Shelf Marshes (first attempt at a cameo layout)

[47137]

I have built a switch panel to let me use the memory functions in the servo controller board, and this panel has a write-up on my blog:

https://www.rmweb.co.uk/community/index.php?/blogs/entry/23872-route-setting-panel-for-megapoints-servo-controller/

 

 

Operationally I am not sure whether this panel should be located on Shelf Marshes or several yards away on the fiddle yard. Technically, if I added a transistor to drive the LED and reduce the current taken from the servo controller, I could have a second identical panel wired in parallel with the first and with LED indicators at both.

 

- Richard.

[47137]

I am now half-way through the 30-day lead time for the baseboard plywood.

 

It's strange but building parts of a layout without a baseboard is harder than I expected. I could set about rewiring the Peco points for DCC, but I would like to bend some of these to fit the layout and it would be best to adjust their shape before cutting rails.

 

So the project goes on hold for a few weeks. I might be able to finish some of the minor models I have started :-)

 

- Richard.

[47137]

On 30/06/2020 at 14:26, 47137 said:

...

This leaves me to focus on how to switch the polarity of the frogs. Really, there are only two ways with any appeal to me: double-pole switches on the control panel (one pole for the servo controller, one pole for the frog) or a frog juicer. I have a collection of old-fashioned switches - a panel from H&M and some Lucas toggle switches from 1960s cars. These would build up into a control panel with some character, but they are single-pole and would need the frog juicer.

...

 

Frogs

 

It has dawned on me. if I use my double-pole switches to control the frogs as well as the servo driver board, then the frog polarities will go out of sync as soon as I try to use my route setting board. Doh!

 

I have also convinced myself, a single servo horn really needs to be applied to only one mechanical task, for example moving points or moving a microswitch, but not both. It might be reasonable and even sensible to use the point tie bar itself to move microswitch, but I have a feeling such a device will be needing adjustment from time to time.

 

I have mixed feelings about a frog juicer. Fundamentally, it seems wrong to design and build a control system which depends on short circuits to make it work; but a fully solid-state solution does have a lot of appeal. If I use a frog juicer and of course disconnect it for analogue operation then all the turnouts will have dead frogs and this is going to cause frustration because nearly all of my loco projects are for smaller prototypes.

 

So really, I've talked myself into using relays to provide power to the frogs. This means the cost of the control system for the points (ignoring the switches at the front end) has doubled to around £148 including eight servos. This is 148 / 8 = £18.50 per point, pretty much the same as Cobalt stall motors worked by local switches and with no route setting. It also means the layout will be using five circuit boards (including my remote switch panel) to control the points, and this seems a bit over the top.

 

As it stands, servos probably still "win" for me against other point motors because they let me have my very shallow aluminium baseboard frame, but it is a close call. I will be happier if I get a real benefit from the route setting feature, or I can add some servos to the unused channels (e.g working scenic features), or I find myself wanting a centralised control desk for the whole of "Shelf Island" and I can invest in a MegaPoints Multipanel to do this.

 

- Richard.

Edited July 21, 2020 by 47137
Simplicity

[47137]

The relay boards and their controller have arrived from MegaPoints and here are the various boards for the layout connected together with a couple of servos on the bench:

 

The green wire with the croc clips is taking the place of switches to control the servos.

 

I think it is a shame, the servo controller (centre) cannot drive relays directly, but realistically the MegaPoints range needs a relay driver module for applications like analogue section switching and it makes engineering sense to pair this board with the servo controller.

 

The relays make a resonant 'click' and they need to go underneath the layout with a cover to muffle the sound.

 

The most important connection here for me is the three-wire bus between the servo controller and the relay driver (the two green pcbs) because this is the very beginning of a possible control bus extending around the whole of 'Shelf Island'. Setting up the servo controller to work from such a bus instead of local levers (toggle switches) would leave my DIY route-setting board redundant, but somehow I don't mind. If the entire layout gets a control bus end-to-end then it is sensible for route setting to be end-to-end as well.

 

To my mind, a control system for a model railway layout has four characteristics - engineering performance, usability, aesthetics, and closeness to prototype. I can think of a dozen ways to do the engineering part, but the rest is a lot harder.

 

- Richard.

[47137]

I have discovered the plywood is on a 30 working day lead, not merely 30 days. So delivery is around 13th August, not yesterday.

 

I am working my way through the options for power supplies at the moment.

 

- Richard.

[shipbadger]

There are shortages of all types of timber at present.  I've just collected twelve sheets of plywood from a specialist panel supplier.  Normally that quantity would have qualified for free delivery but they have just doubled the minimum spend to qualify.  The 'excuse' was that they are rushed off their feet.  This firm now supplies nothing but imported materials and the pandemic has caused problems all over the world so supply chains are disrupted.  My air compressor needs a spare part and that will not arrive (from Italy I think) for another fourteen days but they can't tell me how long it will take to get from dockside to my front door.

 

Tony Comber

[47137]

On Friday evening I received an email to tell me my plywood had been dispatched. There is hope!

 

In the meantime I have been planning the track feeds for the layout. I have picked up on Wago blocks as suggested by RFS. I am planning for 16 pairs of feeds, so  there is a feed for almost every section of track.

 

I am still pondering whether to provide for a few isolating sections for DC operation. On the one hand, the layout will be workable on its own with one engine in steam; but there will be more engines around when the rest of Shelf Island is connected. It is a lot easier to put the wires in now, and even hang them off some spare relays on the MegaPoints bus ... but I must beware of "feature creep" and adding things merely because I can.

 

- Richard.

[47137]

The plywood arrived this afternoon. I nearly had a fit when I picked it up - 21 kg in all - but mercifully about a third of this is the packaging.

• I have a sheet of 6 mm ply to make the track bed, weight 4.8 kg
• A sheet of 4 mm ply to make a bottom cover or dust cover, 3.1 kg (this is already looking "optional" to keep the weight down!)
• Five pieces of 4 mm ply to make the outer supports for the backscene and the wings of the proscenium arch, total 6 kg

All of these cut to size as ordered, and all being nice quality Baltic ply without voids. The lack of air gaps might explain why they are so heavy.

 

I can now stop dreaming up new features for the layout and settle down and draw out the track plan full size. I hope it ends up much like my mock-up, with an extra 40 mm or so of depth front and back but no awkward curves or tight spots.

 

- Richard.

Edited August 11, 2020 by 47137

[47137]

During last week I made a first effort at the baseboard:

 

This had some merits: it was large enough to hold the track plan and small enough to fit into the alcove in the hobby room. Unfortunately, it was 3 mm too wide to do through a domestic doorway and it already weighed 9.5 kg.

 

The 1-inch square aluminium box section I bought from Aluminium Warehouse is of good quality but with a 2 mm wall thickness is weighs as much as 2 x 1 timber. Araldite glue falls off it, and the head of my pop riveter is broader than the inside of the channel, so my only practical fixing is self-tapping screws.

 

Aluminium is strong as well as rigid, and to me this strength is not necessarily a good thing. To straighten a slightly curving length you have to support the ends, stand on it, and hope it moves the desired distance. Pulling a length of timber into line is a lot easier, and wood glue is as strong as the wood.

 

Many years ago, someone told me "aluminium is unsuitable for model railway baseboards". This seemed a bit abrupt at the time, but now all I can add is "in my hands". It ought to work - after all, it won't warp or twist, but not this time. Perhaps, for a semi-permanent layout.

 

I have stripped down most of last weeks' effort and started to rebuild it using pine strip wood. This is looking good.

 

- Richard.

 

 

[Allegheny1600]

Hi Richard,

That's an impressive baseboard!

Regarding using Araldite on aluminium, did you de-grease and also 'key' the surface? A gentle sand over with a fine emery cloth should be sufficient. You certainly don't want to scratch it as I was once told by a Rolls Royce engineer, that scratches in aluminium can propagate until the piece fails! However, I'm sure he meant under constant vibration, as in aviation.

Otherwise, I too have heard that aluminium is best joined by mechanical means.

Cheers,

John.

[47137]

2 hours ago, Allegheny1600 said:

Hi Richard,

That's an impressive baseboard!

 

It was a monster of a baseboard! The second version is going to be c.30 mm narrower and maybe 1 kg lighter. The backscenes will taper down towards the back too.

 

The track plan simply does not lend itself to a split across a baseboard joint, nor will it go into a shorter area. I cannot have a longer baseboard either, because it needs to go into a recess in the room. So the size is fixed, and all I can try to do is to minimise the weight i.e. use as little timber as I can to build it.

 

At the end of the day, I can turn the thing over in the room, stand it on a trestle table to work on it, and stand it on end on the floor to get it out of the way. It will be a two-man lift to get it down the stairs if it goes to a show. A sack truck might be good to move it across a smooth floor.

  

2 hours ago, Allegheny1600 said:

Regarding using Araldite on aluminium, did you de-grease and also 'key' the surface? A gentle sand over with a fine emery cloth should be sufficient. You certainly don't want to scratch it as I was once told by a Rolls Royce engineer, that scratches in aluminium can propagate until the piece fails! However, I'm sure he meant under constant vibration, as in aviation.

Otherwise, I too have heard that aluminium is best joined by mechanical means.

Cheers,

John.

 

No I didn't. A gentle sanding would have helped a lot.

 

In retrospect I think a broad U section aluminium would have been a better choice, say 40 x 20 mm. This would make room for pop rivets or screws in pairs at the joints. And much thinner walls would be fine - say 1 mm thick not 2 mm.

 

- Richard.

[47137]

Guess where the chemical plant goes! Here is the floor of my baseboard before I add the sides and while I can still carry it in one hand.

 

The materials here are 6 mm Baltic ply for the track bed, 15 x 25 mm pine strip wood and 4 mm ply gussets. Grace is here to balance the photo, and is of course reconstituted marble.

 

One length of the aluminium box section stayed because it was too difficult to take it out. The cable trunking is 25 x 40 mm, this is for the track wiring and 12V DC distribution. I envisage the servo wiring being fairly direct from the servo controller out to each servo.

 

Taking out the sections of ply for the chemical plant and the marshes took away about a quarter of the track bed.

 

This section weighs 5.9 kg, so about 0.5 kg per square foot. Adding some narrow strips of ply around the outside edges would turn it into all a sturdy and lightweight conventional baseboard, but this one will be getting ply sides.

 

- Richard.

[ISW]

1 hour ago, 47137 said:

*snip*

 

*snip*

The cable trunking is 25 x 40 mm, this is for the track wiring and 12V DC distribution. I envisage the servo wiring being fairly direct from the servo controller out to each servo.

 

*snip*

 

- Richard.

Richard,

 

That's a very interesting baseboard construction you have there. I can see why you wanted aluminium with that 6mm plywood top.

 

I originally planned to have a 'cable trunking' along the front edge of my baseboards; either a 'U' shape or 'C' shape, both with a lid. A 'U' seemed better as the cables wouldn't fall out each time I opened the lid. However, once I'd decided on separate baseboards with joins the trunking didn't make sense for me.

 

How to you plan to retain the cables in your 'n' shaped cable trunking? Will there be a lid? 

 

The other issue with trunking is the necessity to drill holes for the cable entry / exit points and, with aluminium, you'll need to ensure the holes are de-burred.

 

Looking forward to seeing the same photos after the wiring is installed.

 

Ian

[ISW]

On 25/07/2020 at 16:29, 47137 said:

The relay boards and their controller have arrived from MegaPoints and here are the various boards for the layout connected together with a couple of servos on the bench:

 

*snip*

 

The relays make a resonant 'click' and they need to go underneath the layout with a cover to muffle the sound.

 

*snip*

 

- Richard.

Richard,

 

On my layout all the stripboard / veroboard / MegaPoints boards are screwed down to the underside of the baseboards with blue silicone 'washers' about 2mm thick. These provide the necessary 'standoff' to clear the solder connections on the underside. The 'washers' were cut from a ~3m length of aquarium tubing I found at a local PoundStretchers shop for a couple of pounds! I think I'll have enough for the entire build. The silicone would also provide you with some sound isolation to deaden the relays clicking.

 

Ian

[ISW]

On 21/07/2020 at 08:37, 47137 said:

I have also convinced myself, a single servo horn really needs to be applied to only one mechanical task, for example moving points or moving a microswitch, but not both. It might be reasonable and even sensible to use the point tie bar itself to move microswitch, but I have a feeling such a device will be needing adjustment from time to time.

Richard,

 

I think you may be over-complicating things a bit.

 

Here's a photo of my servo / microswitch / bracket assembly. It simply screws to the underside of the board. I found that the adjustment of the turnout throw was by rotating the bracket about the rear fixing, thus getting the servo arm in the correct position under the toe of the turnout. The servo has way more 'throw' than is required by the turnout itself [1]. Thus, the adjustment of the servo is more for getting the correct throw for the microswitch. Therefore, the 2 adjustments are reasonably independent.

 

This is a very cheap solution (microswitch £0.50 + bracket £0.50 + servo £2.50 + MegaPoints £60/12=£5) at ~£8 per turnout.

 

So far there has been almost no subsequent adjustment. 

 

[1] - My baseboards are 12mm thick, but I'm using the hole in the servo arm near the pivot. You can use one of the other holes to overcome your 6mm plywood.

 

 

Ian

[47137]

9 hours ago, ISW said:

...

 

How to you plan to retain the cables in your 'n' shaped cable trunking? Will there be a lid? 

 

The other issue with trunking is the necessity to drill holes for the cable entry / exit points and, with aluminium, you'll need to ensure the holes are de-burred.

 

Looking forward to seeing the same photos after the wiring is installed.

 

Ian

 

My cable trunking will have snap-on lids to hold the wiring. One of these is fitted in the photo beside Grace, the horizontal length above the level of her head. The long length of trunking (vertical in this photo) will have its lid cut into shorter sections to go between the timber cross members.

 

I want to use my bradawl to make holes in the trunking for my 7/0.2 wires. The bradawl makes a conical hole and when I work out how to do it right, I reckon the hole will shrink back onto the wire and grip it. If not, I can use cyanoacrylate. The cable trunking is large enough to hold the Wago blocks and I want these out of sight.

 

For the aluminium, I have drilled two 8 mm holes straight through and lined these with plastic tube cut from a garden ornament. The plan needs very few layout wires at the front of the layout - two servos and, if I am feeling very thorough, two pairs of track feeds. These tubes are just about visible a few inches below Grace's knee. She is being more useful than I expected :-)

 

- Richard.

 

[47137]

11 hours ago, ISW said:

Richard,

 

I think you may be over-complicating things a bit.

 

Here's a photo of my servo / microswitch / bracket assembly. It simply screws to the underside of the board. I found that the adjustment of the turnout throw was by rotating the bracket about the rear fixing, thus getting the servo arm in the correct position under the toe of the turnout. The servo has way more 'throw' than is required by the turnout itself [1]. Thus, the adjustment of the servo is more for getting the correct throw for the microswitch. Therefore, the 2 adjustments are reasonably independent.

 

This is a very cheap solution (microswitch £0.50 + bracket £0.50 + servo £2.50 + MegaPoints £60/12=£5) at ~£8 per turnout.

 

So far there has been almost no subsequent adjustment. 

 

[1] - My baseboards are 12mm thick, but I'm using the hole in the servo arm near the pivot. You can use one of the other holes to overcome your 6mm plywood.

 

I can think of four ways to switch the frogs and none seems ideal:

1. double pole switches - rules out remote operation over MegaPoints bus

2. frog juicer - depends on short circuits to function

3. microswitches - are not quite 100% reliable and are asking the servo to do two things

4. relays - seems rather over the top

 

I wish I could write out the same list with the merits of each!

 

At the end of the day, unable to choose anything I really liked, I asked myself "suppose I am building this layout for someone else - which method would I choose?" and the relay seemed the obvious answer. I still admire your microswitches and I think the bracketry looks ingenious.

 

For my 6 mm thick track bed I have tried out the largest servo throw with 0.4 mm piano wire and the servo seemed to move a Peco tiebar (minus its over-centre spring) very nicely. The only setting up was to centre the servo and install the horn to the mid-point.

 

I can look again at microswitches for a point position indicator but I like the reliability and lack of mechanical tweaking offered by using relays for the frogs.

 

- Richard.

[nick_bastable]

12 or 5 volt servo boards are cheap easy to activate  using a DPDT  switch  require no mechanical aligning ,   just a thought  (although they are the only thing I use )

 

Nick B

 

 

[47137]

On 21/08/2020 at 01:05, nick_bastable said:

12 or 5 volt servo boards are cheap easy to activate  using a DPDT  switch  require no mechanical aligning ,   just a thought  (although they are the only thing I use )

 

Nick B

 

I like the idea of route setting and the possibility of remote control, because "Shelf Marshes" is part of a larger scheme. But I will remember these servo boards if I add servos for scenic features like gates and barriers which won't be part of routes.

 

Many thanks.

 

See

https://www.rmweb.co.uk/community/index.php?/blogs/entry/23872-route-setting-panel-for-megapoints-servo-controller/

 

- Richard.

Edited August 22, 2020 by 47137
Added link

[47137]

I have discovered bendy ply aka "flexible plywood" - absolutely brilliant!

 

This is 3 mm thick sheet. Two layers glued together become a rigid assembly. "Shelf Marshes" has the foundation of its sea wall separating the railway from the mud flats.

 

- Richard.

[47137]

A final photo of the empty baseboard:

 

The backscene will be in two layers. The outer layer is 4 mm Baltic ply, braced with 15 x 15 mm pine stripwood. The inner backscene is 3 mm bendy ply attached with Evo-Stik contact adhesive.

 

I have installed the right-hand inner backscene. This activity consumed almost a whole 250 ml tin of Evo-Stik for barely a third of a square metre. Coverage of glue is much like paint - about half what the manufacturer claims. Perhaps I should have primed the plywood first.

 

The plan now is to lay the track for the 90-degree curve and the tram depot. I want to see how the tram depot will look before I decide the curvature of the backscene behind it.

 

- Richard.

[ISW]

8 hours ago, 47137 said:

A final photo of the empty baseboard:

Richard,

 

Any plans for legs or supports of some kind?

 

And don't forget to allow for overhead lighting ...

 

Ian

[47137]

10 hours ago, ISW said:

Richard,

 

Any plans for legs or supports of some kind?

 

And don't forget to allow for overhead lighting ...

 

Ian

 

Patience! I started the lighting rig yesterday evening. I have two bi-colour LED tapes inside a length of L-shaped aluminium:

 

The tapes have two sets of LEDs, warm white and cool white. I have bought two drill speed controllers (because they use pulse width modulation) and to begin I want to try driving one set of warm LEDs from one controller, and the other set of warm LEDs and both sets of cool LEDs from the second controller. I suppose ideally the first group will provide a sort of "default sunrise" look and ramping up the brightness of the other set will take the scene from sunrise to full daylight.

 

The LEDs are running lukewarm to the touch and I think the aluminium is taking away a lot of the heat they produce. These Chinese-made LED tapes seem like a lottery but cool running should help them last longer.

 

The table is pale grey not blue and the LED lighting has fooled the auto white balance in my phone camera. I took the photo to find unwanted bright spots on the inner sides of the backscene. I think I've got the spread of light about right; the LEDs stop c.100 mm short of each end. The tank wagon is in the front-most location of any train in the plan and the pattern of light and shadow on and behind it looks ok; I want Northern European not Tropical sunlight. 

 

At the moment, the layout sits on this table on four rubber feet. At home, the layout will rest on timber battens along each side of an alcove, and for shows I want to try the scheme using short trestles on top of the picnic table as proposed by Iain Rice in his book "Constructing Cameo Layouts".

 

- Richard.

[ISW]

Richard,

 

I can tell you've planned and thought-out your layout design and build in some detail. Very impressive.

 

Ian

[47137]

7 hours ago, ISW said:

Richard,

 

I can tell you've planned and thought-out your layout design and build in some detail. Very impressive.

 

Ian

 

The lockdown gave me plenty of time to think about it!

 

When I built the main baseboard for Shelf Island, I ignored the control panel, backscene and lighting to my cost. This time, I am doing them first.

 

Thanks for posting here. Sometimes I feel I am writing a monologue, and it's nice to see the topic livened up a bit. My "plan" is to build Shelf Marshes as a working layout as far as the track and point motors and controls, and then go back to the main baseboard, which is five years old and deserves some scenery. Whatever I learn on the main baseboard will make for better scenery on Shelf Marshes, and looking at some of the work on RMweb and at some shows last year I have a high bar to reach.

 

- Richard.