Steve Hewitt

The Main Stem (just brass tube at present) Trimmers and one Doll assembled:

The bracket work will be assembled from 1mm x 1mm brass angle, and is supported on a short length of channel section where it meets the main post or stem.

To locate the channel section closely on the stem I machined the webs of the channel section as shown, using a 3mm End Mill in my miniature Mill/Drill machine.

This ensures a good fit, and that the channel section is automatically square to the main stem.

By keeping the channel over length, it is easy to get its position correct, and hold it for soldering:

Still over-length, but nearer to the finished size:

To ensure the main parts of the bracket are identical, two lengths of angle are soldered together back to back before shaping with saw and files:

With the "vee" cuts made and the angles bent, the two parts were separated and fitted as shown.

The minor bracket components were added over-length again before cutting back to size.

(By now the second Doll has been fitted).

The supports for the wooden staging, complete with stanchions for the handrail are made like this:

from a length of 1 mm x 0.5 mm brass and my favourite Nickelled pins.

Added to the trimmers:

Finally the bearings for the balance arms are made from shaped square brass tube, and fitted to the main stem at the appopriate height:

The next stage is the signal's "foundations", but with the discussions about working or fixed arms having started, that is all on hold until I agree the way forward at the Manchester show this weekend.

More will follow...

Steve.

Hello,

How do you do the see through glazing on the arm spectacles in 4mm please?

trustytrev.

I use the appropriate glazing material from Model Signal Engineering.

For round spectacles, I use a "leather punch" to cut circles of the correct size.

For other shapes, I scribe the inside of the spectacle onto the glazing sheet, using a very sharp scriber.

This is then trimmed to size, a little at a time, using a small pair of needlework shears.

The glazing is secured in the spectacle with "Canopy Glue" from a model aircraft shop.

Steve.

This is a great thread and I look forward to more regarding the construction and operation of model signals.

[Tongue in cheek] Have you thought about offering your services to produce a Right Track video about building model signals and the methods of operating them. I'd buy it. :-)

So would I !!!!

JF

Thanks for your comments.

I haven't thought of a video, but Phil Atkinson of Hobby Holidays has asked me to run a weekend workshop on Signal Construction and Operation for him.

The dates will be April 6th & 7th, 2013 if you're interested.

Steve.

Thanks for pointing that out. I'd missed it.

Steve.

Hi Mike,

I'll see you at the Manchester show all being well, and we can agree the spec for this and any other signals David will require.

The model is a bit further advanced than the stage I've documented on here.

I'm away from home at present, and won't be in the workshop for a few days.

Just spent the day in "Steam" at Swindon.

Not my favourite company, but the Museum is well presented, all the displays work and its not too expensive at £4.40 for oldies like me.

Steve.

Hi Steve,

 

I don't know if you've noticed but both the distant arms are fixed (and the left hand one doesn't have a green spectacle) hence the inter-arm gap is smaller than normal

 

Bog Junction was always an excuse for some toilet humour, especially as the base has been repaired / strengthened to make it as strong as a brick shi house

Thanks for pointing that out. I'd missed it.

Steve.

Next the dolls.

Both are almost identical, just the diameter of the lower sleeving is less for the left hand doll, which will fit inside the top of the main stem.

As shown this is straight off the assembly jig which I use to get the arm bearings and lamp brackets in the correct positions.

Bearing length will be reduced considerably at a later stage.

More later....

Steve.

The operation sequence for both the Station and the Sector Plate have been stored in the spreadsheets and printed out for checking.

John and I spent today operating the whole sequence to check its accuracy.

A few corrections were of course necessary, and a couple of additions where one train hadn't arrived back in its starting position by the time we'd performed all the instructions!

At its busiest, there were some seven trains in the station simultaneously, occupying six platforms and one siding.

These snaps might give an impression:

Looking over the "Back Road" to trains in Platforms 1, 2, 3 and 4.

Looking the other way at Platforms 9, 8 (empty), 7 and Siding D adjacent to the Roof Pillars.

Platforms 10 and 11 are off-scene to the lower right.

Next step is to print out the instructions on the correct index cards.

This will then allow two operators to control the station at the same time, and we'll see how well balanced the workload is for the operators.

Fingers crossed we don't have to make too many alterations before we can "freeze" the schedule and start proper training sessions for the Warley show.

Steve.

Time to start another signal.

This is for a very large model of Carlisle, to which I've contributed a few signals in the last few years.

The owner has provided this picture of the signal required:

Its a late style LMS bracket, with low mounted arms for sighting requirements.

The main stem is tubular steel, as are the signal dolls.

The model is to be servo operated with working lights.

Having drawn a schematic of the signal at 4mm to 1ft scale, and with frequent reference to Graham Warburton's excellent book on LMS Signals, I made a start:

First the trimmers with the sockets to take the dolls and the two brackets in which the "rocking shafts" will revolve:

The two sockets were turned from Brass bar and also ensure the two channel sections are correctly spaced and parallel.

The hole in each is drilled 1/8in to take the main stem and the sleeved-up bottom of the second doll.

The first doll will eventually be located in the top of the main stem.

This shot of the trimmers upside down shows the crude representation of the prototype socket, which in real life is a welded assembly of plates and tube with fins giving the taped profile.

The brackets to support the Rocking Shafts or Torsion Bars are7mm scale "Medium Handrail Knobs" opened out to 1/32in dia. to take the brass tube I'll eventually use for the shafts.

To facilitate the assembly of these components I used an off-cut of MDF, drilled 1/8 at the correct spacing with two drills in the holes.

The sockets were placed over the drills and the trimmers lined up and held for soldering.

More to follow.....

Steve.

John and I spent the whole of yesterday trying out a new operating sequence for Lime Street.

I had drafted out the operations for the station controllers, and these were corrected as necessary as we progressed.

The strategy (if its not too grand a title) was to make all the sector plate operations sub-servient to the needs of the station. (Obvious, really!)

Each operation at that end of the layout was documented on pre-prepared sheets as we progressed through the sequence.

I'm now in the process of transcribing these to a MS Excel spreadhseet, where an operation (each of up to six distinct moves) is represented in one Row.

These can then be printed onto the required Instruction Cards (roughly 6" x 4") using a Mail Merge in MS Word.

We intend to check all the documentation out next week, after which we should be able to start Operator Training ready for our appearance at the Warley show at the NEC.

We all hope to see many of you there, so please drop by and make yourselves known.

Steve.

Excellent , is the "bounce" done by software?

Yes, is the quick answer.

A servo arm (horn) moves through an arc.

The position of the arm is determined by the bandwidth of the control signal.

The extremes of the position are set with a bandwidth of 1msec at one end, and 2msec at the other IIRC.

The control signal for controlling semaphore signals comes from a dedicated servo controller, such as the MERG Servo4, the Embedded Controls units or in this case the GF Controls unit.

These all use software to generate the control signal.

The speed at which the servo moves is determined by the speed at which the bandwidth is changed.

Adjustment of the controller is typically limited to changing the two end positions i.e. setting the bandwidth for the Danger and Clear positions of the signal.

The more sophisticated the software, the more realistic the movement of the signal arm can become.

See my topic on servos and controllers for more detail: http://www.rmweb.co....vo-controllers/

Hope this is helpful,

Steve.

Here at last are the video clips.

First the Starter Signal:

This is the Bracket Signal:

Finally the Ground Signal:

You can see the operating wire more clearly in this clip than in my still shot!

This video is a bit cruel, as the whole thing is only just over 1/2 inch tall.

Steve.

First a shot which shows (just about) how the operating wire is routed:

I've taken some video, but its taking for ever to process it onto YouTube. !!*f--***

Steve.

Everything got a coat of Halford's White Primer, then the main signal body was finished in Halford's Satin Black.

The arm was held by its shaft in a pin vice for all the painting process.

The overspray just washes off with a soak in Cellulose Thinners when its all complete.

At this stage I'd forgotten that the signal may be "passed at danger on routes to which it doesn't apply".

i.e. it should be Yellow not Red!

Having the main items painted, I could tackle the final assembly.

The Blinder/Crank was beefed up where it will fix to the arm shaft:

Then the operating wire could be fixed.

This passes up through the base plate to the balance arm, a right angle bend each side allow it to continue up to the Blinder/Crank.

Another right angle bend allows it to passes through the Crank and be terminated with a pinch of the pliers.

Once the arm and crank are correctly positioned, a quick touch of solder on a hot iron fixes the crank to the shaft, trim off excess shaft etc. and everything is complete above ground.

Moving the operating wire upwards, raises the balance arm and turns the crank, and vice versa.

Unfortunately, I've failed to get a clear shot of this.

Must try harder next time.

At this stage my error was pointed out to me, and a repaint resulted in this:

The servo was mounted "below ground" in such a way as to avoid underground obstructions indicated on a template from the layout owner.

I'll get a video clip of this and the other two signals as soon as possible.

Steve.

Latest from John, he's got the basic painting and wood staining done.

 

 

 

Ready for Rob to work his magic .........

 

Steve.

And here we are.

Rob returned the turntable deck to John at the Lancashire show last weekend, and John has sent me these few shots of it back in place on the layout:

Steve.

Now for the Ground Signal.

This is to be a single arm version and is based on the etch for the two arm version from Scale Signal Supply.

This is a link to an image of the etch I used when building the signal for Liverpool Lime Street: http://www.rmweb.co....ttach_id=101648

The double lamp housing was made first - a simple fold up of the etching:

The prototype single arm signal has casting which raises the lamp housing above the balance bearing to the height required.

This I represented with a small length of square section brass to which two wing will be attached later to give the characteristic profile. I turned each end of the brass section to allow it to fit into both the lamp housing and the bearing housing:

The lamp housing was centred and drilled with the 1mm dia hole needed:

Cut off the surplus lamp housing, then when its all soldered together you get:

The rear part of the arm etch is attached to its bearing shaft:

After polishing, the half etched front of the arm is attached:

The balance arm had its holes re-inforced to ensure smooth operation and longevity:

So here are the main components:

The back-blinder / crank has been left on the etch for now, until final assembly.

If it escapes there's no way I'll find it!

Here the "wings" have been added to the "main casting":

The signal has now been attached on its base plate and foundation tube - 5/16th in. dia this time:

Clean polish and painting next......

Steve.

Sorry if this a bit Off Topic....

Although you'll have to wait another year to see Liverpool Lime Street at the Lancashire Show, many of the team will be there all this coming weekend:

John and Brian will be demonstrating; Les is the Exhibition Manager, John R. is the Treasurer and Rob and I will be around most of the time, and Spike will be operating on Blackmill.

So there will be 7 of the Lime Street team at the show, all willing to chat to you about the layout and answer any questions you might have.

Interested? Come along to Accrington and visit the 47th Lancashire Show http://www.rmweb.co....september-2012/

Steve.

Hi Jon,

 

Do you think this problem could be due to the fact that on boot-up, servos tend to centre their positions (ie going clock wise; anti clockwise; stop)? If you've got 25 all doing this at the same time, this would represent quite a significant current draw to the system, and possibly accounting for your transformer shutting down?? The answer might be to 'switch in' smaller groups of servos after boot up of the system, thus avoiding this large current drain on boot-up?? Once boot-up is complete, servos will just sit there with very little maintenance current draw, and thus the system can cope with this - no problem.

 

Hope this is helpful, and not a case of me trying to teach grandmother to suck eggs!!

 

Best wishes

 

Mike

There has been much discussion in the past about problems with "Servo Twitch" on start-up.

To avoid this, the controllers used on Liverpool Lime Street ensure that the "position" signal is sent out before the power supply is connected.

Being entirely digital, the position of the servo at last use is stored in the controller.

This seems to work, and Lime Street has currently 32 servos driving semaphore signals.

The same technology (software) is used in the commercial version of these controllers, from GF Controls.

Steve.

OK. Back from the paint shop and final assembly has been done.

First the Starter:

In this shot you can see the signal in its test / transport rig:

The main piece of plywood represents the baseboard at the signal location.

The thinner peice of plywood carries the servo.

The alignment of these is ensured by the 1/2in. dia brass tube fitted to the base of the signal. (See earlier posts)

This method allows the signal to be delivered fully assembled.

The owner can then test it out, and when happy with it all, can dismantle it and repeat the installation on his railway.

Next the Bracket Signal:

I'll try to post some video later.

Next job is the Miniature Arm Ground Signal.....

Steve.

A bit more on the use of a Stepper Motor to drive the Sector Plate.

This short video - very wobbly and shaky - shows:

First the sector plate on the move. A captive nut beneath the deck is pulled along the revolving screwed bar.

Next you can see the drive train, the large motor with toothed belt and the casing for the processor which controls the motor.

The shot ends with Sector Plate Road 3 aligned with Fiddle Yard Road 5.

Ease of operation is shown with shots of the Control Panel.

The illuminated LEDs in the track plan show the previous route set (Roads 3 to 5 above). This also indicates that traction power is connected to those tracks.

The large illuminated button is pressed to cancel the previous route, and Sector Plate Road 4 and Down Fast are selected.

The shot concludes with the sector plate aligning as required and a close up indicates the accuracy of alignment.

Steve.

Hi Steve

 

Are the platforms scale length. 45 feet overall seems very short for such a large station? Also out of curiosity where were the larger locomotives turned?

 

Cheers

 

Jim

Hi Jim,

I believe the platforms are scale length for the period we are modelling, which is 1947-ish, post War and immediately prior to Nationalisation.

There is much very relevant information in the article by Reg Instone and Graham Warburton in LMS Journal no. 36 about Lime Street re-signalling. The LMS were very concerned about the problems at Lime Street before the War with LMS minutes of 1938 referring "considerable inconvenience was caused in working traffic due to the station platforms not being long enough to accommodate the trains requiring to use them".

The solution, extension of some platforms, took place in the period immediately following that modelled, when the signal box was replaced with a new power box located adjacent to the turntable. This was brought into service on 25th January 1948, and the old signal box (the one on our model) was demolished and the track re-laid over its site and in use from 29th February, 1948.

At the time of our model, larger locomotives had to travel up to Edge Hill to be turned.

Adjacent to the turntable on our model is the outline of the new larger installation, which we will show as a development site.

Steve.

The turntables and sector plate use "Stepper Motor" technology.

Some details have been posted earlier in the topic.

You might find them if you use the Search facility at the top of each page.

 

The turntable mechanism was built by team member Geoff, and is now commercially available from GF Controls http://www.gfcontrols.co.uk/.

 

I'll try to get some video of it this weekend, both above and below ground.

 

Steve.

Thanks Mick, for your comments.

Here's a short viddeo I took at John's place today:

 

http://youtu.be/JB_ygsKzDdg

 

I'll try to get some more shots of the mechanism, and also the versions used on the Sector Plate and its own high-speed turntable.

 

Steve.

And here is a very low quality video clip of the mechanism from below:

http://youtu.be/iAsX2dIwU6E

If you look carefuly at the close-up you can see the double reduction Toothed Belt drive from the stepper motor itself to the turntable's drive.

This gives a resolution of 7200 steps per revolution of the turntable deck.

In other words, to move the deck as far as the minute hand of your watch moves in one minute, there are 120 separate addressable steps.

Alternatively, its 20 steps per degree of angle.

Steve.

A bit more progress.

First the balance arm bearing was fitted to the post:

The lamps glued (again) in place, having checked the alignment with the spectacles for the upteenth time:

Almost ready for the paintshop:

Everything had to be scrubbed and polished, odd bits of surplus solder etc.scraped off and so on, before a final soaking in Cellulose Thinners.

Once that was dry it was outside on a warm dry day with the Halfords White Primer.

A hair dry to hand helps the several thin coats to dry quickly.

Resisting temptation, all stored securely for 24hrs to harden before adding the black details etc.

Steve.

Those etched cranks from Masokits look like they'd be good for 7mm scale. Thats if they're not too fine for my "sausage" fingers to handle!

JF

I'd had similar thoughts Jon.

I'm sure you'd manage them. They are well thought out as you'd expect.

Each crank is folded up from two halves, with a brass tube through the pivot to ensure the halves line up.

Once soldered up the crank is then mounted on the next size down brass tube firmly fixed in place.

This gives the pivot function.

A wire through the fixed tube, bent over at the end prevents the crank coming off its pivot.

That's the instructions in words, if I remember correctly.

Its much easier to understand if you see the diagram on the instructions!

I think the cranks are sized and shaped for wire runs in 4mm scale and are definitely intended to be functional.

Well worth a go in 7mm scale for smaller cranks, if like me you're attracted by well designed bits and pieces.

Steve.

The dolls for this signal are white metal castings from the MSE range.

Each was prepared with bearing, lamp bracket and final:

The cast brackets from Scale Signal Supply were fitted:

Each doll has a socket from square section tube fitted between the trimmers:

The supports for the staging were also added.

To get the alignment of the dolls correct I laid the signal on its back, supporting it on its main post and trimmers, allowing for the thickness. Having made the arm bearings over length, these were carefully reduced in length until the dolls were perpendicular to the trimmers. A thick superglue smeared round the base of each doll secured them in place. I don't normally like gluing things, but soldering looked like being difficult.

The result starts to take shape:

Next came the cranks for the push-rods.

Push rods rather than pull wires were used on lower quadrant signals to transfer the movement from the balance arm to the signal arms.

The operating length of the cranks is much shorter than those used for pull wires.

My estimation is about 6in. operating radius, as opposed to 12in for a typical wire crank.

Another Masokits etch was supplied for the cranks.

The sophisticated design would make delightful working cranks I'm sure, but they were all too large for my requirements.

I resorted to an Alan Gibson etch from my stock box.

The result is shown here alongside the Masokits etch for comparison:

Here are the four cranks stacked up on a pin, which is the type I'll use for their pivots:

The first pair of cranks, for the left hand doll, with their connecting puch rod:

The completed set of cranks. These can be removed as a sub-assembly when the signal goes to the paint shop:

More soon.

Steve.