The signal controls already cater for this arrangement and it is all interlocked with the travereser etc.
So the type of signal chosen is known as a “Running line to loop” signal, with a LH bracket and a start was made looking on-line for any drawings or details to aid construction. The photo shown in the first post above is what it should look like and whilst effort will be made to make something like the real thing, it won’t be an exact scale replica....it’s going to be next to a raised track for one thing!
During my research I found some very excellent threads on building O & OO gauge model railway signals on the “rmweb” forum and thanks to the efforts of the two fellows in question, plus contributions from others, I’ve learned a lot about semaphore signals!
Next time, I’ll show some photos of how I started and where I got the basic dimensions from to enable me to cut metal....
Cheers Don
|
|
weary
Part of the e-furniture
Posts: 290
|
Post by weary on Sept 26, 2018 21:10:08 GMT
Hello Don,
I see that you have already 'cut metal', so this information may be too late to be of use: But,........ Don't know if you are aware of this source, but O.S.Nock wrote a series of articles for Model Engineer in 1940/41 covering Standard Semaphore signals of the various pre-grouping railways, from memory the articles included detailed dimensioned drawings of the arms, spectacle plates, lamps, posts, and typical individual signal layouts.
If you can get access to these Magazines there may be something of interest there too.
Regards, Phil
|
|
don9f
Statesman
Les Warnett 9F, Martin Evans “Jinty”, a part built “Austin 7” and now a part built Springbok B1.
Posts: 960
|
Post by don9f on Sept 27, 2018 17:54:58 GMT
Hi Phil, thanks for that info....I had decided beforehand to build an LMS/BR style of tubular post signal and found various drawings attached by contributors on the rmweb forum I mentioned but to be honest, none of them were very clear so some dimensions couldn’t be determined with certainty. They were only general arrangement type drawings and for details of such things as the bracket (or “Stage Angles” as they are called) you were directed to other drawings, same for all the detail like spindle brackets, lamp brackets etc....which of course were never shown on the forum. Anyway the GA did show some basic dimensions and the layout and spacing of all the drilled holes in the main post and the “doll” for the subsidiary signal, just that some weren’t really clear. A breakthrough came with being recommended the book on LMS signals by G Warburton and some of the same drawings were present in the book, but much clearer and easy to scale off for things not dimensioned. So, armed with enough info to make a start, I began making hardware....first up were the spindle bearing brackets for the semaphore arms. These are available from Doug Hewson’s range but seem to only be suited to flat sided posts, not the tubular ones, so I made my own which would clamp to the posts (or dolls as they can be called). A length of 1/2 inch square brass to make the three brackets was drilled, then milled to leave only half of the holes showing, turned over and the radiussed clamping face milled with a 1/2 inch ball nosed slot drill. Bolt holes were added, spaced as per the post drilling dimensions. Lengths of “bearing” were made and silver soldered into the half-holes referred to earlier and hopefully the photos make sense of this. Incidentally, 1/2 inch steel tube will form the upper part of the posts as real ones are 5 1/2 inches dia. at this point and 6 5/8 dia lower down. Many of the fittings clamped to the posts have similarly radiussed “washers” at the opposite side and fortunately these are also available as castings, cleverly arranged as a block of twelve that easily separate into individual washers and saved a lot of extra fiddly work. They will be visible in later photos. Cheers for now Don
|
|
don9f
Statesman
Les Warnett 9F, Martin Evans “Jinty”, a part built “Austin 7” and now a part built Springbok B1.
Posts: 960
|
Post by don9f on Oct 2, 2018 19:58:51 GMT
Hi, when I was involved with a couple of different Ground Level tracks in Lincolnshire a good few years ago, I had intended to make a signal or two and acquired some Spectacle Plate & Lamp castings from Doug Hewson for that purpose, but never got round to making any. Now some 30 years later, knowing that I would never have thrown such parts away, of course they cannot be found! When I made the spindle brackets described above, I was still waiting for the new Spectacle Plate & Signal Lamp castings to come from the present day supplier of DH items, so refrained from drilling any holes in the posts until I could verify that the Lamp would sit properly behind the Spectacle Glasses in the “On” & “Off” positions of the Signal Arm. I was going to have to make my own Lamp Brackets anyway, as none were available without a long delay, but I still wanted to check the fit before drilling. Once the parts arrived, it was soon verified that there was no problem and with suitable homemade brackets, the Lamps would be in the right places relative to the Spectacles with the posts drilled as per the full size drawing. All that remained then was to settle on the overall height of the finished signal. The design of any real signal could be influenced by many factors of its particular location (sighting, clearances etc.), so post heights and spacing of individual signals carried on them etc. were tailored to suit within a range of “standard” dimensions, but obviously the relationship between spindle bearings / lamp brackets etc. would remain the same. I just chose an arbitrary overall scale height of 27 foot 6 inches which seemed about right and drilled the posts accordingly. The main post of this type and height of signal is made up of two different diameter tubes, the transition occurring just below the point of attachment of the lower part of the “Stage Angles” or in other words....the bracket! But more on that later. I realised part way through drilling the second group of holes in the first photo, that I had gone a bit wrong with my calculations for their spacing from the first group near the top (RH end) and had to fill 3 of them in and re-drill, but in the second photo, all is sorted and by this time, I had assembled a few more bits together with temporary screws, including a first attempt at a Lamp Bracket. Also you can see the curved “washers” that fit round the opposite side of the post, plus of course the Spectacle Plate castings (two different sizes for this signal). Next time I’ll show the making of the bracket assembly for the doll/miniature arm. Cheers for now Don
|
|
|
Post by Deleted on Oct 2, 2018 20:18:44 GMT
I'm enjoying your build Don, very interesting, something different and new to learn.
Keep it up sir
Pete
|
|
don9f
Statesman
Les Warnett 9F, Martin Evans “Jinty”, a part built “Austin 7” and now a part built Springbok B1.
Posts: 960
|
Post by don9f on Oct 2, 2018 20:29:45 GMT
Hi Pete, glad you are enjoying it....it’s not technically very challenging but yes, it’s an interesting diversion. Today I’ve been messing around with the LEDs I'm using in the Lamp castings, plus making acrylic lenses for them. All a bit of trial & error and “eyeball engineering”!
Cheers Don
|
|
don9f
Statesman
Les Warnett 9F, Martin Evans “Jinty”, a part built “Austin 7” and now a part built Springbok B1.
Posts: 960
|
Post by don9f on Oct 8, 2018 20:07:38 GMT
Hi, you may recall that I mentioned that when I started building my signal, I had only been able to find limited dimensional details during my research....a situation that hasn’t really changed since. I have however found a few more photos on line of this particular type of signal, but as is often the case, they don’t really show close-ups of some of the detail I was after. In order to make the actual “Bracket” or “Stage Angle” as it is called, I have had to estimate a few dimensions and try and work out how it’s all put together and how it’s attached to the posts etc. from the aforementioned photos, plus reference to the one reasonably good (but rather small) GA drawing in my book on LMS signals. The basic structure of the Bracket is from 1/4 x 1/4 x 1/16 brass angle which looks about right and in the end, I devised my own method of attaching it to the posts. I was concerned that just using brass angle alone might be a bit flimsy and allow the Doll for the miniature arm to wobble about a bit, so the top part of the Bracket actually includes a piece of 5/8 x 1/4 steel, with the post holes accurately bored in it thus giving good support where it’s needed. The lower part of the Bracket fits to the posts with square pieces, also bored and all the attachments have an 8BA stud passing right through.....hopefully this photo makes this clear:- On this type of signal, movement of the miniature arm on it’s Doll to the left of the main post is achieved via a “Rocker Shaft”, as the operating wire from the Signal Box initially pulls in line with the main post. The relieved areas you may have noticed on the 5/8 x 1/4 steel in the photo above are to accommodate the rectangular plates that carry the bearings for this Rocker Shaft, as seen in the next photos:- Earlier designs of similar signals used various combinations of either multiple cranks and bits of wire, or pulleys, chains and more wire to operate the arms on Bracket Signals, but this Rocker Shaft system is simple and effective and much easier to model in this size! The bearings and crank arms are made from bits of brass, the shaft is 2mm stainless:- Here a cross hole is being drilled for a 1/32” pin.....thanks Roger for introducing me to pcb drills! Finally for tonight, the signal is beginning to take shape, with Lamp Brackets in place, wiring for the Lamps installed, post caps fitted and the main Bracket in place. This will have to come off again for drilling the holes for the angles that support the “Landing”....ie the platform where someone could stand to attend to the Lamp and other Maintenance jobs on the miniature arm....more on this another time! Cheers for now Don
|
|
|
Post by Deleted on Oct 8, 2018 20:36:09 GMT
|
|
don9f
Statesman
Les Warnett 9F, Martin Evans “Jinty”, a part built “Austin 7” and now a part built Springbok B1.
Posts: 960
|
Post by don9f on Oct 8, 2018 21:09:23 GMT
Thanks for the links Pete....I must have looked at hundreds of pictures of signals over the last few weeks and still couldn’t find the detail I was looking for and I don’t suppose there are many (if any) like mine still in existence to go and actually look at these days. Doesn’t matter anyway....I’m happy with the way it’s coming along.
Cheers Don
|
|
don9f
Statesman
Les Warnett 9F, Martin Evans “Jinty”, a part built “Austin 7” and now a part built Springbok B1.
Posts: 960
|
Post by don9f on Oct 17, 2018 23:14:30 GMT
Hi, this update is about the “landing” as I think it’s called, or the platform that someone could stand on safely to work on the Miniature and the Distant signals....attend to lamps, cleaning etc. As with the main bracket assembly, this was something I didn’t really have any dimensions for, so it’s my best guess after scaling the small drawing I have, plus reference to the few photos I’ve found. I believe that the full size design was for stanchions 1” dia. and the handrails 3/4” dia and that they were connected together by some sort of split clamp castings that are often visible in many photos....but never in close-up! Anyway I started with these “clamps” and machined part of a piece of 7/16” square brass bar to 5/32” thick, intending the finished clamps to be 7/32” square. 1.7mm holes were drilled using the milling machine such that I could make four clamps, this photo probably shows this better than my description:- One end of each clamp needed to be rounded and this was done to each one in turn, whilst it remained on the stock....again the photo shows what I mean and how it was (crudely) done, but the finished result was as intended:- The stanchions were cut from 3/32” round steel and one end of each was flattened in the vice, then a 1.3mm hole was drilled for a 12ba screw. The other ends were turned down and threaded 10ba:- A rummage in an old drawer full of model aircraft “leftovers” produced some linkage hardware that just happened to be .066” dia....perfect scale 3/4”! Fortunately there was enough in the drawer, as it took four attempts to get the shape of this one-piece handrail right:- After assembly to its angle frame support and with a bit of paint, this is how it looks on the signal....I’ve still got to make the wooden floorboards yet but I’m quite pleased with it so far and think it looks the part. Access on to this type of landing would be via the gap in the handrail that is to the side of the ladder at the rear of the main signal post:- Cheers Don
|
|