Help sought to check locking design

Discussions of the prototypes and how to model them. Show us how you do it.
John Palmer
Posts: 447
Joined: Fri Jul 15, 2011 11:09 pm

Help sought to check locking design

Postby John Palmer » Mon Jan 07, 2013 3:30 am

Here is something on which I would appreciate the erudition of fellow subscribers to this forum.

The lever frame I have been constructing has almost reached the point at which I can start to install the locking. The plan is for this to be accomplished by direct lever locking in a two-channel box, with each channel accommodating five bars. I’ve already devised both the locking table and a dog chart to implement it, and have checked as best I can that my design will supply the locking required. However, I’m acutely conscious that I may have overlooked a required lock, or introduced unnecessary locks that will prevent proper operation of the layout.

What I’m hoping, therefore, is that some kind person will take a look at the design and bring any such errors to my attention. I can supply diagram, locking table and dog chart, together with the spreadsheet I used as an aid when checking whether I had implemented all the locks required. I think I’ve done most of work required to get this right, but believe that it needs a fresh pair of eyes to pick out the errors to which I’ve blinded myself.

At the moment the dog chart is in CAD format with separate layers for each bar position, as this made it easier to check the design of the locking as it progressed. If anyone fancies taking a look at this within a CAD package I’ll be glad to discuss file format preferences (e.g. I should be able to export it in .dxf format). Otherwise I can generate a jpeg of the whole dog chart.

To give you an idea of what's given rise to this enquiry, here is a picture of the frame in its current state, with the lockbox mounted on brackets on its nearside.
Attachments
Frame.jpg

User avatar
grovenor-2685
Posts: 3032
Joined: Sun Jun 29, 2008 8:02 pm

Re: Help sought to check locking design

Postby grovenor-2685 » Mon Jan 07, 2013 8:30 am

John,
Email me the docs and i will go through it. .dxf should be fine, but best to send the .jpg as well, sometimes the CAD programmes don't translate properly.
Use keith@norgrove.me.uk
Regards
Keith (FIRSE if that means anything to you!)

John Palmer
Posts: 447
Joined: Fri Jul 15, 2011 11:09 pm

Re: Help sought to check locking design

Postby John Palmer » Mon Jan 07, 2013 10:05 am

Keith, very many thanks - email winging its way to you. FIRSE does indeed mean something to me; can't do better than get a professional to point out the error of your ways!

JFS
Posts: 625
Joined: Wed Dec 16, 2009 3:47 pm

Re: Help sought to check locking design

Postby JFS » Sat Apr 06, 2013 6:17 pm

John Palmer wrote:The lever frame I have been constructing has almost reached the point at which I can start to install the locking.



So what is the news John? I am itching to see the outcome - not least to know how you will get 18 levers of locking into just 10 channels.

Please keep us posted!

Best wishes,

Howard.

John Palmer
Posts: 447
Joined: Fri Jul 15, 2011 11:09 pm

Re: Help sought to check locking design

Postby John Palmer » Tue Apr 09, 2013 7:40 am

Hello, Howard; bit of a hiatus, I’m sorry to say, as I have had to give my hands a break to recover somewhat from repetitive strain injury, and have also been devoting some time to a Nu-Cast GNR tender for the first of my K2’s.

I believe I now have a workable locking scheme for my proposed layout, which is to bear an uncanny resemblance to Mallaig, although, in my alternate reality, it will be located some miles to the south in Moidart. I am indebted to Keith for his assistance in checking the arrangement and suggesting amendments, but any remaining errors there may be are ‘all me own work’.

Here is the final version of the diagram and locking table.
Aonach SB Diagram revised.jpg

And here the dog chart.
Aonach Dog Chart revised.jpg

As you imply, getting 18 (well, 17) levers of locking into 10 channels is quite a challenge. As Keith pointed out, my original design entailed some duplication of the signal locking. I have managed to eradicate some of these duplicates, although some may remain. They should not inhibit any legitimate movements.

So far as possible my design incorporates stock sizes of bar in order to minimise the amount of cutting required, and I overlooked the need for some 2mm X 6mm bar from which to make the lock dogs themselves. Although I now have a length of this to hand, it’s looking awfully short to cover all the dogs required, so a repeat order may well prove necessary.

As you can see, the lockbars play a role in holding the road for movements in the Down direction. I have been considering the use of light sensors with a short time delay to simulate this on the model, but this may be another instance in which I shall need the help of others, as electronics are not my strong suit.

JFS
Posts: 625
Joined: Wed Dec 16, 2009 3:47 pm

Re: Help sought to check locking design

Postby JFS » Tue Apr 09, 2013 5:55 pm

John,

Many thanks indeed for publishing this - it certainly looks pretty splendid and thoroughly thought through!
At a very quick glance all looks present and correct and I am pleased to see an Advanced Starter and Outer Home which simplify operations massively.

The only thing I can't see is any sequential locking between 4 and 2/3 or 16/17/10/14 and 18 - perhaps you are doing that electrically (or leaving it off)

I am looking forward to looking through a few things and will report back when I am getting somewhere.

One question - how are your levers held over? I can't see any catch handles - are you using an over-centre spring?

I posted some pics of my own efforts here:-

viewtopic.php?f=4&t=1772&start=50 (Scroll to bottom of page)

You might be interested in some mechanical signal box simulation programmes which I have developed - http://www.blockpostsoftware.co.uk Exeter West is certainly one way of learning about Locking!!

Best wishes,

Howard.

User avatar
grovenor-2685
Posts: 3032
Joined: Sun Jun 29, 2008 8:02 pm

Re: Help sought to check locking design

Postby grovenor-2685 » Tue Apr 09, 2013 6:49 pm

I don't see much if any benefit from sequential locking between just two signals in a single line terminus, the block working will be controlled by the token/staff so there is really nothing for the sequential locking to protect against.
To simulate a train sitting on the lockbar preventing the bar being reversed the simplest would be electric locking from a light sensor as you suggest. At MERG we have a design for a sensor using Light dependent resistors that would work well in this application, one board will do 3 detectors, (one for each bar) and could link directly to a servo driver.
Its very simple to lock levers with the SG90 micro servos, as the servo has a 90 degree rotary movement just fit a small length of brass tube on the spindle and cut a slot across the diameter, arrange for the lever tail or an extension of the tappet to pass through this slot when the lever moves, any rotation of the servo will block the slot and thus lock the lever. Mechanically simple and does not impose any load on the servo when it is moving.
Regards
Keith

John Palmer
Posts: 447
Joined: Fri Jul 15, 2011 11:09 pm

Re: Help sought to check locking design

Postby John Palmer » Tue Apr 09, 2013 11:35 pm

Howard, the diagram is in fact a near-exact copy of that for Mallaig, which features both Outer Home and Advanced Starting signals. From memory, all the passing loops on the Mallaig Extension featured only Home and Starter, but I agree that it would be hard to do without the two additional signals at the terminus.

The layout shown is more restrictive than at Mallaig in that there is only sufficient room for two 61’ vehicles plus engine to stand clear of the lockbar on 12 points but still in rear of the Advanced Starting signal. Thus, shunting a train of any length into the Up side roads is going to require a movement into the section in advance. Such are the compromises forced upon me by space considerations.

Actually, it never occurred to me to consider the inclusion of any sequential locking, as instinctively I felt that no worthwhile purpose would be served by this – I’m pleased to find support from Keith for this assessment.

I am in awe both of your table locked frame and the work you have put into rodding drive arrangements, but I know my limitations and am planning mechanical drive arrangements based upon under-board steel rods 3mm in diameter.

The kind of slotting arrangement described in Keith’s post is exactly what I had in mind not only to simulate the lockbars but also to control the ground frame release.

I claim no credit for the underlying design of my lever frame, as it’s an adaptation of a design published by Bill Strickland in editions of MRN dating from the late ‘sixties. The difference is that Bill was able to adapt thick walled brass pipe he was fortunate to have to hand in order to fabricate his quadrants, whereas I had to cut them from 1.2 mm thickness brass sheet before boring and turning them to finished shape.

I have always thought Bill’s was an elegant and ingenious design, and had previously constructed two 4-lever frames for our Burnham layout along the same lines. It does indeed depend upon a form of spring loading, and can perhaps be best illustrated by the attached cross section. What this unfortunately does not show is the slotting of the lever where the pivot shaft passes through it. This slot accommodates the downwards displacement of the entire lever when an operator presses its cap to disengage a stud through the lever from a detent in the underside of the quadrant, in which it is held by the upward pressure of the spring operating at the lever’s base. Once disengaged from the detent, the spring continues to press this stud against the ventral face of the quadrant until the lever reaches its opposite end of stroke. At this point the spring-loading of the lever forces it upwards so that the stud engages in the other end-of-stroke detent and thus holds the lever positively in place.
Fram_X_section.jpg
Fram_X_section.jpg (80.63 KiB) Viewed 4201 times

In my book the big advantage of this arrangement is that it provides a robust alternative to fiddling about with small and potentially vulnerable catch handles. The action of reversing a lever involves slight fingertip pressure to disengage, pulling across the quadrant and allowing the spring to click the lever into its reverse position – very ergonomically satisfying!

Quite a bit of machining and jig-making went into construction of a frame this size, and I took photographs of various stages of the beast’s fabrication which I hope to work into some sort of article when the project is complete.

JFS
Posts: 625
Joined: Wed Dec 16, 2009 3:47 pm

Re: Help sought to check locking design

Postby JFS » Wed Apr 10, 2013 7:59 am

John,

Looks like a very good solution with lots of robustness! I can imagine that there was a fair amount of work in making all the bits also!

I also agree about sequential locking not adding much - I did say I was look hard for things to find!!

I still have not yet had chance to have a proper look through but am looking forward to doing so.

Best wishes,

Howard.

User avatar
grovenor-2685
Posts: 3032
Joined: Sun Jun 29, 2008 8:02 pm

Re: Help sought to check locking design

Postby grovenor-2685 » Wed Apr 10, 2013 9:30 am

upon under-board steel rods 3mm in diameter.

That does seem like considerable overkill, I use standard model aircraft control rods for this purpose, about 1.5 mm dia. They have been perfectly satisfactory on runs up to a metre and have the advantage that the rods, cranks and clevis are all off the shelf.
Keith
locoyard-3.jpg

IanSpalding
Posts: 9
Joined: Tue Feb 08, 2011 8:54 am

Re: Help sought to check locking design

Postby IanSpalding » Wed Apr 10, 2013 10:43 am

grovenor-2685 wrote:To simulate a train sitting on the lockbar preventing the bar being reversed the simplest would be electric locking from a light sensor as you suggest. At MERG we have a design for a sensor using Light dependent resistors that would work well in this application, one board will do 3 detectors, (one for each bar) and could link directly to a servo driver.

Keith,
Which Merg kit (or design) are you referring to?
Regards,
Ian Spalding
M600

User avatar
grovenor-2685
Posts: 3032
Joined: Sun Jun 29, 2008 8:02 pm

Re: Help sought to check locking design

Postby grovenor-2685 » Wed Apr 10, 2013 2:08 pm

Technical Bulletin T33/25 http://merg.org.uk/forum/merg-tbs.php#SuperBloc
NB Only MERG members will be able to use the link.Anyone else interested PM me.
Keith

John Palmer
Posts: 447
Joined: Fri Jul 15, 2011 11:09 pm

Re: Help sought to check locking design

Postby John Palmer » Wed Apr 10, 2013 4:09 pm

grovenor-2685 wrote:
upon under-board steel rods 3mm in diameter.

That does seem like considerable overkill, I use standard model aircraft control rods for this purpose, about 1.5 mm dia. They have been perfectly satisfactory on runs up to a metre and have the advantage that the rods, cranks and clevis are all off the shelf.
Keith

I must say I was quite attracted by the ready availability of clevis joints, etc, by sourcing from model aircraft suppliers. Against this, I found a cheap source of 3mm dia. rod and opted to indulge my propensity for over-engineering things! Bear in mind that I am envisaging rodding drives over distances in excess of 2 metres, with compensators/cranks that will put a compression loading onto half the drive length, and will be transmitting the drive across board joints by sprung plungers. Taking these into account, I thought it prudent to adopt something with a bit less whip than 1.5 mm dia., through rod price was the dominating factor in my decision.


Return to “Signals and Control Systems”

Who is online

Users browsing this forum: No registered users and 2 guests