Trewartha Quay (00)

[37114]

Looking forward to seeing some more progress Alan, you know you are more than welcome bring it round one evening so Dave and I can give you a hand!

[aforsyth]

Thanks Rob. I will probably take up your generous offer of assistance in the new year!

 

The photo below shows the current state (actually back in October, but I've had no time since then).

 

 

A number of minor but significant steps were taken in Sept/Oct:

 

- All timber and baseboards were completely coated in a 2:1 mix of water to waterproof PVA, including the legs while removed from the boards. Hopefully this will protect them somewhat, although eventually I will paint it all a suitable colour.

- All trackbed supporting uprights have a layer of 1/8" cork glued between them and the trackbed using Copydex, hopefully to reduce noise travelling down to the frame structure. I was concerned that this would not be a very strong joint between supports and trackbed, but it seems pretty strong when physically pulling on it. However, I have a couple of small screws holding each support to the trackbed as a backup, although these obviously create a path for noise/vibrations to some extent.

- M6 adjustable height 'feet' were added to all legs (see photo below). I got 12 of these for £8 from a very nice seller on Ebay, who even threw in the Allen key required!

- 1/8" cork sheet glued on top of the trackbed using Copydex for track height / shoulder where appropriate and again to reduce noise.

 

Here's an example of the 'feet' I bought, which I'm quite pleased with. I don't have a nice drill stand to drill perfect vertical holes, but after a bit of practice with the drill on scrap wood and with the use of some hastily drilled guide pieces, the perpendicular holes in the legs were accomplished.

 

 

Finally, I succumbed to temptation and picked up an Arduino Uno microprocessor board, along with an Adafruit 16 channel servo board and 9 model servos to power my points.

I've never played with this stuff, nor done anything with electronics in general since my school days decades ago, so I decided to first build a prototype unit, to test this out on.

This is shown below, with an unseen servo mounted below the point using 'U' shape aluminium channel, and an additional temporary switch board for simulating the control panel.

It works quite nicely, but movement is still too fast for my liking, so I will need to adapt a popular Arduino 'slowdown' library, which will enable prototype switching operations to be carried out.

Total cost of the point switching equipment is about £50 (including boards, 2 power supplies, 9 servos, mounting materials), but this could be significantly reduced with Arduino 'clone' boards instead of originals, which I may use next time.

This is a winter project, but I may try to devote a separate thread to it in the future once it's ready for mounting on the layout.

 

 

 

So - next step is track laying, which I hope to be able to start over the Christmas holidays, if time allows. I also need to finalise the electrical wiring plan before final trackwork is fixed, so thank you to all who made very helpful suggestions on this earlier in the thread.

 

Finally, a Happy Christmas to all readers here. Hope Santa / nearest and dearest brings you some nice surprises! 

 

Alan

[47137]

It is so nice to read some fresh ideas like the cork "buffer" below the track bed.

 

I'll look forward to fresh instalments on this layout.

 

- Richard.

[aforsyth]

Thanks Rob. We'll see if it makes any difference (or I suppose whether there was any point in doing it at all!). My last larger layout attempt used standard pine legs & frame with MDF top, but both that and previous attempts in the 90s with chipboard tops appeared pretty noisy; admittedly this is the first time I'm using cork sheet.

 

Here's the detail of the cork 'buffer' before I glued/screwed the trackbed back on:

 

 

 

I've also taken a quick photo of the underside of the 'prototype' point controller unit, showing the mounted motor. The mini baseboard here actually served to test the theory of using the cork buffers on wood initially. The left batten used Copydex for fixing, while the right batten used standard PVA. It was immediately apparent that the PVA did indeed harden the cork after drying, whereas the use of the Copydex apparently retained much of the cork's original properties.

 

 

It may not be totally clear in the photo above, but there's a microswitch screwed to the right side of the aluminium 'U' channel, for frog activation of the electrofrog points. 1mm piano wire is bent appropriately to connect the servo to the point tiebar and used to create a springy arm to hit the switch, which works but takes quite a lot of work to get right. If anyone has suggestions, I'm certainly open to them!

 

Alan

 

[Edit: improved servo pic]

[47137]

You could try a SPCO reed switch on a bracket beside the servo, with a magnet on the servo arm. This should be easier to set up.

 

For the switches try eBay, here is an example.

 

- Richard.

[aforsyth]

Hmmm... that's a brilliant idea, Richard, thank you. The only problem is that the changeover type seem quite expensive (£2 each on Ebay + P&P, thanks for the link), compared to the standard NO / NC type for which you can get about 10 for the same price.

On the other hand, extra relays would cost, too, and require extra wiring.

 

I had thought about 'doing away' with these mechanical frog switches at one point, and simply using Arduino-controlled relays on a single board to switch power, which would be cheaper (e.g. an 8 relay board costs around £7 inc p&p):
e.g. http://www.ebay.co.uk/itm/5V-1-2-4-8-Channel-Relay-Board-Module-for-Arduino-Raspberry-Pi-ARM-AVR-DSP-PIC-/252051910091

 

However, again, this would require more wiring, even on a small layout, so the changeover reed switches are probably the best way forward, which I'd never thought about. I'll try to pick up a couple soon to experiment with.

 

What kind of magnets would you use for these, and how would you fix them to the servo arm - just with superglue?

 

Thanks again - great advice.

Alan

[47137]

I haven't tried this, but I imagine you could have two NO type reed switches, one actuated at each end of the servo travel. For me, I'd rather spend the extra money on one switch, less wiring, fewer adjustments.

 

I have a Neodymium magnet about 6 mm diameter and 3 mm thick operating a reed switch on my fiddle yard no problems at all. For fixing the magnet I would reach for the hot glue gun, this seems to fix unlikely combinations of materials together. Smaller magnets would almost certainly work, but they will need more careful alignment over the switch.

 

There is an alternative I would like to try (or have someone else try!!) which is to convert Peco electrofrogs into dead frogs. This would mean slicing across the metal crossing with a cut-off disc, leaving the running rails as long as possible but not quite touching each other. Fill in the gaps with filler or Araldite. No switching required, and the surgery will disappear below paint on the sides of the rails. The success of this will depend on how many wheels on the locos and how many wheels have pick-ups. Not recommended for 0-4-0s but should work perfectly with modern bogie diesels. 0-6-0 locos (which are the majority of my own locos) would be a bit of a gamble.

 

- Richard.

[mjcampbell]

No, stick to switched live-frog points, they are well worth the extra effort for smaller locos running well. Larger micro-switches might be easier to set-up - with longer travel they will be less sensitive to position, and a tweak of the arm or the actuating wire with pliers should be all that's needed. Set the switch so the actuation is near the end of the arm.

 

The arduino set-up looks interesting, can you tell us more about it? What's the operator interface - buttons on a panel? tablet or PC screen? How do you code it?

[nigelb]

Just a thought  but instead of cork between the supports   have you considered  laminate flooring underlay to fix the base board down a screw through the center of the support  in a clearance hole into another  block below the support  this would allow you to tighten the track-bed down to slightly compress the underlay

hope that makes sense

[rue_d_etropal]

I would not recommend cutting up points, but just using dead frog points with graphited rail works well.  As long as the physical load is not too much, then the extra slippery track should not cause any problems. Ideal for shunting layouts, and known about by many, but rarely mentioned in general forums. I also wire up points so not dependent on power going through tips. Standard practice on continental type points.

As long as the plastic part of the frog is exactly the same level as the metal rails, then at least one pair of pickup wheels should stay in contact. It is not the pap which is the problem but the unevenness in track which result is not all heels being in contact with the rail. That is simple physics and maths. The plastic and metal will probably expand and contract at different rates so that might be the main cause of pickup problems. It is possible to put some graphite on the plastic, and it seems to help as well.

Having locos properly warmed up is also important.

[aforsyth]

Thank you all - responses inline.

 

I haven't tried this, but I imagine you could have two NO type reed switches, one actuated at each end of the servo travel. For me, I'd rather spend the extra money on one switch, less wiring, fewer adjustments.

 

I have a Neodymium magnet about 6 mm diameter and 3 mm thick operating a reed switch on my fiddle yard no problems at all. For fixing the magnet I would reach for the hot glue gun, this seems to fix unlikely combinations of materials together. Smaller magnets would almost certainly work, but they will need more careful alignment over the switch.

 

- Richard.

 

Thanks Richard. I think in the end you're right - I'm all for a simpler life if it just costs a few more pounds. The glue gun is a good idea, I've been meaning to get one for some time. Hopefully B&Q might have one on sale now!

Regarding the dead-frog modification, I'd be interested to hear how that goes if anyone tries it. One of the goals of this layout was to try out new technology (for me), which includes using live frogs, so I'll battle on for now, and consider my options for the next layout based on my experiences.

 

No, stick to switched live-frog points, they are well worth the extra effort for smaller locos running well. Larger micro-switches might be easier to set-up - with longer travel they will be less sensitive to position, and a tweak of the arm or the actuating wire with pliers should be all that's needed. Set the switch so the actuation is near the end of the arm.

 

The arduino set-up looks interesting, can you tell us more about it? What's the operator interface - buttons on a panel? tablet or PC screen? How do you code it?

 

Thanks Michael. Larger micro-switches may certainly help. Getting the holes aligned correctly (and finding tiny screws) for the switch was quite challenging, although it would be easier next time now.

I'll maybe test out the reed switch idea, as theoretically they could last longer than cheap micro-switches (I have no evidence to back that up though, apart from them having fewer components/moving parts).

 

Arduino Point Control

The Arduino Uno has the Adafruit 16 channel PWM shield plugged in to it, and at present I have done little more than customise the Adafruit library example code to control the point.

I have set up 3 switches at present on this test rig, which we can call 'L', R', and 'S'. When the Arduino is switched on, the servo resets to the middle (I have removed the point spring to allow this), so I currently have to calibrate it each time it's switched on. I press the 'L' switch to move the servo in small increments to the maximum position I want it to go to the left, then the 'R' switch to set the corresponding maximum position to the right. The Arduino remembers these settings, and the 'S' switch then moves it at normal speed backwards or forwards between these two positions as required. This basic functionality will be updated during the next stage to save the settings indefinitely when powered off, and to enable 2 or 3 'management' buttons to work for any number of points.

 

I'd be happy to share the current simple Arduino code with anyone who's interested (I don't want to pollute this thread with it, but can perhaps post it to the Arduino thread). Eventually when it's robust and stable, my goal would be to write an article to explain how to set this up for a newbie like myself, and post the code on GitHub or similar, so that it can be copy/pasted with minimal modification. Quite a challenge, but it's such a cheap and flexible point control solution, the community needs this available. Conversely, if anyone knows of a similar existing 'open source' Arduino/Pi point control solution, I'd be glad to try it and perhaps provide feedback or improvements if necessary.

 

Control Panel

Regarding the layout control panel, while musing on the track electrical plan and building the point switcher prototype, I realised I hadn't given much serious thought to the control panel.

The main complication I've thrown in is that I'd like it to be movable (via plugs & sockets or wireless), to allow me to control from front or back, fiddleyard or main scenic area. A further spanner in the works is the requirement for DC and DCC, necessitating different controllers to be added, even though the main panel should work with either. I'm therefore trying to think through the simplest arrangement of modular controllers and control panel design to accommodate this, without making the whole project too complicated or costly!

 

On the DC side, I have a faithful old Gaugemaster DS, which I plan to use. For DCC, at present I have only a Hornby Elite (picked up nearly new for £50 a couple of years back!), which is fine for testing but won't be ideal for operation, particularly for sound locos. Cost is an important factor, so other than a second-hand NCE system, I'm also considering a new or secondhand Sprog 2/3 and JMRI on a PC or tablet. With the latter, although touch screens and 'point & click' have benefits, I do yearn for a lever frame and decent control knobs, so I need to make a decision here very soon, which will influence the control panel design and thus the layout build.

 

 

Just a thought  but instead of cork between the supports   have you considered  laminate flooring underlay to fix the base board down a screw through the center of the support  in a clearance hole into another  block below the support  this would allow you to tighten the track-bed down to slightly compress the underlay

hope that makes sense

 

Thanks Nigel. Yes - underlay was my initial plan, but then I noticed I had this big roll of cork in the garage from the previous layout attempt, which was a shame not to use! I've heard floor underlay is very good, so that's something I may try next time, so I can compare with the cork used here.

 

I would not recommend cutting up points, but just using dead frog points with graphited rail works well.  As long as the physical load is not too much, then the extra slippery track should not cause any problems. Ideal for shunting layouts, and known about by many, but rarely mentioned in general forums. I also wire up points so not dependent on power going through tips. Standard practice on continental type points.

As long as the plastic part of the frog is exactly the same level as the metal rails, then at least one pair of pickup wheels should stay in contact. It is not the pap which is the problem but the unevenness in track which result is not all heels being in contact with the rail. That is simple physics and maths. The plastic and metal will probably expand and contract at different rates so that might be the main cause of pickup problems. It is possible to put some graphite on the plastic, and it seems to help as well.

Having locos properly warmed up is also important.

 

Wow Simon - sounds like great science theory, thank you! Again, as mentioned earlier, I am going to stick with live frogs this time, just to get some experience with them, but I've seen some impressive slow running over well maintained dead frogs, so again that might be something to try next time.

 

Alan

[nigelb]

Hmmm... that's a brilliant idea, Richard, thank you. The only problem is that the changeover type seem quite expensive (£2 each on Ebay + P&P, thanks for the link), compared to the standard NO / NC type for which you can get about 10 for the same price.

On the other hand, extra relays would cost, too, and require extra wiring.

 

I had thought about 'doing away' with these mechanical frog switches at one point, and simply using Arduino-controlled relays on a single board to switch power, which would be cheaper (e.g. an 8 relay board costs around £7 inc p&p):

e.g. http://www.ebay.co.uk/itm/5V-1-2-4-8-Channel-Relay-Board-Module-for-Arduino-Raspberry-Pi-ARM-AVR-DSP-PIC-/252051910091

 

However, again, this would require more wiring, even on a small layout, so the changeover reed switches are probably the best way forward, which I'd never thought about. I'll try to pick up a couple soon to experiment with.

 

What kind of magnets would you use for these, and how would you fix them to the servo arm - just with superglue?

 

Thanks again - great advice.

Alan

you could use a no  and a nc  reed  wired up together

              _____/ ____

----------{_____ _ ___