Frequently Answered Questions for rec.models.railroad


Home Site Map

Layout building


All contributions and corrections are welcome.

The FAQ is maintained by Urban_Fredriksson@icl.se
The HTML version of the FAQ is maintained by Tony Lupton (t.lupton@citr.uq.oz.au)
The HTML version of the FAQ is hosted by Rich Weyand (weyand@rcnchicago.com)


Introduction

02 Aug 94 14:00:00 GMT

This digest contains the following parts:


Layout, curves

20 Mar 94 01:00:00 GMT

Cork, radius

How do I get the cork to go around curves?

Cork roadbed is typically provided in a split section, thin enough to follow normal curves in a given radius. One netter suggests soaking the strips in hot water first. This makes them flexible enough to do about 8" radius curves before breaking.

Curves, drafting

How do I make the curve reasonably regular?

Use a pencil on the end of a string and tack the free end of the string to the track base at the center of the curve (i.e. center of the circle formed by the curve). Then keep the string taught and draw the centerline of the curve on the track base with the pencil. Similar methods use a solid beam instead of string.

Curves, kinking

How do I keep flex track from kinking when I have a joint in the middle of a curve?

One simple way to reduce kinking is to stagger the joint (tends to happen naturally in curves anyway) so that the actual break in one rail is an inch or two (or more) away from the break in the other rail. Sliding the rails so that the break occurs over the ties of only one of the sections will also help.

			joint
   	================== =============================  <- rails
    	}  }  }  }  }  }  }  }  {  {  {  {  {  {  {  {   <- ties
   	============================== =================
				    joint

       } = tie from left section         
       { = tie from right section
Eliminating the kink
I solder the two sections together to eliminate the kink. I lay the first section, and glue it most of the way, leaving a few inches at end unglued. Then I solder both rails of the two sections together while the ends are straight. Then I finish gluing the rest of the first and as much as necessary of the second section.

Concerning soldering rail, still use a metal rail joiner, but then flow solder into the outside edge of the joiner/rail area. Don't solder the inside edge, or it will create problems with wheel flanges. The rail joiner helps hold the rails in alignment while soldering, and will keep them together should the solder joint fail (due to thermal expansion or contraction).

A second note about soldering on flex... You'll want to be careful not to melt the plastic "spikes" (I've melted entire ties). Melting them will naturally exacerbate kinking/gauge problems. How do you melt solder (374 deg F) without melting plastic?? Good question! If you are an expert with a soldering gun, no problem. I've found that using a 100 watt gun helps because it heats the rail fast, so you can get in and out quickly. Plasitc meltage seems directly proportional to time, rather than temperature.

Handling flex: I try to keep the sections straight until I really need them, and then I only bend the section for fitting purposes in the one place where it's going to go. This way, no pre-kinking gets into the rail and the rail will on its own seek a smooth curve between the endpoints. This also helps in making straight mainlines. Of course for old, somewhat overused yards or spurs, you might deliberately kink up the section first, and then try to lay it as straight as possible... this produces a good effect when sighted along at track level.


Roadbed, Homasote

20 Mar 94 01:00:00 GMT

What do I need to know about homasote?

[The following description of Homasote was written by Gregg Fuhriman]

My experiences
Homasote is a material made of densely compressed newsprint, and is usually sold in 4' x 8' x 1/2" sheets. It is also available to the model railroad hobby as precut roadbed strips. My experience is with the large sheets, which I purchased from a building supply store. In 1987, one sheet cost around 25 dollars.

I cut it into many 8' strips, using a rotary saw set at 45 degree angle to get the "ballast slope". The ascii drawing below shows an edge-on view of how I cut the Homasote sheets:

				     |   |<---- wide enough for track
	        ___________________________________________________________
		|\   /   \   /   \   /   \   /   \   /   \   /   \   /   \|
		-----------------------------------------------------------
				<---------- 4' ---------->
These strips were then used as-is for straight roadbed. To make curves, I cut dozens of kerfs crossways about 3/4 of the way through the strip and about 1/2" apart. The strip could then be "bent" into the desired curvature by compressing the kerfs on the inside of the curve. The purpose of this exercise was to reduce wasted Homasote; it is not a cheap material.

The homasote strips were then glued to plywood sub-roadbed using carpenter's glue and clamps. I also drove small nails through the Homasote into the plywood to help hold things in the right position while the glue dried. Special shapes, like around switch stands, were shaped by hand using a utility knife to carve the homasote.

Summary of My Experience and Netter's Comments
Benefits of Homasote are that it holds spikes and nails well, and it is a sound-deadening material to reduce train noise. It glues easily, as it is a porous material. It is relatively "soft", so it cuts easily.

Cutting this material with a power saw generated piles of fluffy, gray dust that went everywhere and proved to be a pain to clean up. I reccommend wearing a dust filter to avoid breathing in the dust and fluff. Cutting with a utility knife is neater, but more difficult and tedious as the homosote tends to "grab" the blade (just as it "grabs" spikes and nails).

Forming curves with the kerf-and-bend method was not 100% successful. At least half the time the strip would break in two. Also, the resulting curve is not super smooth ... they are actually several short straight sections. Heavy sanding can smooth out some of this (but with more dust).

Other netters have suggested painting the Homasote with a latex paint to help seal it against moisture (introduced mostly during ballasting, but also to guard against ambient humidity). Its dimensional stability with respect to temperature and humidity has been panned, though I have not noticed problems with my layout so far.

I'm starting a first layout in my dorm room (4X6) and I'd like to hear some of the pros and cons of plywood and homasote for the base of the layout.

Homasote is something you laminate on top of plywood. Don't use it "free hanging" from the benchwork. If you're using something like L girder benchwork with cookie-cutter trackbeds, one sheet of plywood goes quite a way (because you can splice together the cut-outs to make more track bed.

1/2" or even 3/4" plywood with an overlay of 1/2" Homasote is what I've seen used on industrial strength layouts. I'm using 1/2" Homasote on top of 3/8" plywood on my layout -- the downside of this is that it takes more supporting risers to get a decently rigid roadbed with thinner plywood.

...

I'd recommend an open framework as being better for landscaping, but if you're sure you want one big flat area to build on, I'd recommend both: homasote over plywood over a 1x4 frame. Homasote is much easier to drive track nails into, not to mention removing them if you want to re-arrange your trackwork.


Electrical power distribution

20 Mar 94 01:00:00 GMT

I have an HO layout power problem. At the furthest parts of the layout there is a noticable power drop. How can I fix this?

What you need is a power feed to this part of track that needs power. If you use blocks, each block needs more than one feed from the block switch. Most people just put the power feed right in the middle of the block. Instead, the feed should be split in two (or more) and one be put towards one end of the block and the other towards the other end of the block. More feeds, less chance of having power drops.

At my club, each joint has a wire jumper, and the joint itself is not soldered. This controls the expansion/contraction of the rail in the heat/cold environment.

...

Run a second pair of wires off of your transformer and attach them at the farthest point of track, or at the point where the engine seems to be slowest. This should allow the bad connections, or resistance of the track to be eliminated.

...

Don't depend on track to carry your power to the ends of the world (layout). The conductivity of track isn't that good, and NS (Nickel- Silver) is even worse than brass. I've been told that code 100 NS track has the same conductivity as a #22 copper wire. NS is recommended over brass however for it's lack of corosion. What you should be using is feeder wires from the track down to a point below the layout, and a fairly heavy gauge wire to carry the power to the feeders. Our club has a policy where we drop a feeder wire at every point where two tracks join. This insures that every section of track has at least one feeder wire connected to it, and is some insurance against "dead" sections. I like to solder the feeder wire to the bottom of the track connector. That way the wire is not seen on top. We use #10 stranded wire to carry power to the feeders.


Track, glueing

20 Mar 94 01:00:00 GMT

My track doesn't have holes in the ties for nails. How do I hold it in place?

A construction cement called "Liquid Nails" will hold the cork and track in place. Just a thin coating of this with the track pressed down onto it will set in about 30 secconds... even with the 8" radius curves which the flex track will not hold on its own.

An alternative is to use white glue spread with a finger over the surface of the cork. You will need temporary pins along the track to hold it in place while it dries.


Layout planning software

20 Mar 94 01:00:00 GMT

Abracadabra, Design Your Own Train

Could someone please give me the name of the Abracadabra product that is like "flight simulator for trains", i.e., the one that lets you simulate being the engineer on a layout you program into it. There are several products on the market with similar sounding names and when I call around to try and find it at one of the local software houses I want to make sure that I'm asking for the correct one.

"Design Your Own Train" and "Run Your Own Train" are a pair of programs that let you do what you say. RYOT gives you a "hogger's eye view" of the road; that is, the view is supposedly that which you would see from the cab.

"Design Your Own Railroad" allows you to create a layout complete with scenery, then run trains on the tracks. You only get a bird's eye view, though.

I personally have used DYORR, and recommend it. I have not used DYOT or RYOT, but have heard that they are less sophisticated products. Of course, if you are determined to have the flightsim-style view out the windshield, RYOT is your only choice.

All three products are published by Abracadata software, (800) 451-4871. It's cheaper to order the products from MicroWarehouse, however: (800) 367-7080.

[Editor's note: I've also seen advertisements for products which look more like CAD packages tailored to model railroading. A review would be most appreciated.]

REALROAD

REALROAD from Digital Power is a locomotive simulator that operates model railroad trains. With this program, you enter information about the type of train you want to simulate. REALROAD uses this information and monitors throttle, brake switch, direction switch and many other user settings to control the model train, making it perform just as a real train of the same type would. It works with existing 8- to 20-volt dc model railroad and track wiring. No modifications are necessary. An interactive calibration procedure allows the program to learn about each model locomotive.

REALROAD comes with enough information to simulate 150 locomotives. Freight and passenger-car information is also included. You can modify or add to the list of possible cars, as desired. Multiple locomotives can be combined on the same train with varying numbers of cars, proving *(?-providing?)* virtually endless train simulation possibilities. REALROAD's controller card installs in an empty PC/compatible slot, while power-supply and track connections are made via a terminal strip on the back of the card.


Track, ballast glueing

02 Aug 94 01:00:00 GMT

When handlaying track, how/when do you folks glue down the ballast?

Some people ballast their trackwork after they have layed the ties but have not yet layed the track and some lay the ties and ballast the track at the same time. Although I have never layed ties and ballast at the same time, I have ballasted before laying track. Basically what I do is the following:

  1. cut all of the ties to length
  2. sand all of the tie ends smooth
  3. stain all of the ties one of three colors
  4. glue the ties down
  5. spread all of the ballast, making sure that none goes above the tie tops
  6. fix the ballast using the standard wet-water/glue method
  7. lay the track
This process is nice in that there is no way to glue your points shut. However, you do have to be careful to make sure the ballast does not interfere with laying the rail, so I don't know if you coul model fresh ballast where the ties are buried. It is also difficult to model old branch lines where the ties have sunk into the ground. But for modeling logging lines or older mainlines, it works beautifully.

When staining ties, I like to have three colors of stain to give some variety. I usually stain about 75% of the ties one color, 20% of the ties a lighter color and 5% of the ties a darker color. The lighter color represents older ties while the darker color represents newer ties. Because I model logging lines, my 75% is a silvery gray color (20 parts alcohol:1 part grimy black), the 20% is a lighter gray color (20 parts alcohol:1 part gray) and the 5% is a fresher wood color (20 parts alcohol:1 part dark tan).

...

Ballast is added in the same fashion for both handlaid and prefabricated track. The roadbed is prepared and the track laid in position using your favorite method before any ballast is added. The method described below also works just as well for grass, dirt, etc. in the rest of the layout.

There are several methods, but the basic idea is to spread the ballast in place (I use a cheap 1" paintbrush to shape it), soak it with some sort of wetting agent, and then flow a fixative into it. Variants abound - this is a FAQ in the model railroading magazines as well, so look there for alternatives.

The most common wetting agent is tap water with a drop or two of standard dishwashing liquid added to cut the surface tension (the water will just bead up otherwise and won't soak the ballast). This can be sprayed on with a *fine* mister or carefully dripped on with an eye dropper. The most common fixative is a 50-50 mixture of white glue and water, again with a drop of detergent. This is dripped onto the ballast and allowed to dry. All of the water will evaporate, so the ballast should be as wet as possible without floating it away; otherwise you may just glue down a top crust which will chip away later.

...

  1. Get a small paintbrush with loose bristles about the width of the inside of the track.
  2. Get a really fine grind, almost dust-like.
  3. Mix up in a spray bottle half white glue and half water with a little more glue than water.
  4. Fill another spray bottle with water. [Usually it is a good idea to add a drop of dishwashing detergent or rubbing alcohol to the water to reduce the surface tension and allow the water to penetrate the ballast better. Ed.]
  5. Spray the grind with the water to compact it so when you apply glue & water mixture it doesn't blow all over.
  6. Spray with glue & water mix. Do small sections when doing track.
  7. When doing gravel roads or grass, do large sections.
  8. Take your time and be patient.
  9. Have Fun!!!! 8^)

Track, ballast, material

20 Mar 94 01:00:00 GMT

Who makes good ballast material, and do you mix/combine several coarsenesses or make it uniform?

Woodland Scenics is probably the most popular brand, but at least one poster described it as looking like kitty litter; a bit harsh but not far from the mark in my opinion. Their finest grade should be used by N and HO scales, and it's really too coarse for N.

The other major source is actual rock. You can just walk outside if you live in the area you model, you can try the local quarry or gravel operation to see if you can get a small sample, or you can order it from several operations which advertise in the model railroading magazines. If you use real rock you must crush it, sift it to size, and then remove any ferrous particles with a magnet.

Note that most real railroads use ballast that is available locally, so the color of your ballast will differ based on the area modelled. Most of it is a standard gray, but iron ore roads have a distinct reddish hue, and RMC just finished a series about a marble quarrying railroad which used marble chips!


Curves, prototypical

20 Mar 94 01:00:00 GMT

I have a question about track laying and rail curvature. I am under the understanding that rail curves on prototypes are designated by degrees of curvature. Does this mean that a curve in the track points the train in a new direction X degrees from the way it was going? If so doesn't this have to occur in a specific distance to indicate the sharpness of the curve?

Yes indeed. The measure is in "degrees per chain", where a chain is an engineers chain (a unit of distance equal to 100 feet and traditionally measured with a chain).

Just to keep things confused, land surveyors also measure distance in chains (and rods!), but a surveyors chain is not the same as an engineer's chain!

Think of trying to survey a railroad line. You can't just stick a compass point in the ground at the center of the desired curve and scribe the centerline of the track, the sizes are just too large, and in addition, your surveyors frequently can't leave the right-of-way for fear of death because the adjoining landowners are furious about the condemnation proceedings by which the right-of-way was just obtained.

Instead, what you do is work your way down the right-of-way pounding stakes into the ground every chain (100 feet!) to indicate where the centerline of the track belongs. The grading crew follows you and tries to make a level grade along the line you've surveyed (with the help of a leveling crew that uses different surveying instruments, specifically, levels), and then another survey crew follows, re-marking the centerline on the top of the grade for the tracklayers.

To do curves, you set your transit over one stake and sight back at the previous stake (let's assume it was on the tangent, for simplicity). Then you invert the transit to make it point along the same line, but in the forward direction, and rotate it by the number of degrees in the specified curve. Your partner is standing at the other end of a chain anchored to your stake, and you signal him with your hands to shift left or right until he is on the mark, whereupon he drives a stake. You then pick up your transit and your end of the chain, walk to the new stake, and repeat the process.

Of course, it gets more complicated on rough terrain.

The angle between the two stakes (spaced 100ft apart) with the center of curvature as the vertex is equal to the "degree" of the curve. The angle of the right triangle (stake,center,P) is one half of the "degree" of the curve. Thus with "d" = the "degree" of the curve, "r" = the radius in feet, and sin and arcsin calculated in degrees we have:

d = 2 arcsin(50ft/r)
or solving for r
50ft/sin(d/2) = r
	                        stake
	                        |\          ^
	                     /  | |         |
	                  /     | \         |
	               /    50ft|  |track   |
	            /           |P |        |
	    center ----------------|       100ft
	            \           |  |        |
	               \    50ft|  |        |
	                  \     | /         |
	                     \  | |         |
	                        |/          v
	                        stake
	
	           |<----- r ----->|
Note: The 100ft is measured along the cord of the curve, .i.e. a straight line, not along the curve, making "180 degrees" (r=50ft) the sharpest curve that can be expressed in "degrees".

One benefit of surveying curves this way is that, if the track needs to make a total curve of, say, 10 degrees, as measured between the tangents at each approach to the curve, all you do to determine the number of surveying stations between the start of the curve and the end of the curve is divide the total curve by the number of degrees allowed as the maximum curvature. Thus, for the example 10 degree curve, if the maximum allowed curve is a 2 degree curve, then there will be 5 survey stations that must be offset by 2 degrees each to make the total curvature.

Approach spirals into tight curves can be worked out similarly. To start a 9 degree curve (that's pretty sharp), you might require one survey station at 3 degrees, one at 6 degrees, and then 9 degrees per station from there on into your spiral, and then work your way back to a tangent with a 6 degree step and a 3 degree step.

...

Just to confuse matters a bit, Track and Turnout Engineering by G.M. Kurtz mentions that in mountainous terrain the Southern Pacific used a 50 foot chain for curves and in flatter terrain used the usual 100 foot chain.


Roadbed, bending cork

20 Mar 94 01:00:00 GMT

How do you bend cork roadbed?

I have found it useful to soak the cork in hot water first, and then bend it while wet. (As per Life Like's instructions on their package). I was fortunate that the glue I was using (Liquid Nails) was able to bond securely even to the wet cork. I was using N scale roadbed and was able to do about 8" radius curves without breaking the cork.


Track, bending

20 Mar 94 01:00:00 GMT

I've heard that you have to be careful laying curves so that you end up with a smooth curve instead of a series of chords going from one tie to the next. So how do people bend smooth curves? I know that there are expensive rail bending machines available (sounds pretty specialized, and definately beyond my means).

A rail bending machine is easy to build. As long as you're laying wide radius curves, you won't need one. Balance a meter or so of rail on your finger and let it droop under the influence of gravity, whatever scale you're in, even up to prototype sized rail, you'll get it drooping to a wide radius curve! Actually, your finger won't comfortably hold a meter of prototype rail, and you can safely spike rail in curves significantly tighter than the arc to which it sags, but let's get on with it!

Suppose you're building a curve that does need pre-bent rail. Most LGB stock curved track is in this class, most HO stock track isn't, but if you're building scale models of streetcar track or turntables, you'll need a rail bender in almost any scale you work in.

Here's how to build one:

Take three pieces of square wood stock (2 cm or 3/4 inch thick for HO, 4 cm or 1.5 inch thick for O, 6 cm or 2.5 inch thick for G), make the pieces 5 or more times longer than they are thick, and glue and nail them to a solid base like so:

	      _____  _____  _____
	     |     ||     ||     |      The middle piece should be free to slide
	     |     ||     ||     |      but very snugly held between the outside
	     |     ||_____||     |      pieces.  The two outside pieces should
	   __|/////|_______|/////|__    be solidly glued and nailed to the base.
	  |                         |
	  |                         |
	  |_________________________|
Next, build three "pullies" out of stacks of washers rotating on wood screws. Each pully should look like this:
	          _
	    __   | |  _
	   /  |__| |_| |__________ 
	  (      | | | |          \  __ This screw should not quite come through
	  (    __| |_| |__________/     your wood square stock when screwed in.
	   \__|  | | |_|
	         |_|    \ This washer (or stack of washers) should be a tiny bit
	            \     thicker than the foot of the rail
	             \
	               This washer (or stack of washers) should be a bit thinner
	               than the gap between rail head and rail foot, so it rides
	               against the side of the web of the rail.  This washer
	               should stick out above the smaller washer enough to keep
	               the foot of the rail from running into the smaller
		       washers as this washer rolls (or slides) along the web
		       of the rail.
Now, screw one screw (with its stack of washers) into each piece of square stock, about 1/3 of the way in from one end; don't overtighten them, since you want to leave the washers able to roll along the edge of the rail.

From above, it should look like this when done:

	    ___________________________
	   |                           |
	   |       (O)                 |
	   |___________________________|_
	   | |                           |
	   | |       (O)                 | < the sliding member can be slid
	   |_|___________________________|   in and out with a bit of effort
	   |                           |
	   |       (O)                 |
	   |___________________________|
Now you can start bending rail. Clamp the bender in a vise with the slider offset about like is shown and try to force a length of rail so it goes under one guide washer, over the middle one, and under the final one, with the foot of the rail against the boards and the rail head away from the boards. Adjust the offset of the slider until you can just manage to shove rail through, and the three guide washers will put a uniform bend on the rail.

You'll find that you can only bend the rail a bit on each pass through the bending fixture, and the limit on how much you can bend it is determined by how hard you can shove the rail in. Don't pull the rail, you'll lose control of the radius that way. Shove it through, and after each pass, slide the slider a bit farther to tighten the radius on the next pass.

I used this approach to bend perfect circles of rail for a turntable ring rail, and I figure it's good for many more applications.

Note that the washers should be nice and round, with no flats or rough spots. If they have flats or rough spots, they're likely to slide over the web of the rail instead of rolling, and that'll make it much harder to push the rail through. For larger sizes of rail, you might want to grease the washers.


Track, code

02 Aug 94 01:00:00 GMT

What does the "code" associated with track mean?

This is the height of the rail, expressed in thousandths of an inch. Code 100 rail (common in HO) is 0.100 inches high. This is equivalent to prototype rail weighing 152 pounds per yard, which is larger than almost all rail used by real railroads in the United States. Advanced modelers typically use rail sizes closer to that used on the prototype, as shown below:

	----------------------------------------------------
	Nominal |   Mass   |  Scale   | Scale   |  Scale   |
	Scale   |  Market  | Mainline | Branch  |  Mining  |
	----------------------------------------------------
	G, #1   |  .330    |  .250    |  .175   |   .125   |
	----------------------------------------------------
	O       |  .175    |  .148    |  .100   | .070-083 |
	----------------------------------------------------
	S       |  .148    |  .125    |  .083   | .055-070 |
	----------------------------------------------------
	HO/OO   |  .100    | .070-083 |.055-070 |   .040   |
	----------------------------------------------------
	N       |  .080    |  .055    |  .040   |   n/a    |
	---------------------------------------------------
	Z       |  .062    |   n/a    |   n/a   |   n/a    |
	----------------------------------------------------

Custom turnouts

20 Mar 94 01:00:00 GMT

What advice can you give me on making custom turnouts?

I've made a number of turnouts including a double crossover that leads into a double-slip switch for a yard throat. (I did not make the double slip, just the crossover, but it was still an alignment challenge.) Iv'e used several methods and find the best one involves soldering the rail to PC board ties. This is the only way I have found to get the rail in correct alignment and keep it there. I've tried just spiking the rail to various materials as well as various contact cements. The latter are applied to the rail, allowed to dry, and then heated with a soldering iron to soften the glue as the rail is pressed into place. This last method works for running rail in code 55 and 40, and you can put down quite a bit of rail in a hurry. But I find it is too much trouble to get and keep alignment in turnouts.

The best article I have seen on turnout building is by Ed Stimpson Jr. in the May 1978 Model Railroader, titled: Maintenance-free turnouts, beginning on p. 62. You should be able to get this from any decent-sized library. If not, ask for the interlibrary loan librarian. I have tried these techniques and they work. To summarize the article, you begin with a paper pattern. Glue PC ties at strategic points along the length of the turnout. Obviously, important spots are at the frog. Many turnout problems are in the area of the frog and guard rails. The most important tool is a NMRA gauge. Be sure the various gauges are correct. If you are not a perfectionist, don't bother to make your own turnouts. Quick-and-dirty or good-enough approaches eventually lead to trouble. Further, it is not a way to save money if your time is valuable. There might be some savings if you are making a hundred and can set up jigs and do mass production.

BTW, I do not like making turnouts in place on the benchwork as this is hard on my back, but many track layers prefer this method. This way, each turnout is custom-fitted to its location. The techniques in the article work for either method.

I deviate from the article when making the points themselves. I bend a piece of rail to the opposite or reverse of the angle needed (sorry I don't know the term... I flunked geometry). Then I file the bent part to a straight line until it is almost all the way through. Then I fold the rail backwards toward the filed-down part to form the point. Sorry it won't illustrate in ascii. Imagine bending your arm slightly, then filing your elbow down to almost nothing, then forcing your arm backwards so the outside of your forearm touches your triceps (ouch!).

The straight rail is laid first, then the points are soldered in place, followed by the closure rails and the outside curved rail. Guard rails are added last. Gauge on the guard rails is important as they keep the flange on the other wheel from hitting or picking the points. Finally, the copper on the PC ties is slit to prevent shorts across the rails. There is more to it than this, and I suggest you find a copy of the article. Making yor own is often the only way to solve a special problem, and it's not all that hard.


Track, detection units

20 Mar 94 01:00:00 GMT

How should I go about setting up a track detection system?

Get a copy of Bruce Chubb's book "Building Your Own Universal Computer Interface" available from Tab or Walther's. This is a further development from his series several years ago in MR about the Model Railroad Computer Interface.

He describes the system in there. If all you need is a track detector, this book also includes a description of his circuit that is compatable with command control systems.


Track, electrical gaps

20 Mar 94 01:00:00 GMT

What are the recommended methods of installing/cutting/maintaining electrical gaps?

Here on the NEB&W, we use handlaid trackage, but we still have a large number of gaps, especially since we use solid metal frogs. And what do we do about the gaps? The answer is...nothing. That's right, nothing. Our gaps are just that, gaps. In fact, a little over a week ago I cut a curved number 10 turnout into the mainline, and there is a 1/8" gap right off the frog on the mainline route. And it ran perfectly during all twelve hours of our weekly open houses. In fact, the only problem we had with it was due to the angle of the frog. Being such a shallow angle, I had to install the guide rails to keep the wheels from taking the wrong route in the frog. After I did that, there were no further problems. And, all of our curved mainline turnouts are superelevated, which makes it all the more difficult.

On a layout like ours, where we have no humidity control, the gaps should be left open to allow for the expansion of the rail and the benchwork. As long as the rail is spiked securely in place, open gaps will prove no problem.

It also means that areas like turnouts, crossings, etc. and other special trackwork won't be knocked out of gauge by expansion problems.

...

I have a small (3' X 4') N scale layout with mostly hand-laid track. Generally I cut gaps using a moto-tool cutoff wheel which makes nice straight, narrow gaps. I usually use a small shim of styrene that is epoxied in place to fill the gap. The styrene is easy to file down to the contour of the rail and can be painted to become invisible.

As the layout is small, gaps are only about 2-4 feet apart and the amount of expansion and/or contraction would be minimal. For longer sections, I would probably leave the gaps open as the plastic filler might not be flexible enough to take the compression and rails between the gaps would buckle slightly out of allignment.

...

Here's a technique that works for me:

  1. spike track securely in vicinity of gap
  2. cut gap (if it wasn't built-in in the first place)
  3. fill gap with epoxy cement
  4. when set, paint over epoxy with rail color [Some railroads fill their own rail gaps with epoxy. Painting is optional]
...

If the gap is on a curve, (to be avoided if possible) I puddle some epoxy on around where the joint will be, let it cure.

I allow for expansion by allowing the "normal" rail joints on straight track to be a little loose and only solder joints on curves, to keep the rail ends in line. I then run a pair eighteen gauge wires under the track to serve as a bus and run feeders up from this to connect to EVERY piece of rail on the layout. Nickle-silver dosn't oxidize very quickly, but it is not a very good conductor.

...

I cut my gap, usually using a Dremel cutoff wheel. Then I take a flat or square, not round, toothpick and push it into the gap. I then "set it" with some white Elmer's or furniture glue. Next day, I trim up the toothpick to conform to the rail contour and have a permanent, fixed gap. Have some that are still good, and, for all purposes unseen, for over 20 years.


Track, flex

20 Mar 94 01:00:00 GMT

What should I know when laying flex track?

I used a construction cement "Liquid Nails" to hold my cork and track. Just a thin coating of this with the track pressed down onto it will set in about 30 secconds... even under said 8" radius curves (and my flex does not hold curves on its own).

And about Liquid Nails, the stuff skins over fairly quickly (couple of seconds) but will remain usable for a couple of minutes. Just make sure you press the track down into it firmly and you should have no problem.

...

To join track on a curve, solder it while straight, and then bend it. You're best off avoiding joints on curves as much as possible.


Track, nickel silver vs brass

20 Mar 94 01:00:00 GMT

Nickel-silver vs. brass: what are the issues?

Nickel-silver is a copper-nickel alloy (mostly copper), considered to be metalurgically similar to brass but superior in corrosion resistance. Brass oxidizes rapidly to a non-conducting surface, which means that power will not reach the locomotives or cars, resulting in stalls. Both forms of track will accumulate other gunk on them, requiring some form of cleaning. In general, nickel silver is much better than brass, and is worth the small extra cost.

The more recently developed alloys used in high-quality G scale track are much less corrosion-prone. Brass is sometimes favored over nickel-silver for outdoor use, because it expands and contracts somewhat less with changes in temperature.


Track, soldering

20 Mar 94 01:00:00 GMT

Anybody have any advice on soldering HO rail?

If you join the two piece of rail with metal rail joiners, then solder the join, the solder will flow much more smoothly. This assumes you're using nickel-silver or brass rail of course! Make sure that the soldering iron heats all three metal parts (ie. both rail ends and the rail joiner) to the same temp before applying the solder, and it should be drawn into the space inside the rail joiner without any trouble.

Note that solder is not a very good gap filler, so if you've got gaps larger than about 0.5mm or so, you'll waste your time trying to fill them with solder without having a bulging great lump of solder in the way of the wheel flanges.

Also note that you should still have non-soldered gaps every metre of so to prevent the solder joins fracturing due to expansion and contraction. Ideally, straight track should have NO directly soldered rail joints, but you should use bridging wires to leave the rail ends free to move. Curved track will usually need to have soldered joins unless you're particularly skilled at spiking rail on curves.

...

I am one of the proponents that do not believe in soldered joiners. Joiners should be left free to move and expand as nature dictates.

Joiners are a mechanical way to join two pieces of rail, and should be just that. They should not be electric current carriers. No matter how good the contact is, it will eventually fail due to oxidation.I believe that the "bus" method (ie. a wire line that follows each section or block around the layout) is the only reliable method to power a layout, but that is my own opinion. There are a few different approaches to dispense power to all sections on a layout.

A lot of people will say that it is a waste of time to solder the wire under the layout to the track, but think of the fewer headaches you will have with electric problems.

I would like to give two examples:

Back then they used a method used in O gauge layouts. A piece of wire was soldered near the end of each rail, connecting electrically the two rails. Talking to the old members, they hardly had any problems as far as power flow is concerned (shorts yes). The only problems of power flow were due to failure of the solder joints. Even now, after 40+ years we find the need to re-join some of the wire with new solder (after a good cleanup), although most of them are still the original solder joints.

Performance is very reliable, and that is what counts in the end.


Turnout problems

20 Mar 94 01:00:00 GMT

What problems can I have with switches and how do I avoid them?

I have found problems with Atlas switches. The flangeways don't seem to be deep enough through the frog, and the large plastic frogs can cause engines to stall. Also, the switch machine housing sticks up too far, and can entangle steam locomotives with wide draft gear.

Use Peco or Shinohara turnouts for N scale; they run and look much better. Also, if you are nailing down the track, allow a little room under the nail head for the switch to "float"; putting the track nails all the way in can distort the rails.

...

My first fault was that I tacked down the turnout. If the turnout is allowed to "float" it will operate better.

Second, use a jewlers file, and file the points of the turnout, this was described in MR a few years ago.

The third thing I did was to get rid of the N scale electric switches. I was lucky that my local hobby shop took them back. I traded them in for the HO below ground throws, and some piano wire. I cliped the plastic pin off at the base but befort the cone on the HO machine. I then cut the piano wire about 2" long, then using an iron, heated the piano and forced it into the cone stub. I then installed the machine, with the piano wire fitting in to the turnouts throwbar (between the rails). Once the machine was secured to the underside of the layout, I then trimmed the piano wire (about .05) above the throwbar.

This made the turnouts not only work better but look better. For a finishing touch I cemented pieces of ties (left over from flex track) next to the turnout for switch stands.


Turnout numbers

Date: 20 Mar 94 01:00:00 GMT

What does the number associated with a turnout mean? What's the difference between, say, a #4 turnout and a #6 turnout?

The number is based on the angle between the straight and diverging tracks. Skipping the exact definition, a #4 turnout is generally the sharpest practical size and is equivalent to an 18" curve. #6s and #8s are more gradual and are typically used on larger layouts both because they look more realistic and because longer equipment will work better on them.

Traction models and models in the smallest scales (N and Z) can often get good results with much smaller turnout numbers, down to #2.5 in common usage.


Dust control

18 Mar 94 01:00:00 GMT

Can anyone recommend a method of controlling dust on a layout?

Short of covering up the layout after each use, a better method is to have the layout set up so that it's viewed almost from eye level AND has a 'roof' over it about 0.5 to 1 metre (2 to 3 feet) completely covering the layout.

If you've ever seen articles on shelf-style layouts built at eye level, you'll notice that a lot have the scenery 'wrapping over' the top of the layout.

I've got no permanent layout, but the small yard I have built on the workbench in the garage has a wall-mounted cupboard over one end, and it's amazing how dust-free this area stays. The section that's not covered over by this cupboard collects a lot of dust, dead insects, etc.

...

A couple of years ago Jack Burgess wrote in RMC how he licked the dust problem on his layout (located in a CA garage.) As I recall, he did some weathersealing of the garage door (not 100%), carpeted the floor (concrete creates dust as it is walked upon), and lastly, he installed electrostatic air purifiers that run constantly.


Layout design

22 Jul 94 01:00:00 GMT

This, and the two following sections on benchwork comes from Frank Kastenholz, who as an introduction writes this:

I'm going through the process of building a railroad right now. I thought that as I went throught the process I'd journal any hints and the like that I've discovered during construction.

As of 17 April 1994 I've got a design complete, and I have the L-Girders, legs, major joists (i.e. enough to hold everything together under the 'stresses' of construction), and plywood 'surfacing' is in place. I've chosen to build using the 'cookie cutter' method on a plywood base. I'm using plywood because I couldn't find any Homasote. I'm doing cookie cutter for two reasons: 1) I do not have enough space to set up a second work-area to cut roadbed strips out of the plywood -- with cookie cutter I can put the plywood on top of the joists and cut it in situ and 2) I'm doing an urban setting so most everyplace will have horizontal surfaces anyway.

As of 17 May, the benchwork is pretty much done. I still have some trimming and minor adjusting to do, as well as adding any supports or joists that I find that I need.

As of 13 June, the main line (cork roadbed and track) is in place and I am starting to thoroughly test the track.

As of 6 July, I've thoroughly tested the mainline trackwork and have started working on the wiring.

Anyway, here are some hints and tips that I've developed during design and construction:

Design

  1. Read and thoroughly comprehend "Track Planning for Realistic Operation" by John Armstrong (Model Railroad Handbook Series #6, Second Edition, Kalmbach Books 1979, ISBN 0-89024-504-5).

  2. Read and thoroughly comprehend "Track Planning for Realistic Operation" by John Armstrong (Model Railroad Handbook Series #6, Second Edition, Kalmbach Books 1979, ISBN 0-89024-504-5). (In case you missed it the first time :-)

  3. As you do the design, make cheat-sheets. These save huge amounts of shuffling through books and recalculating numbers and the like. What I've done is photocopy Figures 5-1, 5-8, 5-9, and 8-9 out of Armstrong's book (Standards {such as minimum radii} for different types of line, Crossover dimensions, Grade Chart, and Turnout Dimensions, respectively). I've also made out a sheet of paper with various useful calculations pre-computed on it -- stuff like rises for a given length of track at a given percent grade, linear distance around curves for things like 30, 45, 60, 90, and 180 degrees of curve at various useful radii, and so on. I'd suggest that you put these cheat sheets on colored paper so that they will be easily found among the piles of paper that you generate as you design.

  4. Take your time with the design. At this point, things are very easy to change. You should also use this time to get as much input as you can, reading books, looking at other layouts, looking at the real things, and so on. I spent about a year doing the design work, probably with about 10 fairly detailed track plans and maybe another 20 rough sketches, and did plans for 2 different rooms. No doubt my friends and family thought that this would be entirely a paper endeavor...

  5. Make a very carefully drawn diagram of the space you have and then photocopy that diagram. You are going to go through several iterations and it is easier to put a failed plan aside and start on a fresh piece of paper than it is to erase the failed plan, or make another drawing of the space. This also gives you a history, letting you refer back to see what ideas didn't work out and what ideas did in previous iterations.

  6. Leave space in the design. Inaccuracies, errors, and so on will creep into things at every stage of the effort. If you want to try something that needs the exact amount of space that's available, you will probably end up being disappointed. Rooms are not square, walls are not straight, all the errors add up. Under-design for your space.

    I've designed my layout to fit into a space that is 6" (about 15cm) smaller in each dimension. I can always expand the layout to fit into the extra space.

    I've also designed into the 'critical' track work places where some track can be removed, making things smaller, without adversely affecting my basic design rules.

    With luck you will end up with more space than you planned for. I am sure that you will have no problem filling that space. If you are not lucky, then the plan that you have made should still fit.

  7. When making plans for the benchwork, remember that l-girders and their webs, legs, and joists all have width -- don't draw them as single straight lines.

  8. Your paper plan should be considered only a final first draft of the final plan. Draw a full size version of the plan, preferably on big pieces of paper on the floor of the actual space to be occupied by the layout. Unfortunately due to space constraints I did not have this option -- I've had to build the benchwork 'on faith' that things will more or less fit. I'll put the plywood surface in place and then draw the plan on that.

  9. When doing the design, select grades that are easy to measure directly. If you have a sophisticated, culturally advanced measuring system available, such as the Metric system, this is easy, 10mm in 1 meter is a 1% grade... For those of us living in more backwards portions of the world, who have to make do with the English system, a 2% grade is just under 1/4" per foot, a 3% grade is just 3/8" per foot. By choosing grades that are easily measured like this, you will make setting the elevations in the benchwork much easier.

  10. When doing the design, remember that you will have to be able to conveniently reach and work on every spot of the layout. Leave ample room for aisles and access holes and place them so that all parts of the layout are within easy reach.

    Simply being able to reach some spot to put a car back on the rails is not sufficient. In fact, this is a relatively minor concern since the trackwork will (hopefully) be good enough so that there will be no derailed cars. Of more importance is to be able to reach spots to install and clean the track, work on the scenery and so on.

    Another facet of this is to arrange the layout so that the taller elements are farther from the access spot than the lower elements. In one place on my layout I have some elevated track that is right next to the aisle, with 'ground level' track behind it. As soon as I put in a temporary structure to hold the elevated tracks, I could see that the arrangement was 'suboptimal' and I will probably rearrange it when I build the permanent structure.

  11. When placing turnouts on the plan, don't forget that the switch machines require room. They take up 'footprint' on the layout (e.g. an 'above-ground' machine such as from Atlas will require about 1" by 3" (25mmx75mm) parallel to the turnout). "Underground" machines will also require 'depth' (e.g. a Tortoise will require about 6" (150mm) of space for the machine and access). This is most important when you have multiple levels of track.

Benchwork - Tools, Girders and Legs

22 Jul 94 01:00:00 GMT
  1. The wood that is available today seems to be softer, and presumably weaker than that which was available in the past. I've assumed that Linn Westcott used the strength figures for 'old time' wood in his benchwork book. As a result, I'd suggest that you use one size 'quantum' larger wood than his book would otherwise recommend. Not only will this cope with possibly weaker wood, but it will give you a margin for error.

  2. Use screws to put the whole thing together. You'll make mistakes and simply unscrewing the mistakes is a lot easier and cheaper than throwing them away.

  3. Use 2 drills when you are doing the work. I have one drill with the screwdriver bit in it and the other with a drill/countersink bit in it. There are a lot of screws in L-Girder construction. If you constantly have to change bits, you'll never get anywhere.

  4. Make sure that all the benchwork is accessible. I have one section of benchwork that runs parallel to a wall. I did not want to attach it to the wall so I built it as a standard L-Girder table -- with the L-girders attached to the outsides of the legs and the flanges of the girders facing 'out'. Then I moved the section into position. About an hour later I had to make some adjustments to the section, but of course, the screws attaching the 'back' girder to the legs were facing the wall -- and there was not enough room in behind them to make the changes. I had to take the section out and rebuild it so that the back girder was facing the room, making it accessible.

  5. When doing the benchwork construction, make one schematic diagram with all of the relevant plans on it. This should also have all the dimensions you need. Then take all other pieces of paper and put them someplace else. This avoids confusion as to which one has the right dimensions on it. I had two copies of the benchwork plan where I was working. One was an early plan and the height marked on the plan was about 4" (10cm) too high. I built about half the benchwork too high. Once I discovered the error I had to rebuild things to be lower. Fortunately it is easier to make a long piece of wood shorter than to make a short piece of wood longer...

  6. The person who invented the big wood clamp that's sort of like an alligator clip should receive the Nobel Prize for Model Railroading. These can be put into place and 'clamped' with one hand, making it real easy to hold things together with one hand and clamp them with the other. More traditional screw-type clamps require two hands -- which leaves you one short for holding things in place while clamping.

    Get several of these. 2 is probably an absolute minimum. 4 would probably be optimum. They should be big enough to clamp l-girders to legs, but not so big that they can not be opened and manipulated with one hand.

  7. At least one pair of bar-clamps is useful for clamping joists to the l-girder and holding it all in place while you drill pilot holes and screw them together. The more you have, the better since you can then clamp the risers to the joists, set the roadbed's elevations and grades and THEN screw everything together.

  8. When assembling the benchwork, I've found that you should build one section very carefully, making sure that it is level, at the right height and position, and so on. Then build each successive section off of what has already been built. This way you are always working from a known good point in the benchwork. What I did was as each section was finished, I got the L-Girders for the next section and put temporary legs on them at about the right height (connected with clamps so that they were relatively easy to adjust), then I'd butt the 'new' l-girders to the existing ones, get them level and then connect them up. Then I'd replace the temporary legs with the permanent ones.

    I never measured the girder height after I built the first section -- I'd just get the girders level and then connect them to wherever they met the legs.

    My layout is roughtly 'C' shaped. The baseline section that I built was the 'middle' of the C -- then I built the arms out from that section. When I was done, I checked the level from the end of one arm to the end of the other arm and the two arms were within 1/4 inch (~6mm) of each other.


Benchwork - Joists, Risers and Roadbed

31 May 94 01:00:00 GMT

  1. If you are setting any sort of grades, a gauge similar to the one in the February 1994 Model Railroader is extremely useful. The only addition is to have a longer base available, so that you can span between risers.

  2. I've found that there are two sets of risers that one uses. One set are the 'first' risers that are used to set the elevations and grades of the roadbed. These risers are the ones needed to get the roadbed at the right height. They are also the hardest to put in since they must be installed before the roadbed is in place.

    The second set are added after the roadbed is in. They deal with adding additional support where it might be needed, for example, if the roadbed sags someplace. They are easier to add since the roadbed can be used to set the height of the riser.

  3. When installing the risers, the first ones should be spaced no farther apart than you can level. For example, if you have a 24" (600mm) level, then these risers should be no more than 24" (600mm) apart. This allows you to set the riser-tops by levelling directly from the one to the next. You could use long pieces of wood for this, but the wood may have a slight warp and then you would end up screwing up the grades. Over a long distance, these errors can add up.

    The other way to set riser-tops is to measure up from the top of the L-girders, but if the L-girders are not level, then you repeat the error.

  4. If there are spots in the layout that must be at certain, unchangeable, heights (for example, to go under/over fixed obstacles) then start setting the elevations at those spots. If you start someplace else, the small errors that occur may add up and you might find that you do not reach the target elevation where you should.

  5. When splicing sections of plywood roadbed together I've been using scraps of L-Girder, attached under the roadbed with the flange facing down and perpendicular to the roadbed joint. The L-Girder is strong and, because of the flange, will not be likely to twist or bend, so the roadbed joint will be kept in alignment.

  6. When attaching the roadbed to the riser cleats, I've found that temporarily screwing the roadbed to the cleat from the top will hold everything together. Then you can drill the pilot holes and screw them together from underneath. Then remove the temporary screw.

    You have to firmly hold the roadbed to the cleat while drilling the pilot holes and screwing them together since the drill bit, and then the screw, would tend to push the roadbed up, away, from the cleat. A clamp would work, but sometimes getting the clamp in the right spot is difficult, and anyway, it would probably be in the way.

Cork Roadbed

Laying the cork roadbed was pretty simply and straight forward.
  1. After I nailed the cork to the plywood roadbed, I went over it with a 'shaper' such as a Stanley Surform. This eliminated any significant irregularities in the cork surface. These irregularities could affect the track -- causing sudden rises or dips in elevation, or perhaps raising or lowering one rail with respect to the other. These These irregularities occur at the joints in the cork -- usually the butt joint where one strip ends and the next begins, or in switches and other 'fancy' trackwork where there is a lot of cutting and fitting. Generally there is little variation side to side since the two parallel pieces are the same 'strip'.

  2. The actual layout should be designed on the sub-roadbed and the cork laid out on that design. This should all be done as accurately as possible. The best guide for laying the track has turned out to be the center seam in the cork.

Track

The track that I am working on is my main line; it is all double-tracked, code 100, Atlas nickel silver. I thought about doing code 83, or the like, but I've never worked in code 83 before and since this layout is my 'return' after about 10 years, I figured I'd better stick to the more forgiving code 100 until my skills got sharpened.
  1. After long thought, I decided that I'd sacrifice 'prototypical realism' for 'modelling reliability'. Code 100 is bigger than real life, true. On the other hand, it would be more forgiving of my rusty track laying skills, of cars and engines that might have big flanges, or out of gauge wheels and the like.

    If the track 'looks good' but suffers constant derailments then the layout will just sit and gather dust.

    I'm not saying "don't use code 83." I am saying that when you select your track, remember that the real role of the track is to physically support the trains and to provide an electrical distribution medium and the track size you select should be one that you feel comfortable working with to provide this physical support and electrical distribution. Poor trackwork means poor operation which means an unhappy model railroader.

  2. Lay the track slowly and carefully. Do not rush. If you get tired or bored or distracted, go do something else. Good trackwork is the key to having the trains run without derailing. I've found that I can lay 2 or 3 sections of track before my mind starts to wander -- then the curves aren't quite as smooth as I'd like them to be, the straights are not quite as straight, and so on.

  3. I've been soldering the rail joints. A high-power soldering gun is preferred here since you want to get the joint up to temperature as fast as possible. The longer it takes to heat up the joint, the more time there is for the plastic ties to get gooey and melt and throw the track out of gauge.

  4. I am using Atlas Flextrack. In order to make sure that the track curves are right, I bought some curve gauges. However, I've found them to be non-obvious in their use. Basically, if you stick the gauge between the rails, then the nail-holes are covered up so you can't nail the track into place. The best method I've discovered is to use the gauge 'half way'. You place the gauge so that part of the gauge is in track already of the right curvature (the 'back end' of the gauge) while the 'leading end' gets the 'new track' properly lined up. The 'leading edge' of the gauge should come right up to the next nail-hole, but not obstruct it.

    I could machine out a slot along the center of the gauge, but only thought of this remedy after I started writing these notes (and well after I'd finished with the track....)

    I've found, however, that if you are careful in laying out the cork roadbed, the center seam of the cork is an excellent guide. I laid most of my track following the seam and afterwards, after checking with the curve gauges, found that things were 'just right'.

  5. When soldering the track, keep the tip of the soldering gun very clean and hold the soldering gun to the inside of the track and apply the solder to the outside. Enough solder will wick in and around the rail joiner and into the joint between the rails to hold everything together. This technique reduces the amount of solder that gets on the inside of the rail at a position where it could interfere with wheels running by. Less solder on the inside, means less cleanup.

  6. Get a metal wheeled truck and run it constantly over the joints. Any clicking sounds indicate flaws. A Kadee truck works fairly well for this.

  7. Use your files alot. I've found it convenient to do the filing in two steps. First, before joining any track, I file all the ends. Then, after a joint is made, I file the joint so that the railheads are smooth and there are no obstructions.

    Before joining, I file the following spots of the rail:

    It is much easier to work with the track while you can move it around and hold it in the 'optimum' position for filing. Once the track is on the roadbed, it stays where it is and the file (and the person using the file) have to move to get into the right position.

  8. Keep the work area clean. You don't want goop in the rail joints or under the ties -- it can throw things out of alignment.

  9. In working with flextrack, it's easier to solder things together in a fixture to maintain alignment. I did 2 and 3 sections together in a fixture and then installed them on the roadbed.

    Atlas flextrack has one rail that 'slides' along the tie strips, the other doesn't. Things seem to work better (laying out curves and the like) if you join the tracks so that the sliding rail of one piece is joined to the fixed rail of the other.

  10. Try to arrange things so that the 'field soldered' joints fall on straight sections. Also these joints should be made in spots where there is easy access to the track. It is easier to keep the alignment.

  11. I tried to first lay all the track, then check everything for alignment, etc etc etc and then do the field soldering. didn't work. Things would shift slightly, joints would come apart a bit, or get out of alignment. This was not good. Put a piece down, get it aligned, and solder it to its neighboer and then go on to the next one.

  12. Check everything with a rail gauge. I like the NMRA guage since it is made out of thin sheet metal it can not only check the guage but the thin metal will easily pick up any irregularities, burrs, or nicks in the rail. When found, get out the file...

    I have checked the entire track, not just the joints. One common problem that I've found is that if the nail in the track is too tight, it will create a bit of a depression in the cross-tie, which then tends to pull the rails together slightly. The rails are then out of gauge and derailments occur.

  13. Have plenty of light available while working. Once you have what you think is enough, go get some more. You'll do a much better job. Plenty of light will reduce eye strain, making it easier to work. Good light also makes it easier to see flaws in the rail joints. Burrs, nicks, and dents in the track (all derailments waiting to happen) are also more easily seen. Especially useful is a small light that you can use to shine on a section of track from different angles. This way, you could get a good reflection off of a flaw.

  14. After laying the track run trains. Run lots of trains. Spend as much time as possible running trains, using as varied an assortment of cars and engines as you can. The idea is to find and fix bad sections of track now, before the ballast or scenery is in place. (It's also fun :-)

    One particular test I am thinking about running is to take a car and make it 'top-heavy' so any excess swaying will topple the car over, or at least sway a lot. This should (in theory) allow me to find spots where there might be dips in one rail and other 'cross-railhead-level problems'.

    The next edition will have the results of this test.

  15. Try to arrange your track laying so that place where you have to cut a piece of track to fit will occur in easily reached locations. This makes it easy to cut the track, check it for fit, adjust it, and so on. The two most common positions where this will occur are where the track 'meets up' with the next turnout (since turnouts generally have to be in fixed positions, and when we get to the next one, we are left with 'a little bit' of track to lay) and where loops get closed.
I've written this about a week after I completed the main line of the layout that I am currently building. I've been running test trains on the track for much of that time. For the most part, the track is excellent. There are some places where there are problems -- and those places seem to be where I had to learn one of the above hints 'the hard way'.

Switch Machines

I've been using Tortoise switch machines throughout the layout. While these hints are specifically a result of experiences with them, they should apply to just about any switch machines.
  1. Test the machines before installing them. I haven't had any problems with the switch machines, but I figure that if there are any problems with them, it is better, and easier, to find the problems standing at the workbench than once the machine has been installed.

  2. I've attached the switch machines to the underside of the roadbed with screws. One thing to be very careful of is to get the screw holes going straight into the roadbed. In a couple of places access is a bit tight and the pilot holes went in at an angle.

  3. I've found it convenient to run power to the switch machine locations as the machines were installed. This way I could test each machine's alignment as it is installed.

  4. Tortoise machines, including the mounting flanges, are a bit too wide for mounting side-by-side for parallel HO tracks that are spaced the minimum distance apart (2"/51mm).

  5. In general, switch machines will take up space -- either underneath the roadbed or next to the track. The roadbed should be wide enough for mounting the machines, and for under-roadbed machines, there should be enough vertical clearance for the machine, and for your hands and tools in order to install and align the machine.

Control Panel and Wiring

  1. Masonite by itself it too flimsy for building the control panel. It should be either a stiffer material, or some stiffeners should be added.

  2. I've screwed the panel directly to the framework. This is not good. Attach the bottom of the panel to the framework with hinges so that you can open it up.

  3. If you can, leave plenty of room between the individual block and turnout control switches. It makes soldering the connections much easier.

  4. I've daisy-chained the power from one switch to the next. This has led to a rat's nest of short wires. Yuck. Some kind of a bus distribution scheme would be better.

  5. The wiring has run directly from the switches on the control panel to the terminal blocks. If I ever have to remove the panel, disconnecting all the wires will be painful (and I'll almost certainly make mistakes when connecting them back up). I'd suggest using some kind of connectors, such as Molex or Amp between the control panel and the terminal blocks.