Shelf Marshes (first attempt at a cameo layout)

[47137]

I have put in all of the wiring under the baseboard for the track feeds and servos and I hope this passes muster:

 

The cleats are the smallest ones from Screwfix ("3 - 5 mm"), drilled out for No.2 x 9 mm self-tappers. I bought 70 screws and I have four left over.

 

The wiring for the servos and track feeds all worked first time except for the DB-25 (above). I haven't labelled the wires for the track feeds (blue and brown), but with a maximum of three wires of the same colour in a bundle I think changes should be straightforward. The green wires are frogs and the labels beside them match up with their associated servo.

 

The trunking lids could go on now, but I might leave them off until the wiring for building lighting is in.

 

The relays for the frog switching have caused "issues" if not outright "problems" and I will come back to these when I know everything is sorted out.

 

- Richard.

Edited September 21, 2020 by 47137

[ISW]

6 hours ago, 47137 said:

I have put in all of the wiring under the baseboard for the track feeds and servos and I hope this passes muster:

Richard,

 

Very tidy indeed. It does look the 'business'. The only change I'd make would have been to run the MegaPoints network cable in the 'upper' (on the photo) cable duct, with the servo cables, instead of with the power cables in the 'lower' duct. But as you've no interference issues there's no need to modify.

 

Ian

[47137]

Ian,

 

What do you do with the bus when you want to connect a MultiPanel to it?

 

It seems to me, every MegaPoints control board has two bus connections. They daisy-chain happily enough, but if you want to connect a baseboard at each end of a baseboard, there is no spare terminal for the panel.

 

Maybe you just double-terminate the bus?

 

- Richard.

[ISW]

1 hour ago, 47137 said:

Ian,

 

What do you do with the bus when you want to connect a MultiPanel to it?

 

It seems to me, every MegaPoints control board has two bus connections. They daisy-chain happily enough, but if you want to connect a baseboard at each end of a baseboard, there is no spare terminal for the panel.

 

Maybe you just double-terminate the bus?

 

- Richard.

Richard,

 

I'm in the unusual (?) position of having 10-baseboards in a circle with a jumper cable at each baseboard joint. For the track power (DCC) and MegaPoints controller connection I simply remove one baseboard jumper and connect the DCC / MegaPoints to one side of where the jumper was.

So, for the MegaPoints the controller is at the 'end' of a long network cable running around the layout daisy-chaining the various servo boards together. Therefore, all the servo boards use both network connections, but the Controller board only has a single connection. Obviously, one of the servo connections goes to the dead-end at the missing baseboard jumper location.

Does this make sense and/or answer your question?

 

As far as I know, you can also put the Controller board in the middle of the network, and use both network connections to then branch out to 2 strings of daisy-chained servo boards.

 

Ian

[47137]

Of course. The boards for the control panels have two bus connectors.

 

My own layout is five, possibly six baseboards in a U shape. As it grows, I can see possibilities for a few local control panels or one fairly central one. The MegaPoints range seems to be sufficiently flexible to support most arrangements.

 

Many thanks.

 

Richard.

[47137]

On 21/09/2020 at 16:13, ISW said:

Richard,

 

Very tidy indeed. It does look the 'business'. The only change I'd make would have been to run the MegaPoints network cable in the 'upper' (on the photo) cable duct, with the servo cables, instead of with the power cables in the 'lower' duct. But as you've no interference issues there's no need to modify.

 

Ian

 

I have given this some thought, and realised I designed my cable trunking to keep control and power signals away from each other. And, my relay driver and its relay boards are (unsurprisingly) right next to the power trunking.

 

This means, the only sensible routes for serial data cable are through a short length of the power trunking, or along just outside it, or diagonally across the gaping hole in the baseboard where the chemical plant will go.

 

At the moment, the serial cable is 0.9m long with about half of this inside the power trunking and the rest in orthogonal runs on the surface of the baseboard. I think, if there are problems I would look to moving the cable to go parallel with the trunking an inch or so away. But if this happens, I would become a bit twitchy myself about the robustness of the control equipment; I would probably look to frog juicers instead of relays. I suspect, there won't be any problems with such a short serial cable; but I do wonder why the designer went for an unbalanced serial link when something like RS-485 is there for the taking, balanced line and support for multi-drop.

 

(I'm not opposed to microswitches but the track layout is a bit dense in places and finding room to fit them in as a retro-fit looks like a pain; and I couldn't get my trial one to work well enough).

 

- Richard.

[ISW]

32 minutes ago, 47137 said:

(I'm not opposed to microswitches but the track layout is a bit dense in places and finding room to fit them in as a retro-fit looks like a pain; and I couldn't get my trial one to work well enough).

Richard,

 

I'm using what's called a 'Sub-Miniature Microswitch' that are only 12.8 x 5.8 x 6.5mm (Length x Width x Height) and mounted to a short bit of U-channel (the same used to mount the servo) and bolted to the side of the servo mount. Space requirements minimal. You can see them at https://www.switchelectronics.co.uk/bent-lever-subminiature-pcb-microswitch-spdt-3a . I wanted the 18mm leg version, but it wasn't in stock. This one has a long enough leg, but I'll have to straighten them with me pliers!

 

Setup is a bit of a faff using the buttons on the MegaPoints servo board to set the two extremes of movement while making sure its enough to 'click' the microswitch. I've not had to go back and adjust many, just the odd one that I wasn't careful enough with in the first place.

 

Cheap as well, the ones I bought originally were 20p each, but my latest purchases were 24p each. 

 

Ian

 

[47137]

At the moment, I will be happy to have my eight servos working their points reliably using the default settings on the servo controller i.e. a movement of about 90 degrees.

 

I think I am getting there. I have added a shim of leather under two of the mechanisms, This is 3.3 mm thick stuff, cut from an old belt. The total thickness of the baseboard between tiebar and U-section aluminium becomes 6.3 + 6.3 +3.3 = 16 mm. With 0.5 mm piano wire in the second hole out from the servo shaft, and a bend in the wire to set the neutral position of the tiebar to match the neutral position of the servo, the movement seems to be just about spot-on. At least, for these first two servos.

 

I am getting a bit wearied by these servo installations. I know, when everything is just right, I will be happy ... but there is a part of me saying really, a set of stall motors would have gone in for much the same total outlay and they would have much better frog switching.

 

I have discovered two "features" or perhaps "characteristics" of the relay driver, I have this slaved from the servo controller:

 

(1) When you throw a point lever, the associated relay changes its state instantly, before the servo begins to move. This means, if you are using Peco Electrofrog turnouts, you are guaranteed to get a short circuit every time you throw the lever unless you make two cuts in the rails to isolate the point blades from the frog. I hadn't done this for some of my points, so I had to modify them in place on the layout - cut the rails and add bonding from the point blades to the stock rails.

 

(2) If you configure a servo to mode 2 instead of mode 1, to reverse its direction, the operation of the associated relay does not change. This is either a good thing, in that you can choose mode 2 if you wired the relay terminals back to front; or a frustration, in that if you want to reverse operation of the lever you must rewire the relay.

 

Neither (1) nor (2) appears in any of the system documentation I have read, though apparently if you buy the four channel servo board there is an option to have the relay operated half-way through the servo movement. Which is a lot more sensible, but too late for me. And wouldn't have the route setting feature.

 

The truth is, a stall motor has integral contacts for frog switching and they can be designed to operate when we would want them to operate - just before the end of the movement. So the frog can be usefully isolated while the points are moving, and become live when they stop. This would support the Electrofrog design and you wouldn't have to modify the turnout. You cannot do this with the relay board, but you could, with some patience and skill, do it with a pair of microswitches one each side of the servo. Or a frog juicer.

 

Onward and upwards.

 

- Richard.

[47137]

I forgot to mention, I ran a train on Sunday. It had to be an 00 gauge one (because it had its decoder set to address 3 and my DCC handset programmed for my H0 engines had flat batteries) but run it did, without derailing or indeed me cleaning the track.

 

- Richard.

[ISW]

13 hours ago, 47137 said:

At the moment, I will be happy to have my eight servos working their points reliably using the default settings on the servo controller i.e. a movement of about 90 degrees.

 

13 hours ago, 47137 said:

I am getting a bit wearied by these servo installations. I know, when everything is just right, I will be happy ... but there is a part of me saying really, a set of stall motors would have gone in for much the same total outlay and they would have much better frog switching.

Richard,

 

I therefore consider myself very lucky that my installation of servos and microswitches on ~33 turnouts went relatively smoothly. As I was a total novice, and this being my first (and last?) layout, I did at least build a short sample length of double-track with a crossover. I used that to iron out the installation issues.

 

I do have 'modified' electrofrog turnouts, with the frog totally isolated from all other rails, and jumper wires installed between the stock and tongue rails of the switches (although this doubles up as the track feed as well).

 

As to costs:

1. MegaPoints 12xservo board: £35 - say £3 a turnout
2. Aluminium channel: £7 - that's enough for ~20 turnouts, say £0.50 a turnout
3. Servo: £3 a turnout , although my last purchases were nearer £1.50
4. Microswitch: £0.25 a turnout

That's less than £7 a turnout, even allowing for some piano wire and cabling. That compares very favourably with a tortoise motor at ~£20 each.

 

Ian

 

[47137]

Ian,

 

I don't think you have been especially lucky; rather, I allowed myself to be mislead by an instructional video showing a 6 mm baseboard, I compounded this with my own dummy run (I didn't study it thoroughly), and I'm now looking for refinement of a basically sound design.

 

This is my present iteration:

 

I am seeking quieter operation. The brown layer here is the strip of leather. I've ordered up some foamed neoprene to see if this works better.

 

Some of the motor noise goes into the baseboard through the screws. If I dispense with these and glue the servo mount into place, I will make it difficult to experiment with microswitches.

 

Mechanically, this arrangement is looking good. The lateral movement of the piano wire is about 8 mm in the servo horn, the fulcrum is roughly twice as far from the hole here as it is to the hole in the tie bar, and there is a tiny bit of tension on the piano wire at each end stop.

 

I am happy enough; really, I'm tweaking things now.

 

- Richard.

[ISW]

44 minutes ago, 47137 said:

I am seeking quieter operation. The brown layer here is the strip of leather. I've ordered up some foamed neoprene to see if this works better.

Richard,

 

Noise of operation is not something I'd considered or noticed. My own error was not to remove the 'overspring' on the Peco switches, so I get a defined 'click' at each operation. That's something I will correct on the Upper Baseboards. The servo itself seems (?) quiet to my ears. Each to his own I suppose ... and the amplification effects of the baseboard. My 12mm baseboards must be stiff enough to damp the vibrations.

 

Maybe you need to glue a piece of 6mm plywood to the underside of your baseboards, large enough to encompass a number of servo installations? Just an idea ...

 

Glad to hear you are enjoying the 'process'.

 

Ian

[47137]

You can remove the Peco springs now if you like. Just prise up one of the two metal tags on the cover moulding beside the tirebar. Lift out the cover, lift out the spring, and for cosmetic reasons put the cover back and close the metal tag.

 

After you remove the springs you will find a new topic to play with: point blades on some turnouts will tend to work their way away out of their hinges. Because the spring isn't holding them in place.

 

A good solution is to place a strip of styrene, 0.5 mm thick and about 1.5 mm wide, alongside the sleeper beside the tiebar. This will restrict the tiebar to have only one degree of freedom. I picked this up from one of the UK model railway clubs with a YouTube channel, possibly Chadwick but I'm not sure.

 

- Richard.

[ISW]

4 hours ago, 47137 said:

You can remove the Peco springs now if you like. Just prise up one of the two metal tags on the cover moulding beside the tirebar. Lift out the cover, lift out the spring, and for cosmetic reasons put the cover back and close the metal tag.

Richard,

 

That might work with a cork roadbed or no roadbed, but it's not really possible with a foam roadbed. Bending the metal clips for removal is okay, but it's next to impossible to bend them back as there is 'nothing' to push down against. The foam is not stiff enough.

 

I have to remember to remove the spring 'before' installation.

 

4 hours ago, 47137 said:

A good solution is to place a strip of styrene, 0.5 mm thick and about 1.5 mm wide, alongside the sleeper beside the tiebar. This will restrict the tiebar to have only one degree of freedom. I picked this up from one of the UK model railway clubs with a YouTube channel, possibly Chadwick but I'm not sure.

 

Thanks for the tip. I've had a quick look at some of my installed turnouts and there is currently very little gap between the tiebar and the adjacent sleeper. Plus, the piano wire is helping retain the tiebar in a correct longitudinal location.

 

Ian

[ISW]

4 hours ago, ISW said:

Maybe you need to glue a piece of 6mm plywood to the underside of your baseboards, large enough to encompass a number of servo installations? Just an idea ...

Richard,

 

If you are really serious about eliminating the vibration / noise propagation, then use a large piece of 6mm plywood, as I previously suggested, but screw it to the underside of your baseboard with rubber washers inbetween the 2 plywood faces. The plywood will be stable enough for the servos, but the air gap created will eliminate any vibration transmission to your baseboard.

 

Ian 

[47137]

I am intrigued how a servo can be almost silent when held in my fingers, but make a noticeable sound when attached to the baseboard. I suspect, a shallow curtain hung across the front of the baseboard (in front of the void below it) will suppress anything. I do want to do things as well as I can; so experimenting with some different shim materials may not actually achieve much but will show me what I can and cannot achieve!

 

I have discovered, if you put two styrene shims beside the tiebar you hide almost all of the hole for the piano wire:

 

The shim on the left is holding the point blades into their hinges. The one on the right is cosmetic.

 

This is my Setrack turnout, I am experimenting here hence the overlong piano wire. Black styrene would be better than white.

 

You can see how the hole in the tiebar here was elongated when the servo "jumped" for some reason. These servos are almost too powerful for the task in hand, and a thin springy piano wire seems a sensible precaution. This damage was done by 0.8 mm wire, I'm now using 0.5 mm.

 

- Richard.

[shipbadger]

I've had the problem of holes enlarging in tie bars before, as a result of wear.  Rounding the 'slot' carefully and inserting a short length of brass tube cured the problem.

 

Tony Comber

[ISW]

17 hours ago, 47137 said:

I have discovered, if you put two styrene shims beside the tiebar you hide almost all of the hole for the piano wire:

Richard,

 

If you use a foam underlay, that automatically 'hides' the hole as it's impossible to drill a hole in it. When you drill the baseboard, the foam 'moves out of the way' and moves right back. Neat.

 

No sign of the hole, although the tiebar is a bit 'twisted'. Still works just fine though.

 

Ian

[47137]

4 hours ago, shipbadger said:

I've had the problem of holes enlarging in tie bars before, as a result of wear.  Rounding the 'slot' carefully and inserting a short length of brass tube cured the problem.

 

Tony Comber

 

Tony, I was thinking along these lines but you have given me the confidence to give it a go:

 

This is better than the original. I am tempted to make the same mod to all of the points on the layout.

 

Many thanks,

Richard.

[47137]

3 hours ago, ISW said:

Richard,

 

If you use a foam underlay, that automatically 'hides' the hole as it's impossible to drill a hole in it. When you drill the baseboard, the foam 'moves out of the way' and moves right back. Neat.

 

No sign of the hole, although the tiebar is a bit 'twisted'. Still works just fine though.

 

Ian

 

I wonder why the tiebar is twisted.

 

I have achieved a similar effect by bending a Peco turnout. This is a medium radius one shoved about 10 mm out of true:

 

 

This makes for nicer-looking track. It is easier on a large radius turnout and I wouldn't try it on a small radius one. This "adjustment" is as far as I would dare to go on a medium radius one.

 

- Richard.

[ISW]

1 hour ago, 47137 said:

I wonder why the tiebar is twisted.

 

I have achieved a similar effect by bending a Peco turnout. This is a medium radius one shoved about 10 mm out of true:

Richard,

 

In my case it was simply the gaps between the tongue rails and the associated closure rails of the turnout. One was a tad larger than the other. I could easily 'twist' the tiebar back into a better, non-twisted, alignment.

 

Well, you said the tongue rails could move out of their hinges. How correct you were!

 

Ian

[47137]

I have a five-year old layout with three Peco turnouts driven by Tortoise motors. I took the springs out of the turnouts and can I vaguely remember trying to put a tiny bit of tension on the piano wire to push the tongue rails towards their sockets. Certainly, none of them have ever worked their way out.

 

I completely forgot to do this for Shelf Marshes. To be honest, there is so much power in a 9g servo it would be easy to put a gentle bend in the wire to do this.

 

For the record, the Tortoise motors seem to have a piano wire between 0.6 and 0.7 mm thick - an American size I guess.

 

- Richard.

[ISW]

5 hours ago, 47137 said:

For the record, the Tortoise motors seem to have a piano wire between 0.6 and 0.7 mm thick - an American size I guess.

Richard,

 

MegaPoints supply a piano wire of 0.64mm diameter. That's the equivalent of 22 SWG wire; an American size.

 

Ian

[47137]

Ian,

 

I have been thinking about your servo mounts with microswitches. It may be, the Peco over-centre spring is helping them to work (or at least helping with the setting-up) by holding the tongue rails in position against the stock rails.

 

Of course, this concern may be completely unfounded but I am on my fourth iteration of servo mounts without microswitches and reaching my limit of frustration. This time, I am adding brass bushes in the tiebars and foamed neoprene shims to soak up noise. I will never choose servos again - it will be stall motors or wire-in-tube.

 

It would be worthwhile to do a dummy run with one Peco turnout without its spring before committing to removing all the other springs for your upper level.

 

- Richard.

[ISW]

30 minutes ago, 47137 said:

It would be worthwhile to do a dummy run with one Peco turnout without its spring before committing to removing all the other springs for your upper level.

Richard,

 

I already have 3 point ends on without the spring, not by design though ... I can't recall exactly, but I bu&&ered the spring on a few turnouts on the Lower Level during installation. Plus 1 of the secondhand turnouts on my 'practice' layout / crossover was already missing the spring. As a result I'm reasonably confident regarding my servo installation without the turnout spring. In addition, if I try to move the switches by hand, the piano wire is definitely holding the switch over, on turnouts with and without the spring.

 

Thanks for the observation though.

 

36 minutes ago, 47137 said:

Of course, this concern may be completely unfounded but I am on my fourth iteration of servo mounts without microswitches and reaching my limit of frustration.

 

If it helps, I'm using 6.4mm piano wire and the hole in the baseboard is 8mm (I gather 10mm is the 'norm', but that seemed a tad excessive to me). It's obviously important that the hole is on the track centreline.

 

Ian