Jump to content
 

Building an automated sector plate


MattWallace

Recommended Posts

Hi all,

 

My latest layout plan requires a number of sector plates in the fiddle yard to allow for decent length trains *and* the ability for the loco to run around the freight services.

 

The layout plan looks like this:

 

13979153413_5de3f5c106.jpgOilTerminalSectorPlate by ProfFalken, on Flickr

 

and a close up of the far right of the fiddleyard shows the three sector plates:

 

13976311632_8f96c5046a.jpgtraverseransector by ProfFalken, on Flickr

 

I'm going to post to this topic in three parts:

 

  1. Prototyping the design using foam-core board
  2. Building a second prototype out of plywood (the materials that will be used for the final construction)
  3. Building the sector plates into the traverser table

Thanks for watching and please provide any comments! :)

 

Matt

Link to post
Share on other sites

Part one: Prototyping in Foam Core Board
 
I love foam-core board.  It's strong enough to simulate thin plywood, cheap enough to not worry about throwing it away and much easier to cut.  This combination allows me to prototype things without needing to get the workbench and the jigsaw out, so it keeps my wife happy too!
 
======== INSTRUCTIONS ========
 
The first thing I did was print out in 1:1 scale the track layout for a single sector plate in XTrckCad and glue it with prittstick onto a piece of board:
 
13993262503_3d65bbd687.jpgUntitled by ProfFalken, on Flickr
 
The second step was to cut out the track area:
 
13970085771_d39b46ddb7.jpgUntitled by ProfFalken, on Flickr
 
Once that was done, I used a circular cutter (like a compass but with a blade instead of a pencil) to cut the curve for the sector
 
13993255443_45f7b5efcc.jpgUntitled by ProfFalken, on Flickr
 
I then marked the centre lines in Biro of the "middle" track and sector plate to help with lining up later on
 
13993251473_ecf8f381bd.jpgUntitled by ProfFalken, on Flickr
 
I then measured 2cm either side of the track centre on the sector plate to give me a reasonable width shoulder either side of the track
 
13973253895_70a89989a8.jpgUntitled by ProfFalken, on Flickr
 
And cut it out:
 
13973249065_f8e45309eb.jpgUntitled by ProfFalken, on Flickr
 
At this point, I had a sector plate and the run-in track:
 
13973242005_b49d9de9ca.jpgUntitled by ProfFalken, on Flickr
 
I used double-sided tape to secure the stationary part of the sector plate to a cutoff from the foam core board and a small screw from a 9g TowerPro Servo to secure the sector plate:
 
13973235855_2f3f311a5d.jpgUntitled by ProfFalken, on Flickr
 
I then used a paperclip to provide some method of operating the plate
 
13970042632_52b11a9017.jpgUntitled by ProfFalken, on Flickr
 
NB: This was the point at which I realised that you need to change the curve of the sector plate slightly to allow it to traverse across the arc correctly.  it's well worth spending time adjusting this if you are going to give this a go as it prevents sticking!
 
Test the length of the plate (no, this wasn't the first test, I know that a Bachmann 66 is just under 1ft so that's the length of the track in XTrckCad. I also tested it after the printing of the plan and before sticking to the board!)
 
13973224635_fc8628b13c.jpgUntitled by ProfFalken, on Flickr
 
Start to cut the track out:
 
NB: I chose to cut the track slightly biased toward the static part of the plate (i.e. a very small overhang on the sector plate itself) to provide an extra little bit of support.
 
13993211983_e74d89be3c.jpgUntitled by ProfFalken, on Flickr
 
13973651954_c481756aaf.jpgUntitled by ProfFalken, on Flickr
 
Affix the track with double-sided tape so the track can be reused:
 
13973635234_99f3fde04e.jpgUntitled by ProfFalken, on Flickr
 
========================
 
The final implementation of this setup will be powered by a servo to allow for full automation of the layout.
 
I had an old "servo sweeper" script that I installed on an Arduino and connected up a 9g servo to the paperclip to see if there was enough torque.
 
Surprisingly, there is, although I need to extend the servo arm to get the correct range of motion:
 

 

I'm hoping to get this electrified tonight so I can test it properly.

 

Cheers,

 

Matt

Link to post
Share on other sites

I don't want to poop in your pool party but why not just have a couple of sidings on the right hand side of the traverser long enough for your loco?  You can then run around any wagons simply using the traverser itself.  This then does away with the need for having a functioning sector plate on a sliding traverser with all the attendant wiring and physical challenges that presents..

Link to post
Share on other sites

Hi Jon,

 

The main reason is that the way I need to build the baseboards in my loft to fit around existing things makes it a bit more difficult if I need to add more legs.  

 

The other reason is that I want to be able to use the various parts of the traverse to assemble services and by doing it this way, I can be shunting a service together across multiple lines whilst still using the other parts of the traverser to accept incoming services.

 

As far as the challenges of wiring etc are concerned, I plan to have a single cable loom that feeds the DCC bus to the traverser.  I hope to be able to use the Arduino DCC libraries to drive both the traverser itself and the sector plates through JMRI.

 

Thanks for the sanity check though :)

 

Matt

Link to post
Share on other sites

Hi Matt,

This looks to be a pretty clever bit of engineering going on, I shall enjoy keeping up to date with this!

Cheers,

John E.

 

Thanks John, hopefully I'll be able to post some more before the end of the week.

 

Im not an expert but wouldnt the pivot have to be closer to the traverser? It just seems the swing is wrong.

 

Possibly, yes.  I've not done this before so everything is up for grabs here! :D

 

Thanks,

 

Matt

Link to post
Share on other sites

Hi Matt

 

Looks a great challenge - will be watching with interest.

 

Regarding the pivot point - Based on the arc you have cut, the pivot wants to be EXACTLY where the point of the comprass cutter was in your third photo (looks near the centre of the piece of sector plate track). That way you won't need to fettle the arc.

 

If you want to pivot from nearer the right hand end of the sector plate (where the screw is in photo's 8 & 9) you will need to draw / cut the arc based on that pivot point.

 

Hope that makes sense!

 

Happy modelling.

Link to post
Share on other sites

Hi Matt

 

Looks a great challenge - will be watching with interest.

 

Regarding the pivot point - Based on the arc you have cut, the pivot wants to be EXACTLY where the point of the comprass cutter was in your third photo (looks near the centre of the piece of sector plate track). That way you won't need to fettle the arc.

 

If you want to pivot from nearer the right hand end of the sector plate (where the screw is in photo's 8 & 9) you will need to draw / cut the arc based on that pivot point.

 

Hope that makes sense!

 

Happy modelling.

 

Makes a huge amount of sense and had I not been in a rush I probably would have realised that last night!

 

Thanks Sam,

 

Matt

Link to post
Share on other sites

We have Power!!!!

 

13986716245_b58194f461.jpgUntitled by ProfFalken, on Flickr

 

13983511342_6ec6bd0cb9.jpgUntitled by ProfFalken, on Flickr

 

Now I just need to install JMRI onto the laptop and test the connectivity.

 

I've moved the pivot as suggest by a couple of people above and that now looks a lot better, so I need to crack on with the code for the Servo next!

 

Cheers,

 

Matt

Link to post
Share on other sites

This looks interesting

 

Will it be operated by DCC or via a laptop .

 

Hopefully both!

 

The plan is to install an arduino as a DCC decoder (probably based on http://www.mynabay.com/arduino/14-arduino/dcc-monitor/14-arduino-dcc-monitor as this seems to be the most popular reference design at the moment) and then have JMRI send signals to it from the laptop to move the 9g servo to the correct position.  I'm undecided as to whether a servo or stepper motor is the best option here, however as I have a servo and don't have a stepper motor, it will remain a servo for the time being! :)

 

Eventually, I plan to replace the laptop with either a Raspberry-Pi or a Beagleboard, but that's subject to approval from the management committee...

 

My ultimate goal is to have one of the currently unused function buttons on the controller mapped via JMRI to a Jython script that uses the block detection built into the traverser and sector plate (yet to be fitted!) to slowly shunt the loco from the traverser onto the sector plate, move the sector plate to the correct road and then slowly move the loco back out onto the spare road on the traverser, however that is a long way off yet!

 

Thanks for all the interest so far, hopefully I'll be able to get more photos/videos up soon.

 

Matt

Link to post
Share on other sites

  • 2 weeks later...

So, so more time this evening and the prototype in foamcore (before I prototype in Ply, before I build the real thing!) is really starting to come together:

 

First, I cut out some cross-braces and sides for the board to give me clearance for the servo underneath (the holes are to thread the wires through!):

 

13904563948_f6502a00b6.jpg2014-05-02 20.13.07 by ProfFalken, on Flickr

 

I then attached the ends and sides to the base with UHU:

 

14091147365_e43419c573.jpg2014-05-02 20.21.23 by ProfFalken, on Flickr

 

Before cutting a hole for the servo and mounting it in place:

 

13904523489_3a74bd38f4.jpg2014-05-02 20.26.00 by ProfFalken, on Flickr

14088035052_973277039f.jpg2014-05-02 20.26.10 by ProfFalken, on Flickr

 

I noticed that the servo didn't quite marry up to the underneath of the board, so I used some of the off-cuts as a mount:

 

14088036062_9bbb66cc9e.jpg2014-05-02 20.31.20 by ProfFalken, on Flickr

 

Then I added the rest of the bracing and threaded the wires through:

 

13904562190_d6f50baa06.jpg2014-05-02 20.36.59 by ProfFalken, on Flickr

14087945971_dd2872e34f.jpg2014-05-02 20.41.51 by ProfFalken, on Flickr

 

Next steps are:

  1. Wait for the glue to dry (it's been put away for the night now)
  2. Attach the sector plate to the servo
  3. Pass the wires from the sector plate through the holes in the end brace and connect up to the DCC feed
  4. Run the wires for the CBUS, Servo, Arduino and other bits through the base board
  5. Turn on the servo and hope the UHU holds it in place
  6. Test it!

I also have an idea for an indicator as to which track is lined up and ready to be used by the sector plate through the use of BarePaint - a conductive paint that I hope I can place on the base of the traverser and the underside of the sector plate and use that to make the connection.  I've not used this paint before, so it's completely experimental... :)

 

Cheers all,

 

Matt

Link to post
Share on other sites

Matt

 

Very interesting concept I wondered at the start why you were using foam board but it became clear as I read on.

Have you thought about using self adhesive copper strip as opposed to bare paint this will definitely make a good contact.

Is is used for slot car tracks and can be purchased quite cheaply from eBay the other option is micro switches although these may be harder to get in the exact position to line up with the respective tracks.

 

Just search eBay for self adhesive Multi Purpose Copper Tape it comes in approximate 6mm width and can be soldered to

Link to post
Share on other sites

Matt

 

Very interesting concept I wondered at the start why you were using foam board but it became clear as I read on.

Have you thought about using self adhesive copper strip as opposed to bare paint this will definitely make a good contact.

Is is used for slot car tracks and can be purchased quite cheaply from eBay the other option is micro switches although these may be harder to get in the exact position to line up with the respective tracks.

 

Just search eBay for self adhesive Multi Purpose Copper Tape it comes in approximate 6mm width and can be soldered to

 

Thanks Gary,

 

Just to be clear, the foam-core is only being used for this prototype and will not be used for the finished version.  That will be made from a combination of 9mm and 3mm ply with softwood bracing as it is going to be around 6' long and 3' wide!

 

It has made me seriously think about building a small layout from foam-core, but I think I have enough to be getting on with for now!

 

I wasn't aware of the copper tape other than the "anti-slug tape" that you can get from garden centres, so I may well use that on the final version.

 

Once I have this foam-core mockup running well, I plan to recreate just this part in plywood and then I'll move on to the traverser proper and sink three of these into the boards!

 

Cheers,

 

Matt

Link to post
Share on other sites

5471301402857c5cb9db6d3c46b9c33c8ec6a132

The concept sounds seriously flawed to me, would not a few "kick back" sidings set back on the approach side of the traverser maybe hidden by a cutting or building help as you cannot run round in the original design without the loco emerging onto the visible section of the layout under the bridge. 

Link to post
Share on other sites

Hi David,

 

Thanks for the comment.  You may well be right about the kick-back sidings and I'll take that into account when I move on to the layout and the traverser so thank you for pointing that out.

 

I had some progress over the weekend in that I managed to get the sector plate pivoting fine and relatively accurately... until I put my class 20 on top of the track :(

 

Turns out that a Bachmann class 20 is too heavy for a 9g Tower Pro servo to move.

 

Ah well, at least it proves the design, now I need to save up for a more powerful servo.

 

Cheers,

 

Matt

Link to post
Share on other sites

  • 2 weeks later...

This sounds a fascinating project getting the electronics and software to operate something like this and I like your approach of playing with a prototype.

 

There are several points that I have some thoughts on. The geometry of the tracks approaching the sector plate would work better if the pivot point was as far to the right as possible. This would reduce the radius required on the approach. As Sam has pointed out the arc you cut has to be centered on wherever you place the pivot point. As you have already found the friction in turning something like this can be high and it would be much easier if the track end of the sector plate is on some wheels or preferably ball races. It doesn't have to look pretty.

 

You haven't said what operating and control systems you are using for the main traverser. I would look at using exactly the same means to operate a small traverser in the end of the main one instead of 3 separate sector plates. This would then be able to access all the tracks, not just a group of three.

 

David has already pointed out that whatever you do you are only providing a way of releasing a loco from the end, not having a true run round ability. A kick back would be one way of doing it but if you have the length available a second small traverser would give full functionality but would introduce the problem of needing to have it in the right place to run over it to enter the fiddle yard so perhaps its not such a good idea.

 

Hope this gives you something to think about.

 

Stuart

Link to post
Share on other sites

I've seen all the the same issues crop up on a US forum, where a similarly just flat sector plate was tried out to replace the end turnout of a turning "Y", so as to fit it in limited space.

 

IMHO, Just for functionality, you need to think about lowering the"ground level" under the sector plate an inch or so, so you have room for a sufficiently strong and rigid piece of sector track, (for heavy locos) plus some turntable-like wheels or bearings underneath to allow it to rotate with less friction.

 

I'm also confused as to what happens in the tunnel section. Who or what turns the trains there?

 

Andy

Link to post
Share on other sites

Hi Andy,

 

I think you're right about lowering the base of the sectorplate to allow for some kind of bearing arrangement, but this is why I built it in foam-core first so I could find out all these things!  I'm starting to think that I could use an old wagon bogie that I have in the loft on some 2nd radius settrack but I need to work out how well that would fit.

 

The tunnel is mainly there because I can't do a tail-chaser in the loft owing to plumbing etc. so by having that in place it allows me to run DMUs and other services (MPVs, Tampers etc) up to 4ft in length into the tunnel on the "down", across to the headshunt on the "up" once inside the tunnel (or vice-versa) and then back out onto the layout after a suitable delay - I suppose you could think of it as a two-track hidden fiddle yard!

 

The entire layout will have block detection and a start-of-day sequence will allow JMRI to track the locos via the various blocks allowing me to automate a large part of the running of the layout.

 

Hope that answers some of your questions,

 

Matt

Link to post
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...