the switch throw wire can be passed through the fulcrum to engage the hole in the throwbar.  Once it is screwed to the actuating arm of the Tortoise, the switch is operational. 

Switch stand actuation

It doesn't take a great deal of testing a newly installed switch before we reach the conclusion that it's really difficult to tell which way the switch is thrown.  The obvious resolution to this problem is to install a target on the switch stand, just as the prototype did.  In HO, the common way to build a working switch stand is to follow the prototype's example, but use the action of the points to rotate the target, rather than the other way around. 

In our case, this is infeasible as the switch only moves about a millimetre.  To rotate the target through 90 degrees would require a crank a mere 0.7 mm long.  Not only that, but getting the dimension wrong even by 0.1 mm causes the rotation to be either much too great (110 degrees) or much too small (76 degrees).

Fortunately, we have another piece of the mechanism that moves through a distance of about 12 mm --  the actuating arm of the Tortoise.  This amount of motion means that we can make the crank about 8 mm long.   To take advantage of this motion, we need a switch stand spindle that passes through the roadbed and pivots down near the bottom of the switch motor. 

Installation started with drilling a hole in the roadbed for the switch stand's spindle.  For high level switch stands, this hole should be approximately five feet from the near stock rail; low level switch stands would be perhaps three feet [Rench, 26. P54].  The prototype spindle is 7/8 in. square section rod, which I represented with .25 mm (.01in) wire. To keep the spindle vertical, I glued a small capillary tube in the hole through the roadbed. 

Like the throw bar, the pivot for the spindle mounts on the Tortoise's rails.  However, the rails end about 2 cm (.8 in) above the actuating arm.  In order to obtain the maximum motion, the actuating arm needs to be close to the actuator on the Tortoise.  So we need to extend below the rails, as shown in Figure 5.  My pivot is mounted on a piece of plastic that measures 3 cm by 6.5 cm (1 1/8 by 2 1/2 inches).  At the lowest part of the pivot mount, I glued two pieces of the larger diameter of square tube.  These tubes guide the target actuating arm. 

The pivot itself is simply a piece of 1mm (.04 in)styrene glued to the mount.  The pivot hole is located by dropping the spindle through its hole in the roadbed and marking its position.  The location is not critical provided the spindle is vertical above the roadbed.

The actuating arm is a piece of the smaller section of square tube with a 1mm styrene rectangle glued to the end.  The styrene rectangle has a slot to accept the end of the spindle crank and allows it to travel through a radius as the arm travels back and forth.

With the actuating arm centred so the slot lined up with the pivot hole, I marked the middle between the two

guides.  This located a slot that accepts the switch throw wire.  As the switch throw wire moves back and forth it engages the sides of this slot, causing the actuating arm to move back and forth as well, and rotates the target.  As shown in Figure 4, this slot must be V-shaped.

The spindle is made from fine wire, bent to provide the required 8 mm crank.  It passes through the pivot and should have enough flexibility to snap into the slot in the actuating lever.  To install it, I started the spindle in the hole in the roadbed and slid the pivot mount onto the rails of the Tortoise. 

I tested the spindle's motion by moving the actuating arm back and forth.  Only then did I install the actuating wire, and apply power to test.  To fine tune the target's rotation, I moved the actuating arm up and down until I was satisfied.

References

Gourley, René. "Building a 1906 Turnout", Proto:87 Journal #6, Proto:87 SIG, 1997.

Lowing, Doug. "Thin throwrods for turnouts", http://www.geocities.com/ResearchTriangle/Lab/5872/throrod1.html, September 1998.

Ottaway, Stephen. "Making a Point", Proto:87 Journal #6. Proto:87 SIG, 1997 (reprint).

Rench, Walter F. Practical Trackwork Switch stands, switches, frogs, crossings and slip switches. General Publishing Co. New York, 1926.