Atlas Code 55 Curved Turnout

[Rossford Yard]

3 of the 4 I had on my last layout worked fine, including one entering the helix in an inaccessible location. One was out of line at the plastic frog as noted above, but was fixable so no real problems.

That said, I avoided them on my new layout, just in case. I still remember having trouble with Atlas HO curved turnouts years ago. Seems like it's just tough geometry to make work, although my Peco curved turnouts never gave any trouble.

[signalmaintainer]

I appreciate that we can have a reasoned discussion on TRW of the pros and cons of Atlas code 55 curved turnouts. There are known issues with them that perhaps vary in frequency and magnitude from one production run to the next, I wager.

It's a real contrast compared to another forum where the general attitude of some is, "Well, I haven't had any issues with Atlas code 55 turnouts, so you shouldn't either." That's not really solution-oriented.

[dem34]

I found that a large portion of Atlas Turnout issues are at least mitigated if you run a bead of thin CA along the spikeheads at the ends of the rails. Especially the outer frog rails.

[mmagliaro]

Upon re-checking, I think the #10 22/18 is a better match than the 20/16.   It's a close call.  It really depends a little on what your trackwork is like around it.   In any case, since this thread subject has come up multiple times (i.e. "How are Atlas c55 curved turnouts") and I always say it is the "first thing I'd replace"...   I decided to put my FastTracks where my mouth is and order the darn template and supplies and just make one.  I went with the 22/18.

-

[ednadolski]

Upon re-checking, I think the #10 22/18 is a better match than the 20/16.   It's a close call.  It really depends a little on what your trackwork is like around it.

Is it a pure curve or does it have (small) straight sections in it?

If it is a pure curve then one could string a complete circle of them together in a track drawing program and see what kind of circle fits best.

Ed

[mmagliaro]

Is it a pure curve or does it have (small) straight sections in it?

If it is a pure curve then one could string a complete circle of them together in a track drawing program and see what kind of circle fits best.

Ed

That's a pretty darn good idea.  Looking at the Atlas turnout, I *think* it is a big curve if you just look at the outer stock rails.

The Atlas product description says they are 21-1/4" / 15" for the two radii, which is probably why
the FastTracks 20/16  and 22/18 each don't quite match, but are close in different ways.  They do make a #6 22/14 and a #8 20/15. I think I would still prefer the more gradual divergence of a #10, seeing as I know from the paper template that it will still drop right in.

[mmagliaro]

Just to close the loop on this...
I went with the 22/18 and built the FastTracks. 

It fit right back in where the Atlas curved turnout was.  I think I had to make a slight adjustment to the track on the diverging route of about 1/8".

I DID make some minor improvements to it.  I added a few extra soldered PC board ties in addition  to the ones specified in the jig.  One was about in the middle of the turnout, and the other were at the ends of all 3 legs (the incoming track and the outgoing two diverging tracks).  I just wanted more certainty that it was going remain rock solid in its shape and geometry over time.

And although you can't see it in the photo, I did drill diagonally through the lower heels of the ends of the point rails and put .015" phosphor bronze wires through the point rails into the PC board throwbar before soldering, just to make sure those solder joints never break.

[pedro]

And although you can't see it in the photo, I did drill diagonally through the lower heels of the ends of the point rails and put .015" phosphor bronze wires through the point rails into the PC board throwbar before soldering, just to make sure those solder joints never break

Looks great, and I like this KISS method of simply pinning the points to the throwbar for reinforcement. Any closeup pics? Does the wire go entirely through the throwbar and get soldered to the underside?

[mmagliaro]

Not a great photo, but this is all I've got.

I didn't do anything fancy.  The bronze wire is bent in a kind of L shape, where the short leg of the L goes through the heel of the point rail down into a hole right through the throwbar, and the long leg of the L lays down on the throwbar itself.  Everything gets squeezed gently so it lays down flat and out of the way, and then I solder, and then file (which is why the "near" one in this photo looks square shaped.  The far one was a smaller.  But honestly, after a bit more filing and painting, you can't even see this thing.

[Ed Kapuscinski]

@DeltaBravo paying attention here? I used a number of these in your plan.

I know I am, cause I've got a number of them in my OWN plan as well.

[bman]

So been reading this over and over. Thanks for all the good input. I do like the idea of building my own. I build several as a kid in HO when we moved to the country and the large plywood central I had was dismantled (not well either) for a move. I literally took curved and straight sectional track and made a couple of spring track switches for a small switching layout I had room for in the small farm house my parents purchased.
Now that being typed, I only need 1. And that kinda sorta defeats the purpose if I have to buy the jig. Using the Atlas code 55 with the advertised 15" inside radius helps me keep the curvature of that end of the layout as it is a 15" radius and fits nicely into the plan. And that is becuase I'm using the free version of Atlas track planning software and only get Atlas track options. Not sure my hands are steady enough to scratch a turnout out of N sectional track. But I do have a supply of old Atlas N code 80. I guess it doesn't hurt to try what's the worst thing that can happen but I waste some time? It only has to be a spring switch anyway. 

[davefoxx]

I'd suggest just paying someone to build the necessary curved turnout that you need.  That will be a fraction of the cost of a jig used for a mere one turnout.  You might find someone here on TRW that will build one, and, if not, I think you can purchase them on eBay, too.

Hope this helps,
DFF

[mmagliaro]

I don't see how Atlas specs their curved turnout as having a 15" inside radius.  I replaced mine with a 22/18.  No way would a 15" radius have worked.   It must be that their turnout has a more gradual radius on the two diverging exit tracks than it does through the points, so it averages out to a 15" piece overall, but it had more like an 18" on the exit so it mated up with my 18.75" radius curve.  That's really awful.  It means the point area must be pretty sharp on the Atlas turnout.  All the more reason to get rid of it.

[nickelplate759]

Related to that (the true radius of the inner curve of the Atlas curved turnout), I found that the model of this turnout in SCARM doesn't seem to quite match reality on that inner curve.  I printed out a SCARM plan full size and the curved turnouts didn't quite fit.  I was able to get them to work with a little twiddling and adjustment, so not a big problem.

and on another related note, anyone know of a library for the FastTracks curved turnouts for SCARM?  The supplied libraries don't include curved turnouts.

[Pomperaugrr]

At one point I had six Atlas code 55 curved turnouts on my layout.  I now only have two left on the layout.  Those two remaining curved turnouts are preproduction samples provided to me by a certain, unnamed, hockey playing now former employee of Atlas for testing and feedback.  These have operated flawlessly since I got them.  The other four purchased curved turnouts have all failed, either electrically or mechanically.

If I need curved turnouts in the future, I would definitely go with FastTracks either homemade or sourced from a builder. 

The 60 other #7 and #10 Atlas code 55 straight turnouts have been reasonably reliable, with only a couple of point failures caused by the round retaining clips falling off.