Shelf Marshes (first attempt at a cameo layout)

[47137]

I think, there is one detail I need to put my mind at rest:

 

"How much tension should there be in the piano wire when the points are at rest?"

 

I'm guessing, enough tension to move the tiebar another 1 mm or so if the stock rail wasn't there.

 

Perhaps, the 9g servos are so powerful and robust the amount of tension isn't terribly important, as long as it doesn't cause damage to the tiebar.

 

- Richard.

Edited September 30, 2020 by 47137

[ISW]

16 hours ago, 47137 said:

"How much tension should there be in the piano wire when the points are at rest?"

Richard,

 

I'd say about 2-3mm deflection from straight would be about right. Some of mine are probably nearer 4-5mm as I have to ensure contact with the microswitch on one side. The turnouts are all PVAed down so are very unlikely to move laterally. With a 'large' diameter piano wire there is less 'slop' in the tiebar hole, and consequently less chance of damage due to the large contact surface area.

 

Ian

[47137]

I have been aiming for 1 mm or so of tension, I expect I am setting myself too high a bar.

 

Rightly or wrongly I am using 0.5 mm diameter piano wire. I used up all of my 0.6 mm and the 0.4 mm is not strong enough. I have put brass inserts into most of the tiebars, this is a scrap of brass tube:

 

The exception is the turnout where the servo ripped the end off the tiebar. The same servo has run amok again and ripped into the hole in the middle of the tiebar, and I cannot install a bush here:

 

Funny thing is, it still works fine.

 

I will put my neck on the line - I think servos are unsuitable for model railway points because they exert far too much force on the tiebar. We would never experience damage like this with a stall motor.

 

Of course, many people seem to get along with them just fine!

 

- Richard.

[ISW]

3 hours ago, 47137 said:

I will put my neck on the line - I think servos are unsuitable for model railway points because they exert far too much force on the tiebar. We would never experience damage like this with a stall motor.

Richard,

 

I'd say that was true, but only in the circumstances where this is a short distance between the servo arm and the tiebar. When this distance is short, the piano wire will not be long enough to bend sufficiently. Once that distance reaches ~15mm there seems to be no problem as the piano wire is long enough to bend.

 

Ian

[47137]

12 hours ago, ISW said:

Richard,

 

I'd say that was true, but only in the circumstances where this is a short distance between the servo arm and the tiebar. When this distance is short, the piano wire will not be long enough to bend sufficiently. Once that distance reaches ~15mm there seems to be no problem as the piano wire is long enough to bend.

 

Ian

 

I still think my piano wire, most of 36 mm from servo horn to tiebar, worked like a cheese wire.

 

- Richard.

Edited October 1, 2020 by 47137

[47137]

On Parole

 

I think it is time to show my servo mechanisms to the World. I am fed up with altering things, and "they seem to be working at the moment".

 

These two are representative. They are fixed onto a shim of 6.3 mm plywood and a strip of 3 mm foamed neoprene. The ply doubles the thickness of the roadbed (I have a 6.3 mm baseboard top) so I get enough movement at the tiebar. I do not accept it is possible to use this style of mechanism with a 6 mm thick baseboard, and I think claims made to this effect are misleading and wrong. The neoprene does an excellent job of reducing the motor noise transmitted into the baseboard. It is much better than the leather I tried before. The neoprene is compressed just enough to let the two wood screws seat themselves properly:

 

Incidentally, in the beginning I tried to use the screws supplied with the servos to fix the brackets onto the layout. These screws might as well have been made out of butter and gave up after I sheared off the third one I tried. It is really difficult to get the scraps of thread out of the baseboard.

 

I wanted to have eight identical mechanisms but I had to cut the corner off one to fit it up against the baseboard structure. This is the servo I originally installed to use the hole in the end of the Peco tiebar:

 

If servos have one merit, it is the ability to squeeze them into a small space. My baseboard framing is 50 mm deep in total and I would have to cut it away to fit a stall motor here; the servo slips in underneath:

 

The tongue rails (and tiebar) on a Setrack Point have to move further than those on a Streamline point. This was the servo for my Setrack turnout and you can see, I adjusted the end stops of the servo to give me more movement:

 

The next day, the piano wire fell out of the tiebar. So I added a further 4 mm of plywood to let me have a smaller movement on the servo horn:

 

I do not understand how people get away with much more than 100 degrees of servo rotation. There is too much up and down movement on the piano wire, and if the wire is long enough to stay in place in the tiebar then it will foul the trip pins on Kadee couplers.

 

So - this is what I have done. For completeness:

• Peco turnouts laid directly on a 6.3 mm baseboard with no track underlay
• Over-centre springs removed
• 9 mm total shims for Streamline turnouts, 13 mm total shims for the Setrack turnout
• 6 mm diameter hole through the road bed and the shims
• 1.5 mm diameter hole for the fulcrum of the piano wire 
  (I would prefer 1 mm but the 1.5 mm drill bit is stronger for drilling aluminium)
• 0.5 mm diameter piano wire
• One lug cut off each servo to seat the servo in the U channel
• Servos fixed into the channel with a thick, gap-filling cyanoacrylate
• Throw of 6 mm on the servo horn (the second hole out from the axis)
• Servo throw a shade more than the default MegaPoints settings, so about 95 or 100 degrees.

I think, I have done much as what others have written up, but I have found the exercise dreadfully difficult. If anyone fancies having a go, I suggest building at least one complete mechanism and testing it thoroughly before blindly copying what you see here and indeed elsewhere online.

 

- Richard.

[ISW]

3 hours ago, 47137 said:

Servos fixed into the channel with a thick, gap-filling cyanoacrylate

Richard,

 

Wow. I don't think your servos will be coming out anytime soon. MegaPoints 'recommended' just using a friction fit in the aluminium U-channel. I upgraded that by adding 2 nuts / bolts at the ends of the U-channel to 'clamp' the servo in place. You have taken it to another level ...

 

Ian

[ISW]

3 hours ago, 47137 said:

If anyone fancies having a go, I suggest building at least one complete mechanism and testing it thoroughly before blindly copying what you see here and indeed elsewhere online.

Richard,

 

I agree completely. I did that on my 'practice' crossover and learnt quite a few lessons.

 

Ian

[47137]

These are the MegaPoints relay boards I am using to change the frog polarities.

 

In my MegaPoints system, the relay driver is controlled by the servo controller. The eight relays are synchronised to eight of the servo channels.

 

This is one of those wiring tasks I find much easier to do by just connecting everything up and then debugging it:

 

I swapped around the wires on the relay terminals until everything worked, and then wired it all properly:

 

Later, I decided to reverse some of the servo actions (this is the MegaPoints "mode 2") and discovered I had to rewire all of the associated relays to match. However, all of my turnouts now have natural "normal" positions when there are no local levers thrown.

 

I think this is all a bit over the top in terms of hardware, but if I used double-pole toggle switches to control the servo channels and the frogs, I would deny myself the opportunity to use my route setting board and indeed any ability to control the layout remotely via the MegaPoints bus.

 

I have fixed the two relay boards onto a floating panel attached to the baseboard by four thin pins. The idea here is to reduce the relay noise transmitted into the baseboard, and I think it has a modest effect.

 

- Richard.

[ISW]

2 hours ago, 47137 said:

Richard,

 

You really need to invest in some multi-core cables .... It'll make wiring easier, simpler to troubleshoot, and tidier. I use 6-core cable from Hobbytronics (just a customer ...) like this https://www.hobbytronics.co.uk/ribbon-cable-6core-15ft?keyword=cable

 

Just a thought.

 

Ian

[47137]

16 hours ago, 47137 said:

 

13 hours ago, ISW said:

Richard,

 

You really need to invest in some multi-core cables .... It'll make wiring easier, simpler to troubleshoot, and tidier. I use 6-core cable from Hobbytronics (just a customer ...) like this https://www.hobbytronics.co.uk/ribbon-cable-6core-15ft?keyword=cable

 

Just a thought.

 

Ian

 

Do bear in mind, this was a lash up to make everything work. It lasted less than a day. I find it easier to connect everything up and swap pairs of wires to the relays as needed to get the polarities right. If I wrote everything down and worked to a plan, knowing my luck I would still end up swapping wires.

 

I do have some multi-core cables to hand for the layout. I will use them where I need to take a set of connections from one location to another. This will be from the servo controller to the route setting panel, and from the DB25 connector to the control panel.

 

I had three engines on the layout a few days ago, and convinced myself the wiring to the track is complete. The most important model is my Matchbox MB-24, because it runs over all of the track without stalling. This is before I clean the tops of the rails too.

 

https://www.rmweb.co.uk/community/index.php?/blogs/entry/18275-mb-24-no-5-matchbox-Bachmann/

 

If the MB-24 copes, everything else should follow.

 

- Richard.

 

 

[ISW]

12 minutes ago, 47137 said:

Do bear in mind, this was a lash up to make everything work. It lasted less than a day. I find it easier to connect everything up and swap pairs of wires to the relays as needed to get the polarities right. If I wrote everything down and worked to a plan, knowing my luck I would still end up swapping wires.

Richard,

 

Apologies for the misunderstanding, I'm not really a 'lash up' type of person. If I did that I'd end up spending more time fixing the lash up than I would installing the final cables! Each to his own though.

 

12 minutes ago, 47137 said:

The most important model is my Matchbox MB-24, because it runs over all of the track without stalling. This is before I clean the tops of the rails too.

That's a really good 'testing' loco, nice short wheelbase. As you say, if that works without stalling then any loco will be fine.

 

Somewhere, probably in my parents attic, I have a pair of those Triang 'Dock Authority' locos:

Remember those? Absolutely ridiculous locos with huge pulling power (thanks to the 'ribbed' wheels) and a top speed way off the chart (certainly high enough to cause overturning on tight bends). Maybe I should retrieve it (Covid19 Rules permitting ...), convert to DCC and give it run on my layout. Totally un-prototypical but, I imagine, loads of fun!

 

Ian

[47137]

1 hour ago, ISW said:

Somewhere, probably in my parents attic, I have a pair of those Triang 'Dock Authority' locos:

 

Remember those? Absolutely ridiculous locos with huge pulling power (thanks to the 'ribbed' wheels) and a top speed way off the chart (certainly high enough to cause overturning on tight bends). Maybe I should retrieve it (Covid19 Rules permitting ...), convert to DCC and give it run on my layout. Totally un-prototypical but, I imagine, loads of fun!

 

Ian

 

A friend had one, we would have been about 10 years old. I envied it at the time because of the headlight.

 

I think it would take a taller cab and make a nice 3 foot gauge shunter for 1:55 scale, but those wheels would have to go!

 

- Richard.

[ISW]

2 hours ago, 47137 said:

but those wheels would have to go!

Richard,

 

What's wrong with the wheels?; they are all metal (no traction tyres), serrated to give excellent grip, and power pickup was never an issue (the loco was surprisingly heavy for its size).

 

Me wonders if you can get them as a retrofit for old Lima/Hornby locos .... 

 

Ian

[47137]

12 minutes ago, ISW said:

Richard,

 

What's wrong with the wheels?; they are all metal (no traction tyres), serrated to give excellent grip, and power pickup was never an issue (the loco was surprisingly heavy for its size).

 

Me wonders if you can get them as a retrofit for old Lima/Hornby locos .... 

 

Ian

 

Just a matter of taste I suppose!

 

There was a smooth wheel variant:

https://www.elaines-trains.co.uk/index.php?cat=20211

 

- Richard.

[ISW]

2 hours ago, 47137 said:

There was a smooth wheel variant:

Richard,

 

Now that I didn't know. Unless the previous user(s) used the loco to the point where the ribs were worn away.

 

Ian

[47137]

I put the lighting rig into the build today. The lighting rig itself is a strip of L-section aluminium holding two tapes of LEDs. These tapes are dual warm and cool white ones, i.e. two sets of LEDs on the tape. The idea being to be able to vary the lighting effect from "dawn" to "noon" to "dusk".

 

This is the general arrangement. I have "made" a dedicated control panel for the lighting and incorporated this into the left wing. The knobs are, like taps, hot and cold; the toggle switch is for lighting in the model buildings and street lamps and so on:

 

This is the arrangement on the inside. You can just about see the lower of the two LED tapes:

 

The LED tapes are 1.4 m long and they consume about 0.6A per circuit on full power, so up to 2.4A in all. I doubt I will want the layout to be quite this bright, but I have included a time delay module in the supply to the lighting circuit to let the MegaPoints servo board take as much of the 3 amps available as it needs while it initialises itself and its servos.

 

The time delay module has a 555 timer and a relay, and I put it at the top of the wiring:

 

Underneath this are two speed controller modules intended for electric motors. These use pulse width modulation so they are ideal for the LED tapes. One module ("hot") drives one circuit of warm LEDs. The other module ("cold") drives the second set of warm LEDs and both sets of cool LEDs. The terminal block top right is one of those "double" designs with a set of pins and sockets. This comes apart to let me take out the lighting rig so I can actually build the layout.

 

I cannot take a meaningful photo of the overall effects possible, but the lighting is looking good to me.

 

- Richard.

[ISW]

Richard,

 

That looks very neat and tidy. I like the idea of dimmer switches on the LEDs. Where did you get those modules from?

 

27 minutes ago, 47137 said:

but I have included a time delay module in the supply to the lighting circuit to let the MegaPoints servo board take as much of the 3 amps available as it needs while it initialises itself and its servos

 

Although I have a 14.2amp 12v power supply brick (ex Xbox), I was still planning on having the LED lighting on a separate power supply (2amp 12v). Any reason for putting both on the same power supply?

 

Ian

[47137]

27 minutes ago, ISW said:

Richard,

 

That looks very neat and tidy. I like the idea of dimmer switches on the LEDs. Where did you get those modules from?

 

 

Although I have a 14.2amp 12v power supply brick (ex Xbox), I was still planning on having the LED lighting on a separate power supply (2amp 12v). Any reason for putting both on the same power supply?

 

Ian

 

It will be tidier after I lift out the scrap of green insulation visible near the relay in the last photo!

 

These are the dimmer modules, under a fiver for two, supplied by "dolphin114" through eBay:

https://www.ebay.co.uk/itm/Mini-DC-Motor-Speed-Control-Driver-Board-3V-35V-5A-PWM-Controller-LED-dimmer/163091515028?ssPageName=STRK%3AMEBIDX%3AIT&var=462236844167&_trksid=p2057872.m2749.l2649

 

I have one power supply mainly because I want to take this layout to shows one day. So one mains lead to run everything. Also, MegaPoints tell me the servo controller draws up to 2A during its initialisation.  I'd much rather have the delay module than a second power supply, surely this is much more elegant. When I power up the layout, the servo controller takes about two seconds to ready itself and its own associated relays; and the lighting rig fires up another two seconds later. Just right.

 

At the moment, things are looking good. The lighting rig is taking perhaps 1.3A for sensible illumination leaving say 1A for a DCC controller (one train in steam), 0.3A for relays, 0.2A for the lights in the model buildings and street lights, and a bit of headroom. I will guess, the power supply unit will give its best regulation and run at its best efficiency at its rated output too.

 

Although I suspect in practice, I might be using a different PSU for the DCC controller simply so I can give it more than 12 volts.

 

- Richard.

[47137]

I have built one of these voltmeter/ ammeter modules into the layout so I can keep an eye on the load I am putting on the PSU:

https://www.ebay.co.uk/itm/Digital-LED-100V-Voltmeter-10A-Ammeter-Panel-mount-volt-amp-meter/152885694791?ssPageName=STRK%3AMEBIDX%3AIT&var=452642839382&_trksid=p2057872.m2749.l2649

 

Also from "Dolphin114".

 

A lot easier than connecting an ammeter into the circuit.

 

- Richard.

[ISW]

Richard,

37 minutes ago, 47137 said:

It will be tidier after I lift out the scrap of green insulation visible near the relay in the last photo!

 

Nah, didn't spot it. The rest was too tidy ...

 

37 minutes ago, 47137 said:

These are the dimmer modules, under a fiver for two, supplied by "dolphin114" through eBay:

https://www.ebay.co.uk/itm/Mini-DC-Motor-Speed-Control-Driver-Board-3V-35V-5A-PWM-Controller-LED-dimmer/163091515028?ssPageName=STRK%3AMEBIDX%3AIT&var=462236844167&_trksid=p2057872.m2749.l2649

Thanks for the link. Think I'll be buying a few myself.

 

38 minutes ago, 47137 said:

leaving say 1A for a DCC controller (one train in steam) *snip* 

 

Although I suspect in practice, I might be using a different PSU for the DCC controller simply so I can give it more than 12 volts.

I have an NCE PowerCab at present, so I can't quite see how to use a single 12v power supply for DCC, MegaPoints, & Lighting. As I mentioned, my power supply is more than capable, but I don't see how to use it for DCC as well.

 

If there is a way to 'boost' my NCE PowerCab from its current (sorry, no pun intended) maximum of ~2amp to nearer 5amp using my 14amp power supply I'd be very interested to hear about it.

 

Ian

[47137]

On 04/10/2020 at 21:29, ISW said:

I have an NCE PowerCab at present, so I can't quite see how to use a single 12v power supply for DCC, MegaPoints, & Lighting. As I mentioned, my power supply is more than capable, but I don't see how to use it for DCC as well.

 

If there is a way to 'boost' my NCE PowerCab from its current (sorry, no pun intended) maximum of ~2amp to nearer 5amp using my 14amp power supply I'd be very interested to hear about it.

 

My DCC controller is a Roco z21 Start and I am running this from a 13.8V DC supply. I had an idea, I could use 12V instead but looking at the back of the z21, it expects 18 to 24V. So I expect my 12V is a bit too low; but I could experiment.

 

The NCE PowerCab is part of a larger system, and if I have got this right, expansion involves buying a base unit able to deliver the power to the track, and configuring the PowerCab as a handset able to control this base unit. I don't really know anything about using DCC on larger layouts, but there are plenty of discussions here:

https://www.rmweb.co.uk/community/index.php?/forum/23-dcc-discussion-topics-not-questions/

 

Hope this helps.

 

- Richard.

[47137]

I have completed the "proscenium" (if this isn't too grand a word!) by adding a strip of ply across the top, and also dropped in a dust cover:

 

The lighting rig is running at about half brightness and drawing about 1.1A, so I will have plenty of capacity to spare with my 3A power supply.

 

I fixed the middle panel of the front onto the lighting rig, and the two lift out together to let me do the model making inside.

 

The bit of flex in the foreground is a multi-core cable for my temporary control panel - a terminal block with some croc clip leads to let me change the points. This is driving me mad and some kind of panel with switches is called for.

 

- Richard.

[ISW]

2 hours ago, 47137 said:

 

The bit of flex in the foreground is a multi-core cable for my temporary control panel - a terminal block with some croc clip leads to let me change the points. This is driving me mad and some kind of panel with switches is called for.

Richard,

 

Yes, I know the feeling. I had a MegaPoints controller without switches, just bridging the contacts with a screwdriver. Very frustrating.

 

I'm assuming that the back / sides / top of your "proscenium" are detachable for access to install the scenery?

 

Ian

[47137]

10 hours ago, ISW said:

I'm assuming that the back / sides / top of your "proscenium" are detachable for access to install the scenery?

 

The backscene is fixed, it is a structural part of the baseboard. The two wings (front left and front right) are fixed too. The dust cover is a loose piece of 4 mm ply.

 

The lighting rig (the strip of L shaped aluminium) becomes a structural part when it is in place, it tightens everything up. I am using four screws, two at each end, to hold it in place. But it has to come out to let me build the layout, so it slides up vertically like this:

 

 

My textbook for the baseboard is "Creating Cameo Layouts" by Iain Rice. He suggests a monocoque construction, but admits the arch and lighting rig need to be detachable. Really, my baseboard only becomes a true monocoque if I fix the lighting rig permanently - and I cannot imagine doing this.

 

The biggest difficulty building adding this section has been aligning the two wings to the top of the arch so the front presents a consistent flat surface. I cannot use screws through into the aluminium because the LED tapes are on the inside, so I ended up creating two vertical slots. I have offcuts of ply on the inside and narrow strips of wood on the outside. This is fairly clear in the first photo here.

 

- Richard.