The constraints are:

1. The vane linkage cords must first run down to the first turning point on the axis of vertical rotation of the vane turret, so that they describe a conical shape as the turret rotates, thus not changing in length.
2. They must run more or less athwartships from the second turning points to the little "tiller" on the servo carrier.

In between the two turning points, there is freedom. On Tystie, there is an extra turning point as the cords come down from the vane (mounted on the pushpit) to the first turning blocks in the centre of the davits crossbeam, then at an angle to second turning blocks at the top of the transom, then down the transom to the third turning blocks on the transom, then athwartships to the servo.

The red cords can just be made out in this photo [taken as we were perched on top of an uncharted rock in a remote lagoon in BC, Canada!].

That's right. I drilled two 6mm holes inwards through the groove, either side of the turret support, and hollowed out the underside of the disc to make room for the knots.

That's not the only way of doing it, though. The cords have to be secured in that position, but maybe you can see a way you like better. 

Kris wrote:

Since you located the fix point of vane cords in proximity of servo axle there will be no change in the distance / linking cord length when swinging.

I start today  making the vane assembly and am waiting for the servo drawings.

In fact, Kris, the placing of this point is important - it's where the essential negative feedback is applied in this kind of servo pendulum gear, and there is a controlled amount of change in the vane linkage cord length. The vane linkage cords are attached to the servo carrier at 120mm out from the servo blade's vertical axis, and ~40mm down from the 15˚ inclined servo carrier axis. Thus, as the vane puts an angle of incidence onto the servo blade, and it swings out to one side as a consequence, the angle of incidence is reduced in direct proportion to the amount of swing.

 Detailed drawings, as complete as I want to go at the moment, are at:

https://app.box.com/s/tmwm42be9auvgx7kz0xyo9ycyz7fez9x

I think I'll go ahead and make the detail drawings anyway, because there is a wider application for this - it would suit many small boats where the vane assembly could be mounted on the pushpit. The mounting frame could be narrow, for an inboard rudder or twin rudders, or wide, for a transom-hung rudder. It's more conventional than the Hebridean-principle gear, and as easy to build.

The major assembly drawings are now in:

https://app.box.com/s/tmwm42be9auvgx7kz0xyo9ycyz7fez9x 

You're welcome! Do you want me to make dimensioned drawings, or will you take it from here?