Sunday, February 26, 2012

A code 40 turnout

Today's project was going to be figuring out the best way to solder point jumpers on to my Atlas code 55 N scale turnouts.  I've tried one or two techniques already, but I'm not totally happy with them.  I was also looking into cleaning up the points.  After a few false starts, I made an abrupt course change and pulled out the Fast tracks code 40 turnout jig I bought a few years ago but never actually tried.

A long time ago, when the first article on handlaying code 40 track came out in the late '70s, I built a timesaver switching layout in code 40.  My first turnout worked marginally, the subsequent ones weren't even that good.  More recently I've built a large curved turnout in code 148 rail in O scale, with much better results.  I was more careful, and I know a lot more about turnouts than I did in the '70s.  I was pretty sure I could build a turnout using the Fast tracks jigs without too much trouble.

Actually building the turnout took longer than I thought it would, but in fact it was in line with what the instructions tell you to expect your first turnout to take.  There is still a fair amount of craftsmanship involved in making a good turnout even with the aid of all the various Fast tracks tools and jigs.  My first turnout has a rather despicable frog.  Although it seems to work quite well, which is a testament to how much the jigs help you get everything properly aligned.

First Fast tracks code 40 turnout
The above photo shows my first turnout.  The wheelsets in that truck are Fox valley, it rolls through the whole thing very smoothly.  I haven't installed the turnout on the QuickSticks ties yet, because I haven't gapped the frog rails yet, which is in turn because I can't seem to find the jewelers saw I am quite sure I have somewhere.

The hardest part of making the turnout is, as you might expect, getting the points and frog to be really nice.  The points were considerably easier than the frog, oddly enough - the opposite of my experience making the code 148 curved turnout with no jigs.  I think the frog is trickier because you can't fine tune the point by filing the gage face of the rail - doing so pulls the actual point of frog further back from the theoretical point of the frog.  I think I also need to get a better optivisor, and better lighting on my workbench.  But most of all I just need practice.

A not so nice frog
You can see from the above photo that my frog actually seems to have two points.  It's pretty nasty, in fact.  I'm sure I can do better with practice.  But the interesting thing is the geometry of the turnout as a whole is so good because of the precision of the Fast tracks jigs that the shoddy frog point doesn't seem to matter.

And just for fun, a side by side of this and an Atlas code 55 turnout.

Fast tracks code 40 vs Atlas code 55
Both are number 10 turnouts.  Note that the closure rail curvature of the Fast tracks turnout is a little different, so the points are actually further away from the frog than the Atlas turnout.  The difference in rail height is the same proportions as comparing HO code 70 to code 100, or O scale code 132 to code 183 - the last comparison pushes home what's been bothering my about the code 55 track.  The code 148 track on the newer half of the O scale No-Name Industrial looks bigger than I want (I prefer the look of the code 125 on the older half), and the thought of code 183 for O scale boggles the mind!

So, what's the point of all this?  Didn't I already decide to build the Palmer Industrial Park layout in Atlas code 55?  And have I not, in fact, bought all the code 55 turnouts and flex track for the job?  Well, yes.  But I've been repeatedly reminded recently of how much better code 40 looks than code 55, and the recent thread on the Atlas N scale forum on code 40 track got me thinking about it again.  I'm also feeling less pressure to speed through the track laying since I got my O scale No-Name Industrial back into operating shape.  I'll need to do a few more practice frogs and switches to see if I can consistently produce a good code 40 turnout, and I'll need to try laying some plain old code 40 track.  I've already tried some ME code 40 flextrack, but the batch I got had very inconsistent tie height - it just didn't seem good enough to be worth the trouble.

Monday, February 20, 2012

Northeast Container Mockup 2

This weekend I spent some time to finish the Northeast Container Mockup.  I decided my first pass at the mockup wasn't tall enough.  Phil (UP6903) on the Atlas O forum suggested that the building should have a foundation that was the same height as the loading dock, so I tried raising the whole building up by that much using some temporary blocks cut to the right height.

Up on blocks
I left it that way for a couple weeks to see if it still seemed like a good idea after I got used to it.  And it did.  So I made a less precarious foundation with some poplar ripped to the right size, with a little rabbit for the 1/4" ply building to sit in.  The foundation is screwed to the bench work - I wanted to keep the building accurately positioned but still removable.  I'm not sure that was the right choice, but I'm not sure it was the wrong choice either.  Perhaps time will tell.

The foundation
The above photo shows the rabbit on the foundation.  It's screwed down with Kreg pocket screws through slightly over-sized holes, which let me tweak the position slightly before the final turn or two on the screw.  The screw eyes in the top edge are to hook rubber bands on to hold the building in place.

High tech hold downs
The little metal thing the rubber band is looped through on the wall is for picture hanging, it's got two little spikes that you drive into the wood to hold it in place.  Here's the main building in place.  I gave it a quicky paint job with a spray can.

Main building
And the finished mockup with the loading dock in place, and some 1/16" paneling for a roof.

Finished mockup
The whole structure is slightly over 11 feet long - the main building is 64", the narrower pieces is 72".  I'm pretty happy with the overall footprint and size of the building.

An overview
You may be wondering why the backdrop doesn't go all the way down to the layout at the right end of the build.  At the time I put the backdrop up, the plans called for a solid series of smaller warehouse like buildings up against that wall.  I found I was short a sheet of full width masonite for the backdrop, but I did have a sheet that would go down below the roofline of the buildings.  And ever since the plan changed I've regretted that "clever" time and money "saving" move.  I'm toying with the idea of using photo backdrops for a tree line, which would cover up that problem and possibly solve the white window shade problem at the same time.

A closer view
As you can see from the hi-cube peeking out of the warehouse door, the height is about right now.  I think the color is somewhere in the general neighborhood of what I want, but it's definitely not right.  I'm also having second thoughts about the peak in the roof.  Putting it there tells you how wide the building is (twice the distance from the front to the peak), and that's clearly not as big as it should be.  And I think the pitch of the roof is a bit steeper than the effect I was going for.  For now it gets the idea across, gives guests a sense of what the purpose of the operation is, and lets me put off worrying about a building a better version of it for quite some time.

The building mockup with it's roof in place will also let me test another things over time - how to deal with uncoupling a cut that gets spotted in the building.  A couple experiments suggest that with the kadee 805 couplers, it's reasonably easy to uncouple a car placed just inside the building by angling the skewer in through the top of the door.  I don't have enough cars converted to kadees yet though to get an idea of how that will really work out in an op session.

Looking back at the photos it occurs to me that I should explain the yellow things on the track.  Those are yellow sticky note pieces in a new set of experiments with marking clearance points.  My first pass at the experiment used a number I got from a CSX document stating that the clearance point was the point at which the two nearest rails were 13 feet apart.  Using that number on my layout looked good, but ate up a little too much track length for a layout that already suffers a little from too short tracks.  The new number I'm trying is 8 feet between the two nearest rails.  That leaves a little over two scale feet between cars, and adds a car to the capacity of several tracks.  Once I'm sure where I want the clearance points to be I'll follow the prototype practice of marking the rail or tie with some yellow paint.

Thursday, February 16, 2012

Kadees on Berwicks

After putting it off for years (literally) I've finally started making good progress converting the rest of my O scale rolling stock to Kadee couplers, specifically 805 metal couplers in the 817 plastic boxes - what would now be a #816.  Progress has been so good, in fact, that I've run out of the plastic boxes I had left.  I ordered more which should be at the local hobby shop this weekend, so in the meantime I'm writing this blog post about putting Kadees on Atlas O 60' Berwick high cube box cars.

Not all the Atlas couplers are the same length, so the first thing I did was to measure the distance from the end of the car to the outside face of the Atlas coupler knuckle, to make sure I got the Kadee mounted in the same general place.  It turns out the in the case of the Berwicks at least, the holes are in the right place so just re-using them puts the Kadee coupler knuckle face exactly the same distance from the end of the car.  But you can't tell that right off, the first problem is removing the Atlas coupler.

The Atlas couplers are held on with two screws that are inside the coupler box - you have to pop the lid off and remove the spring and coupler to get at them.  Popping the lid off an Atlas coupler box without breaking anything is a real challenge.  Someone once posted a picture of a tool they came up with that makes it easy, I wish I could remember who or find the post so I could give credit, but I can't.  However I was able to remember what the tool looked like, so I made one by cutting a few pieces of 1/16 thick brass, filing the end of one into shape, and soldering together the sandwich with a resistance rig.  Here's the tool in action.

Coupler box tool
Basically, you slip a knife blade in under the cover and exert a little pressure, pull the tab back with the tool, and once you get it pulled back far enough the lid will pop off.

Popping the cover off

Once you've got the Atlas coupler box off you'll notice that the screws are very different lengths.  This is important - the screw that goes in right at the end of the car has to be shorter, because it goes into the little spacer block right over the plastic end of the car.

The spacer block under the coupler box

This is the time to put the car back on the track and use the body height end of the Kadee coupler height gauge to see whether and spacers are needed.  I use Micro-Mark coupler shims.  In the case of the Berwicks, 1 shim is needed.

By comparing the length that stuck out of the Atlas box against different length screws sticking out of the Kadee box & 1 spacer combo, I decided an 8mm screw was about right for the end hole, and the other hole I wanted more bite so I used a 10mm screw. I bought my screws from NWSL - my packages say 31208-5 for the 2.0x 8mm panhead screws, and 31210-5 for the 2.0 x 10mm panheads.  The Atlas couplers are held on with 2.0 mm metric screws, spaced very slightly closer together than the Kadee holes.  By using the same screw size, and enlarging the holes in the Kadee box with a #43 drill I find I can re-use the same holes.

Next step is to prepare the coupler.  With a metal coupler in a plastic box most of the tedious filing and burnishing that's needed with an all metal installation is unnecessary.  All I do is burnish the outside of the knuckle with Kadee greasem, since it's the part that needs to slide by another coupler really easily to couple slowly.

Burnishing the knuckle

And more burnishing just to be sure

It's not that hard to burnish the knuckle faces on an installed coupler, but it's trivial before you put the coupler in it's box so I like to do a thorough job of it.

Then put the coupler together, get the coupler, spacer, and screws all sandwiched up, and mount it.

All the pieces

Putting in the center screw (that goes over the car end) first since it's the shorty makes sure it gets as much of a bite as it can, and of course you have to be careful not to over tighten it.  The other screw will be a little reluctant to find the hole (even after enlarging the holes the spacing isn't a perfect match), but by tipping it a little it will catch and you can then straighten it out.  No need to worry about over tightening the 10mm screw, at least not with a jewelers screw driver.

Installed Kadee
The end result sticks out the right amount, is the right height, and works great.  The Tighe warehouse at the back of my layout (track center line 26" from the layout edge) is the industry the Berwicks head to on my layout.  Uncoupling the Atlas couplers back there can be pretty annoying - visiting friends have sometimes been unable to do it at all.  The switch to Kadees has improved the experience immensely.

Now if those plastic boxes would just come in so I can finish off my Berwicks and get on to the coalveyors...

Wednesday, February 15, 2012


I've planned to equip all my cars with Kadee couplers, because of the improved operation. Kadees will couple much more gently than Atlas couplers, which generally take a moderate slam. Kadees are also much easier to uncouple than Atlas couplers. I loath uncoupling magnets - they're never where you need to uncouple, except on rare occasions. If you do uncouple over one, you either leave the car with it's coupler sprung to the side so you can't couple to it, or you have to back and fill to shove the car back past the magnet. And of course they always seem to show up under spots where you stopped without intending to uncouple. I prefer to uncouple with a skewer like tool.

At the recent Springfield, MA train show Kadee displayed their new O scale couplers for the first time. My first thought was "Aha! My procrastination has paid off!" So I bought a couple pairs of the #745 "prototype" head medium centerset shank metal couplers with plastic boxes. Previously I'd used only #805 metal couplers in the metal boxes. I was pleasantly impressed at how easy it was to prepare a metal coupler in a plastic box - none of the filing and burnishing I'd done on the metal in metal ones appears to be necessary. And the #745 couplers look nicer than the #805s.

The real test came when the couplers were on the cars and in operation. The #745 couplers coupled very easily - pushing a car along at under 1 scale mile per hour to couple up to a lone free standing car still resulted in the knuckles sliding past each other and coupling, instead of just pushing the free standing car down the track on the coupler face like the Atlas couplers do.

Uncoupling was a different story. Here's the two different couplers, coupled and under slight tension for comparison. Click on the photos to see a larger version.

Kadee #745 couplers coupled

Kadee #805 coupled

You can see that there is a lot more open space between the 805s. The 745s still have a little, but they're definitely tighter.

Here's my arsenal of uncoupling tools.

Uncoupling tools - best are center, and right.

On the left is the Micro-Mark tool. Some people seem to like it, but for me it just doesn't work for O scale couplers at all. Second is the Accumate Switchman, which doesn't seem to be nearly as useful as I'd hoped but it does work up to a point. Then we have a narrow skewer that works well for HO kadees, a slightly thicker skewer with a screwdriver like tip carved into the end opposite the point - my current go to tool for O scale since with the mixture of couplers I have I need both ends. The next tool is the same size skewer as the center favorite, but shoved into the carcass of a bic pen after pulling the ink cartridge out. The idea being that you can put it in your pocket without stabbing yourself (with the cap on), but still have enough point to uncouple. I have to admit it falls into the category of nice idea but doesn't work as well as I hoped. Second from the right is an extra large skewer, and finally on the left is the classic #2 uncoupling tool, which can also be used to take notes, mark up switch lists, etc.

The first technique I tried to uncouple the #745 Kadees was the classic - get slack, drop the skewer in the hole in the middle, and twist. That does not work, since the knuckle has a squarer corner which fits more perfectly into a squarer opening. The result is putting the skewer in the middle wedges the knuckles into the couplers, the knuckles can't open far enough to get by either the skewer or each other, and you're stuck.

745 knuckle wedgie

In order to uncouple the 745s you need to have just the right amount of slack - either tension or compression makes it nearly impossible. The screwdriver tip end on the medium skewer tool then works pretty well. But since the opening you're aiming for with it is pretty small, it's pretty hard to uncouple on a rear track (my rearmost track is about 26" from the aisle). Uncoupling on a curve is slightly harder, again because of the tighter fit.

Here's the same uncoupling tool (the center skewer from the above photo) in a pair of 805s.

805 knuckles not wedged
Notice how much bigger the hole between the knuckles is than the 745s - much easier to get the tool in there on a rear track. Also the knuckles don't get wedged, and you can twist to uncouple even if the couplers are under compression. The medium sized skewer will just sit in the knuckle once you twist it open, and you can leave it there while you pull away.

After the twist

The twist open technique doesn't work as well with a Kadee coupled to an Atlas, and it doesn't seem to work worth beans with two Atlas couplers. But in both those cases you can use an alternate technique that requires slightly more careful aim of sliding the skewer in at the knuckle tip as shown in the "after the twist" photo about, skipping the twist step. There is a certain middle aged vintage of Atlas couplers for which that technique doesn't work at all though, which is why I have the screwdriver tip carved in the other end of that skewer.

If you operate with switch lists you're familiar with the need to mark them up and scribble notes. Currently I'm using car cards, but I have done some sessions with hand written switch lists and I hope to eventually get computer generated switch lists working. The point is, you need a pencil anyway, and with Kadee 805 couplers a pencil is also an effective uncoupling tool, using the same drop in the hole and twist method as the skewer. The only difference is the pencil won't stand in the open coupler by itself like the skewer will.

#2 uncoupling tool in action

The bottom line is I like everything about the new Kadee O scale couplers except the way they uncouple. But that is important enough to me that I'm going to stick with the old 805 couplers. However I am sold on the metal coupler in the plastic box, so I've ordered enough of the #817 plastic box sets to cover the rest of the 805 packets I have on hand.

Wheelset Widening

A while back I bought a couple MTH 50' high cubes, with the TTX FBOX reporting marks.  These come as 3 rail, but MTH also sells a 2 railing kit for them that has a set of decent 2 rail trucks.  The cars are already pre-drilled for mounting Kadee couplers.  The conversion went pretty quickly and the cars were on the railroad in short order.

The first move of the next operating session was pulling them out of the interchange track where I had put them on the rails.  The first sign of trouble was some serious bouncing and clunking going over turnouts.  It didn't take long to figure out that the trouble was all the wheels were too narrow in gauge.

A tad narrow
The most annoying part of this was that I had to take the trucks off again, which requires removing the floor of the car since the screw goes down through the floor into the truck bolster.  Anyway, I got the trucks off and dismantled and started working on the wheelsets.  I couldn't seem to shift or rotate the wheels on the axle by hand, and I didn't want to use a puller and mess up the pointed axle tip, which serves as a bearing.

The solution I came up with is a crude but effective wheel gauge widener:

Widener - 2 side pieces, 2 wedges, 1 hammer
I cut a slot in 2 squares of 1/2"plywood, and a longer slot into 2 tapered pieces of pine I cut on the table saw.

Widener in use
Using it is simple.  Stand up the two squares and drop in the wheelset, slide the wedges in from the top and bottom until everything snugs up about in the middle.  Then CAREFULLY tap with the hammer.  Two or three light taps is enough to get the gauge dead on 6 out of 8 times.  The other 2 times I gave it an extra tap, and had to press them back together just a tad, which is also easy by moving the 2 squares to the outside of the wheel and lightly tapping.

The cars now go through all my trackwork with no problems.