More Experimenting with Micro Servos as Switch Machines

[C855B]

I'd still like the frog to participate in detection. I've seen enough flakiness with detection through the years dissuading me from giving a pass to a null region in the circuit, no matter how small. This is especially the case with the plan of only one resistor axle per car.

No, the driver and relay board will likely be mounted on the benchwork frame. The plan, subject to revision, is to put all electronics on the frame against the wall, or bracketed under the frame, on connectored panels removable for debugging.

I did say relay board. I was testing the relay function just a little while ago, and this looks like the plan. Frog Juicers are nice and will likely be the solution for manually controlled turnouts (if I have any!) but the relay setup with the Singlet II is straightforward and has an extra set of contacts for signal logic.

[John]

Nice work .. the cost seems prohibitive for my tastes ..

[nickelplate759]

I note you used CA to glue the servo to the ties.   Given the mechanical stresses involved, I'm wondering how it holds up.   Have you had them in service this way long enough to find out?

[C855B]

It holds up just fine. Absolutely no failures to this point, in eight months with a couple of dozen turnouts in operation this way.

A couple of things I'm doing which may help - first, I wait 24 hours before setting the point alignments, which put shear stresses on the join. While CA establishes a firm bond very quickly, it's not going to be fully cured for a while. I think too many of us are tempted to "test" CA joints after, oh, 10-15 minutes, and IMO this weakens the bond. Second is once the alignments are set, there's no need to operate it until installed in the roadbed. Since I mortise the servo into the cork and foam underneath, and glue the turnout to the roadbed for installation, operating forces are nicely spread around.

But even when operating too much in testing and doing stupid things like letting the servo controller self-align - it cranks HARD(!!!), assuming HO turnouts, I think - I have had zero issues with CA bonds breaking. It just works.

[mmyers]

I don't know if this servo controller would be of any use. The board has terminals for led indication which might be useful to drive a relay or solid state circuit to power the points.
Here's the link: http://www.modelrailroadcontrolsystems.com/servo-controller/

[DKS]

All you need to control frog polarity is a cheap surplus lever microswitch, which would be installed after the servo is in place: the lever simply has to be in contact with any moving part of the system. You can do this almost literally for pennies a turnout.