0 Members and 1 Guest are viewing this topic.
I've enjoyed this thread immensely, particularly the range of clever mechanical solutions to the multiple crossover challenge. Can't help but think though that given the number of moving pieces involved, this will be a high maintenance setup with too many places where the mechanical installation can fail with little room to work in to tweak things. On a larger layout, where there are other things that inevitably go wrong and this just seems like asking for trouble, IMHO.I'm also not sure that making all four points on each double crossover move simultaneously necessarily simplifies things.... If I were doing this, I would use a Tam Valley servo under each set of points and think of "routes" through multiple crossovers as the most efficient way to align the appropriate point and only move the points necessary for a given route rather than twice as many. This could be done through a matrix, as John C. suggested earlier, or in my case because I have DCC, simple "Macro" programming which would throw all the points required at the same time. This would take two servos and one controller per a set of two points, ie one crossover. I have multiple macros set up in large yards with up to 10 sets of points moving in unison and to me, that's the most elegant solution for this situation involving such complex trackwork without too many opportunities for mechanical failure. Just food for thought....good luck!Regards, Otto K.
I'm also not sure that making all four points on each double crossover move simultaneously necessarily simplifies things.... If I were doing this, I would use a Tam Valley servo under each set of points and think of "routes" through multiple crossovers as the most efficient way to align the appropriate point and only move the points necessary for a given route rather than twice as many. This could be done through a matrix, as John C. suggested earlier, or in my case because I have DCC, simple "Macro" programming which would throw all the points required at the same time. This would take two servos and one controller per a set of two points, ie one crossover. I have multiple macros set up in large yards with up to 10 sets of points moving in unison and to me, that's the most elegant solution for this situation involving such complex trackwork without too many opportunities for mechanical failure.
I just can't stand to design something that relies on "slop" to work properly. Where there is slop, there will be more slop as it wears.
Also with the DKS setup, you will have to limit servo travel pretty drasticly which limits the power and control range of the servo.
Servo's are what I do. (RC car racing for 25 years, along with planes and heli's.) Electronically limiting the throw on a servo makes them very erratic and unreliable. You want to try to use as much as the throw of the servo as possible. I know this is not a high stress application like I am used to dealing with but it still goes against all that I have learned over the years.
... I'm also not sure that making all four points on each double crossover move simultaneously necessarily simplifies things.... If I were doing this, I would use a Tam Valley servo under each set of points and think of "routes" through multiple crossovers as the most efficient way to align the appropriate point and only move the points necessary for a given route rather than twice as many. This could be done through a matrix, as John C. suggested earlier, or in my case because I have DCC, simple "Macro" programming which would throw all the points required at the same time. This would take two servos and one controller per a set of two points, ie one crossover. I have multiple macros set up in large yards with up to 10 sets of points moving in unison and to me, that's the most elegant solution for this situation involving such complex trackwork without too many opportunities for mechanical failure. Just food for thought....good luck!
There also is no room for 14 Tam Valley controllers and 28 servos in that confined a space. That is more of a consideration than the cost.
Because there is no room for Tortoises and they are too expensive for this application.
You might have to spend some time adjusting things, but that should only be once. If you want to avoid that part, then this one goes one step further: it includes equalizer arms that make the system self-adjusting. The equalizers will allow the points to fully close even if there are slight variations in the degree of motion necessary for each set of points.It all depends on where you want to focus your energies. The bottom line is that one design can be tweaked a thousand ways to suit the builder.
The equalizers certainly improve the design but it is still a difficult install. You need to build the mechanism on the layout with the bent wires going through the pivot bushings, then getting bent into place. Certainly not a servicable design and difficult to bend 4 throw wires exactly the same way. The equalizers certainly help solve that problem but you still either need to build it on the layout or leave rather large holes for the mechanism to poke through the layout and actuate the turnouts. Part of the idea of the original design was that all the layout required was a small hole for the actuator wire to go through the roadbed and insert into the throwbar. The entire throw mechanism was built on a removable/adjustable base plate which makes building and servicing much easier. Stuff fails and it needs to be serviceable. I may be over-engineering this but experience has told me, servos fail, turnouts fail, things will go wrong and it needs to be easy to fix. Your solution is not one I would consider. Too many moving parts and parts that are hard to recreate repeatably. KISS - keep it simple stupid. Bryan has a couple of these to build, not just one. I will bow out of the discussion. The original solution worked just fine, it just needed stiffer wire.
And he'd have twice as many of them to build as these.
The original design was "proof of concept". I already showed how to expand it to allow one servo control all 4 points of the crossover and it only requires one more part and an additional pivot pocket added to the original design per side.
There also is no room for 14 Tam Valley controllers and 28 servos in that confined a space. That is more of a consideration than the cost. The servos cost $2.60 each, which is next to nothing, and two TV controllers can be thrown with a DPDT toggle, so the costs are comparative.If you have a better way of achieving the task, and have a way to mount the servos on adjacent tracks with 1¼" centerlines seven tracks deep, I'm more than open to suggestions.
Very nice N! I'd love to see a closer shot of a throw bar - and some more shots of your layout.
Author
Topic: Stiffness of Fast Track crossover points / servo mechanism redesign (Read 8661 times)
