Stiffness of Fast Track crossover points / servo mechanism redesign

[DKS]

DKS...you just draw that up like SHAZZAM and there it is???  I'm impressed!

Used to do that sort of stuff for a living, long long time ago. If I couldn't draw it up like SHAZZAM, I'd be shown the door...

[jdcolombo]

Because there is no room for Tortoises and they are too expensive for this application.  It's not just one double crossover — it's a concentration of four doubles and two singles, all packed tightly together.  It makes far more sense to develop a small-footprint servo mechanism incorporating the Tam Valley controller.  The advantage is the lower cost and more straightforward electrical arrangement; the necessity is that the compact footprint of the mechanism fits directly and completely underneath the crossover end with no overhang.  With 1-1/8" between all the track centers and seven tracks deep and 20 points to throw, Tortoises are not a viable option.  And one mechanism, which throws two points, costs less than one discounted Tortoise.

I suppose it is possible to devise a way to have one Tortoise throw two symmetrical sets of points.  But using the Tam Valley controller and a servo serves a number of purposes without pushing the cost into orbit.

Got it.

And I agree that something like David's linkage suggestion would be better for this purpose.  You might find commercial linkage that could be adapted - Circuitron, for example, has a linkage that permits remote location of a Tortoise, and I've seen diagrams of using two of these linkages with cables to operate two turnouts with one Tortoise.  It's possible that these could be adapted so that you could control 4 turnouts at once (and there's no reason you'd have to use a Tortoise as the "driver" for this; anything that would actuate the linkage would work).

John C.

   

John C.

[bbussey]

Someone I guess lurking here sent me a PM on TrainBoard suggesting placing the current servo mechanism directly against the crossover and making cutouts in the layout top to clear the entire mechanism.  I manually tested this configuration and it does work, but I like Mike's/David's linkage idea better so I'm going to pursue that.  One thing I did learn is that a wire length of ¼" moves the points without issue.  So instead of having the pivot wire above the ties, I may see if I can bury it ¼" underneath the throw bar by cutting clearance in the foam.  The hole can be covered with .005" styrene with a slit to clear the throw rod, a trick that Scott Lupia taught me.  So I won't have track running this coming weekend, but I will finalize the new design by then.
 

[Sokramiketes]

DKS's solution will work beautifully but involves more precise measuring/tinkering than what I was thinking.

I was thinking more like the Circutron Tortoise Remote Mount Kit.  The kit includes the crank assembly, a mount for the tortoise, and material for the really tiny bowden wire push rod that connects the two. 

Now, Bryan mentioned cost as a concern...  Since the bowden wire is like .012" diameter at the sheath, and .010" or so for the spring wire inside, what about running it up directly to the throwbar and burying the PVC sheath in the scenery? 

I only started playing with the remote mount kit because I had clearance issues on the first Modutrak issue and was looking for a way to mount Tortoises on their side.  The kit accomplished that and the discovery of the bowden wire was a bonus.  At the time I had sourced extra PVC sheath through Small Parts Inc.  Music wire will work for the wire.  Then you can run straight from servo to throwbar with a flexible setup that needs no fancy planning!

[Chris333]

You spent a lot of time on those servos. Could you cut the 1/8" ply and 1" foam and mount your servo block right to the balsa wood? Seems like that would be the easiest and quickest thing to do.

[DKS]

I only started playing with the remote mount kit because I had clearance issues on the first Modutrak issue and was looking for a way to mount Tortoises on their side.  The kit accomplished that and the discovery of the bowden wire was a bonus.  At the time I had sourced extra PVC sheath through Small Parts Inc.  Music wire will work for the wire.  Then you can run straight from servo to throwbar with a flexible setup that needs no fancy planning!

This sounds vaguely similar to a technique I used on Rick Spano's layout in some tight spots where a Tortoise would only fit on its side. I just very roughly aligned the pivot point of the arm in the Tortoise with the tube carrying the crank; accuracy was unnecessary because the end of the wire just dropped into the screw hole on Tortoise arm. Lots of slop, but it didn't matter; in fact the slop made any fancy linkage unnecessary. Should also work just fine with a servo.

[nkalanaga]

Dwyane:  I used the same method on my dual gauge turnouts.  The problem there is that, even with #8 turnouts, the points are too short to bend easily, and there's no room for a rail joiner hinge. 

For anyone worried about durability, the filed bases don't seem to weaken the rails, as none have broken.  I doubt that there's enough bending to make metal fatigue a factor in normal use.  On a club layout, maybe...

[SkipGear]

Bryan,
 Be careful, DKS's linkage setup won't work without a few more parts. It can be done but each pivot will need it's own link. As the pivots rotate, they will get father and farther from each other. One link tying them together will not let them move.

Your existing mechanism can be used to control 4 turnouts instead of two, espcially since all 4 linkages will move in unison. Just stretch it a little.



Bend the turnout actuator wire  U shaped so it connects to both throws on one line of the crossover. Solder a another wire to that so it will become the servo lever connection. You may need to use slightly thicker wire and stronger servo to overcome the rails because they are doing double duty now.

Hopefully that makes sense. The fewer the moving parts, the better. Less things to go wrong.

A slightly stronger method might be to create 4 Z bent wires and join them on the verticle by soldering a brass tube over them.

[robert3985]

This post deals with closure point hinging techniques.  I'll leave the bellcranks and linkages up to the rest of ya!

Dwyane:  I used the same method on my dual gauge turnouts.  The problem there is that, even with #8 turnouts, the points are too short to bend easily, and there's no room for a rail joiner hinge. 

For anyone worried about durability, the filed bases don't seem to weaken the rails, as none have broken.  I doubt that there's enough bending to make metal fatigue a factor in normal use.  On a club layout, maybe...

I have manufactured and used the "filed base" hinge for decades and I've never had one fail.  However, using code 55 rail (or larger) if the points are soldered to a PCB throwbar, and a thick actuating wire is used to throw the switch, the throwbar soldered joints will eventually break, especially if they're short (of prototype length).

The problem is that although they work really well and robustly as a simple hinge, if the points are solidly attached to the throwbar (even if a tab is bent from the rail foot at a right angle to the points and used as a large pad to increase the area to be soldered to the throwbar) they will eventually fatigue and break.  Of course, you can simply solder them back to the PCB throwbar again, but sooner or later, the throwbar will delaminate due to both the torsion and heat.  It's a PITA to replace them, especially if the actuating wire is in the middle of the throwbar as you've got to disassemble what's underneath the switch to be able to insert a new one, then you gotta re-register the wire through the hole in the throwbar, which when there's scenery on top of everything, means it's most likely a two-man operation.

Photo 1: "Filed Base" hinges which, as you can see, involve filing the head of the rail also.  I file mine so that I just kiss the rail web on either side:


Photo 2: "Filed Base" hinges, prototypically short closure point rails with unprototypical right-angle tabs at the toes of the points acting as attachment shoes for the point toes:


These look like they ought to work, and they do for a while using code 55, then the point solder joints fail.  However, using code 40, they DO work reliably, but I don't like the unprototypical look of the bent tabs at the toe end of the points.  I consider this a failed design, and I don't build my points/hinges/throwbars like this any longer because I want more realism and total reliability, neither of which this design allows.

What happens is that even with hinges, short hinged throwbars form a rigid parallelogram and the soldered throwbar joints are the weakest link when they are thrown.  Ideally there would be hinges on each corner and both the heel blocks and throwbar hinges would not just be hinges but would allow for the end-angles to change also.

There are a couple of ways to accomplish this one of which is the traditional railjoiner slip-hinge at the heelblocks.  This keeps the heels of the closure point rails aligned with the ends of the closure rails coming from the frog, AND allows linear movement as well as angular movement, which de-stresses the points soldered solidly to the PCB throwbar. Micro Engineering #6 closure point hinges are made this way.

Another way is to hinge the toe of the closure points at the throwbar, and build a little slop into the throwbar so it can have angular movement but still be confined between the headblocks, which will do two things (1) completely de-stress any attachments at the closure rail toe and (2) eliminate any solder joint between the toe and throwbar.  But, this method, although prototypical, is complicated.

The quickest, easiest method is to use Proto87 Stores heelblocks/hinges and solder the toes of your closure points to a N-scale PCB tie between PCB headblocks.  The Proto87 Stores parts are pretty cheap (although the price went up recently) and are etched from nickel silver and allow both linear and angular movement. Photo 3 shows them installed on a recent batch of turnouts for my Emory Center Siding section.  Photo 4 shows them installed along with prototypically short closure point rails and throwbar toe hinges of my own design.

Photo 2:  Proto87 Stores etched heelblocks/hinges:


Photo 3: Proto87 Stores etched heelblocks/hinges with Closure Point Rails installed & Toe Hinges:


Although this information probably won't assist the OP, maybe anybody else who either builds or is thinking of building their own turnouts might benefit from my mistakes over years of making my own turnouts, as well as solutions that are easily available and/or can be built.

[DKS]

Be careful, DKS's linkage setup won't work without a few more parts. It can be done but each pivot will need it's own link. As the pivots rotate, they will get father and farther from each other. One link tying them together will not let them move.

Actually, it does work as shown, because I've used the very same design to actuate an animation device having four parts that moved symmetrically, just like the crossover points. The parts that connect the point linkage pairs (the parts that look like large washers or short sections of tubing) just need to be slotted very slightly, or simply large enough in diameter to allow a small amount of play. The key factor is this: the point movement is small enough that the separation which takes place between point links as they rotate is very small. Where it wouldn't work is if the points moved more than just a small fraction of an inch, or if the parts connecting the point link pairs together were very tight. Trust me, I've been making devices like this for quite a long time.

I've illustrated the crux of the issue below. Even when the amount of point travel is exaggerated, the amount of play that must be accommodated--even if doubled--is minuscule. The parts would need to be machined to high tolerance in order to prevent it from working.



Tony's arrangement should also work, except that you'll have the same problem you have now of overcoming the bending of wires across the 1.5-inch distance through the roadbed. Where there may be additional trouble is if different point pairs require different amounts of force to throw completely; his design does not easily accommodate compensating for the possible need to move the rods different amounts. Z links exhibit relatively little wire flexing, so the movement of the links should be more consistent.

[bbussey]

You spent a lot of time on those servos. Could you cut the 1/8" ply and 1" foam and mount your servo block right to the balsa wood? Seems like that would be the easiest and quickest thing to do.

Chris, I did try the current mechanism butted directly against the underside of the crossover and it worked.  But I'm not feeling adventurous enough to cut 3"x2" openings under each crossover end and find out I have additional problems.  Fortunately I don't have much invested up front, because I wanted to get one working before moving ahead.  I like the crank idea, one reason being that I only need one controller and one servo per crossover.  I've got the 3D model for the new base roughed out.  Once I decide on the washer/spacer that will serve as the pivot point, I can send it out to Shapeways for rendering. 

[bbussey]

... The parts that connect the point linkage pairs (the parts that look like large washers or short sections of tubing) just need to be slotted very slightly, or simply large enough in diameter to allow a small amount of play ...

David, do the pivot washers travel in a track of some sort, or are they free-standing?  And do they consist metal, or some slippery plastic?

[DKS]

David, do the pivot washers travel in a track of some sort, or are they free-standing?  And do they consist metal, or some slippery plastic?

I'd envisioned them free-standing, since they really don't need any sort of track; they would, however, be rigidly mounted to the ends of the links that tie them to the rocker arm. As to the material, I leave that up to you. I think they could be made out of just about anything, so I'd choose whatever material you're most comfortable using.

[nkalanaga]

Bob: I agree that the rigid throwbar joints are the first to fail, but my turnouts have two advantages.  First, the throwbars are 1/16 inch square brass tubing, with a piece of sheet brass between the points, so the rail base is mechanically held top and bottom by the solder.  Not suited for DCC, maybe, but filing the brass thinner outside the points and ACCing a piece of paper insulation on each side works fine.

Second, the throwbars themselves are short, and connected to the slide switches I use for throws by brass rods, which are not soldered to the throwbar.  Thus, the throwbars themselves can rotate a little, which seems to help.

Even with that, I have one that comes loose semiregularly, probably because the outside (standard gauge) point is hard to brace on a dual gauge turnout.  The narrow gauge point is in the way, preventing the use of a long brace.

Your turnouts look beautiful, and I'm sure they work just as well, but it's too late to change mine.  They've been down for 16 years!  If I had to start over those Proto87 parts look like the way to go.

[robert3985]

Bob: I agree that the rigid throwbar joints are the first to fail, but my turnouts have two advantages.  First, the throwbars are 1/16 inch square brass tubing, with a piece of sheet brass between the points, so the rail base is mechanically held top and bottom by the solder.  Not suited for DCC, maybe, but filing the brass thinner outside the points and ACCing a piece of paper insulation on each side works fine.

Second, the throwbars themselves are short, and connected to the slide switches I use for throws by brass rods, which are not soldered to the throwbar.  Thus, the throwbars themselves can rotate a little, which seems to help.

Even with that, I have one that comes loose semiregularly, probably because the outside (standard gauge) point is hard to brace on a dual gauge turnout.  The narrow gauge point is in the way, preventing the use of a long brace.

Your turnouts look beautiful, and I'm sure they work just as well, but it's too late to change mine.  They've been down for 16 years!  If I had to start over those Proto87 parts look like the way to go.

I'd like to see some photos when you've got some time.  Yeah, I've got about 20 turnouts on my layout that are built using my "old" method.  The code 40 ones operate reliably, but the C55 ones break periodically.  I'm still figuring out a method to change the point attachment protocol at the toe (I think I'm homing in on it) on the "old" ones.

Most likely, I'll just keep re-soldering until the PCB throwbars delaminate...then fix each one on an individual basis when that happens since replacing the throwbar involves dismounting the Tortoise underneath, then re-aligning it when the new throwbar has been installed.  Don't wanna do that more than once!

Thanks for the positive comments.  I think we'd all like to see close-ups of your complicated trackwork as well as the mechanism underneath when you get that designed and installed.