Town and country - automated N-scale layout


This was unexpected :)

After looking through my track bin more carefully, I was able to salvage 4 Atlas Code 55 #5 turnouts along with a few Micro Engineering code 55 #6. I also discovered that I have some brand new Micro Engineering code 55 flex track! So, I basically have most of what I need to build this layout. Only thing is I have to see if the deeper flanges in European N-scale rolling stock can run on code 55 track - hopefully this is no longer a problem.

IMG_6780.jpeg
 
2 steps forward, one step back 🤷‍♂️

I am glad I tried this experiment: I tried running a Kato Glacier Express passenger car through an Atlas code 55 turnout and it did not work well. The car's flange are sufficiently deep that it rode up on the frog - not enough to derail but it looks bad. On the other hand, the car went through the Micro Engineering #6 turnout with no problem at all.
In the pic below: Atlas code turnout is on the mat and the Micro Engineering turnout is on the table. Have to rethink this now...

IMG_6785.jpeg
 
As I mentioned earlier, Kato European rolling stock works fine with Micro Engineering N-scale turnouts. The great thing about Micro Engineering turnouts is that they have an isolated metal frog. The only question in my mind was whether I could solder a wire to the frog. I am happy to report that it was not hard at all.
I was able to solder a wire to the little metal nubbin. I cut a slot in one side to allow the wire to exit and then I filed down the solder joint. It has good electrical connectivity. The frogs will be powered by a Tam Valley Frog Juicer. I used this in my HO scale layout and it worked great there.

IMG_6786.jpeg

IMG_6788.jpeg
 
Last edited:
Next step is to figure out a clean way to connect linear servos to the Micro Engineering turnouts. These linear servos are meant for model aviation use but they ought to be great for model train turnouts too. I will be controlling the linear servos from a Raspberry Pi. Its very easy to program these single board computers to generate the pulse width modulation signals for servo control.


IMG_6795.jpeg
IMG_6796.jpeg
 
Iron Horseman said:
My opinion is that a number 8 is large for all but around the room, filling the basement, or filling the attic, or separate building type layouts.
Click to expand...
Completely agree - I just can't use a #8 in a layout as small as this one.

My next step now is to build a small test layout to find the best way to install a turnout servo motor on an extruded foam base. I have more than enough track for that so let's see how it goes.
 
In order to figure out how to mount the the servomotors for the turnouts, I am creating a small testbed on a piece of extruded foam:

IMG_6804.jpeg


What I am thinking is the rather than dig out a hole to mount the servomotors, I will build up the landscape around the servomotors. Need to go and get some more thin foam sheets. I also need to accurately bend some piano wire to connect the servomotor with the throwbar

IMG_6805.jpeg
 
I have finished building the testbed to figure out the mechanical and computer tricks to get this linear servo to work as a switch motor.
I used a piece of stiff piano wire to connect the servo to the turnout thrower but this still needs some experimentation. Also, I wrote a little Python code to control the servo and this is also still in the early stages as I have to find the right parameters.

IMG_6807.jpeg


IMG_6808.jpeg
 
After a few days of experimentation, I just gave up with the micro-servo idea for two reasons.
  • I was not able to build a clean simple linkage between the servo and the throwbar of the turnout. The wire linkages looked really clunky and did not work well.
  • I was just not able to precisely control the servo using software PWM signal (Pulse Width Modulation) generated by the Raspberry Pi. Now, I could get some hardware assist. Adafruit makes a servo controller board for example, or I could use an Arduino. But this just makes the overall system more complex.
I have more than a dozen Tortoise switch motors left over from old layouts so I decided to save some money and just figure out a clean way to install the Tortoise in 2" foam. The other big advantage of Tortoise is that I can use the additional contacts for detecting the position of the switch at startup time from the computer.

I created a couple of templates using 0.060" thick styrene sheet and it all came together very easily. I will be screwing the tortoise to a couple of styrene sheets and then I will glue the styrene to the foam. Hopefully, I ought to be able to replace a tortoise if it stops working.
  • IMG_6814.jpeg
 
Good luck with the tortoise machines! I like the concept of a slow moving switch, but didn't want to fork out the cash, and since I was using Peco turnouts, I ended up using their electric switches that attach directly to the underside of the track. So all I had to do was carve out a hole in the foam for the switch to sit in. They work great, but do make a noticeable click when you actuate the points.

I envy people who have complete digital control of their layout. Looks like that's where you are heading! I may haed in that direction, now that I acquired a more robust DCC system that can interface with JMRI.

Please keep posting updates!
 
JeffH said:
Good luck with the tortoise machines! I like the concept of a slow moving switch, but didn't want to fork out the cash, and since I was using Peco turnouts, I ended up using their electric switches that attach directly to the underside of the track. So all I had to do was carve out a hole in the foam for the switch to sit in. They work great, but do make a noticeable click when you actuate the points.

I envy people who have complete digital control of their layout. Looks like that's where you are heading! I may haed in that direction, now that I acquired a more robust DCC system that can interface with JMRI.

Please keep posting updates!
Click to expand...

This layout will indeed be under full digital control. One of the main intentions of this project is to run a few trains automatically on the layout. So, in a way, it is the direct opposite of my other layout (Woodlawn Industrial Park) which is a completely manually operated switching layout.

I want to make this layout lightweight, self-contained and portable so I can take it to train shows. The control computer is going to be a Raspberry Pi driving the layout through a Sprog 3 DCC. I will be writing the control software in Python - so lots more to come :)
 
I assembled the first mounting plate for the Tortoise and used some small wood screws to fix the tortoise onto the mounting plate. I then cut a rectangular hole in a piece of foam. I assembled it all together and it all works well. Note that I had to laminate two sheets of 0.060" styrene so that it was the same height as the Midwest products N-scale cork roadbed. I don't like the way the screws stick out so I need to find a better solution.

IMG_6815.jpeg


Not bad for a first attempt but I have to see if I can remove the projecting screws. Maybe try some nuts and bold instead. Also, it just occurred to me that I should change the orientation of the tortoise so that more of the mounting plate is under the turnout.

IMG_6816.jpeg
 
I just received the Sprog 3 that (along with a Raspberry Pi) is going to be the DCC base station for this layout. The Sprog 3 will be controlled by JMRI running on the Raspberry Pi. A Sprog 3 can put out 2.5 amps, which is way more than I will need for such a small layout.

IMG_6817.jpeg
 
I discovered that I have a few sheets of very high quality basswood plywood meant for scratch building wooden models. So I tried creating a Tortoise mounting plate using this plywood and it works much better than styrene. Firstly its more rigid and secondly the mounting screws bite into wood much better than they do into styrene.

IMG_6818.jpeg
 
I think I have most of the infrastructure for this layout worked out, the picture is getting clearer:
  • Trains will run under DCC and a Sprog 3 (along with JMRI software) will be the controller
  • A single Raspberry Pi will be the control computer
  • I will write my control software in Python to work with JMRI
  • I will use either 2 Wabbit turnout decoders or 1 DS64 to drive the turnouts via DCC
  • Tortoise switch motors will be used.
  • Train location feedback will be via reed switch sensors buried under the tracks. I will not use current draw detectors.
  • All sensors will be directly connected to GPIO pins on the Raspberry Pi. I may need to add debounce/monostable but this remains to be seen. I might be able to handle debounce in software itself.
  • I will mount small rare earth magnets in the trains to trigger the reed sensors
  • I will use Atlas code 55 10" radius sectional track sections for the loops at the end of the layout. From my experience is that flex track does not work well for such sharp curves.
  • I will use some Micro engineering code 55 flex track that I have on hand for the other track - might need to buy a little bit more
  • Turnouts will be Micro engineering code 55 #6 turnouts. I will power the frogs from the extra contacts on the Tortoise
  • This time, I will definitely use real rock ballast from Arizona Rock and Mineral. I really don't like the Woodland scenics crushed walnut shells.
  • I just discovered that Joe's Model trains has their rusty rail paint back in stock so I will order that again. I am delighted! This paint is absolutely perfect for rails.
 
I was not happy with the rough cuts that I was making in the foam base so I went over to a local Michael's and got a hot foam cutter. It works really great on 2" foam. I hated how ugly the holes look even though its covered by the mounting plate - and its very quick!

IMG_6821.jpeg
 
kleiner said:
I discovered that I have a few sheets of very high quality basswood plywood meant for scratch building wooden models. So I tried creating a Tortoise mounting plate using this plywood and it works much better than styrene. Firstly its more rigid and secondly the mounting screws bite into wood much better than they do into styrene.
Click to expand...
Give Luan plywood a look too, usually stronger than basswood with multiple ply's in 1/4" stuff.
 
ctclibby said:
Give Luan plywood a look too, usually stronger than basswood with multiple ply's in 1/4" stuff.
Click to expand...
If I was building this layout in HO scale, Luan plywood would have been my choice. But the thinnest Luan plywood I have seen in the big box stores is 1/4" thick. This plywood is just 1/8" thick - which works out to be exactly the height of the Midwest Products N-scale roadbed. Also, I have a lot of this stuff lying around anyway :)
 
You must log in or register to reply here.



Back
Top