More detail of my 'Cape Horn' experience -
In the late 1990's a long time friend of ours bought a centre cockpit Bristol 41 and fitted it out for cruising. As he lived in Ottawa he got the designer Yves Gélinas to install the Cape Horn gear, but by the time they had sail down the St. Lawrence and crossed the Atlantic to south Spain he admitted he couldn't get the gear to steer the boat for more than a few minutes. Joan and I flew down to Spain to spend a few days with them, but our only chance to try out the gear was for about an hour in a near flat calm, with not enough wind to move the heavy boat at 2 kts.
The boat was Med-moored, bows to, so with no dinghy afloat it was not possible to see the geometry of the installation. The steering quadrant output of the gear was inside the lazarette with a cat's cradle of lines and block coupling it to a small axillary quadrant on the head of the rudder stock, with a 2:1 mechanical dis-advantage, before going to a pair of jambing cleats which were the 'engage' method. It looked a theoretical attempt at a totally impractical lash up. When I tried to drive the rudder by swinging the paddle I was horrified to find how difficult it was as the friction was so high, but even worse, the mechanical connection to the large centre cockpit steering wheel could not be disconnected so that when the wheel did start turning with the rudder the flywheel effect took over, and it did not want to stop. It was a disaster area, and I could not see how it would ever work.
In our short time under way, in an effort to see the gear work we motored up to about 6 knots and then stopped the prop so that the paddle was not effected by the prop-wash. I tried to steer the boat by tilting the vane by hand, and it did not work. The paddle would float up to the surface to one side or the other, not necessarily the correct side for the vane inclination, and when the vane was returned to upright the paddle would not go back to vertical. As I was concerned that the coupling to the boat's steering was inefficient I also ran the gear uncoupled from the rudder/ wheel, and again found the paddle would not follow the commands from the vane even when I was controlling it by hand. I did check that the vane/ paddle steering rod was not reversed.
When we returned to the marina I went through the manual and found that the paddle was much longer than recommended when self installed, and from memory had over a metre immersed when vertical. At the time I guessed that the extra volume under water was causing the paddle to float to the surface, but when moving at over 3 kts it still should have followed the vane commands. I measured the paddle chord and balance about its turning axis and found it to be just under 25%, which should have been stable but sensitive.
After we flew home my friend shortened the paddle to the recommended immersed length, but to no effect. The final solution was to remove the gear and return it to Yves in Canada at a significant financial loss to my friend.
My overall impression was that it was a pretty looking gear, but of a theoretical rather than a practical construction for the job. The paddle was of rather narrow chord compared to the accepted successful gears and quite finely balanced and I guessed that this was the reason for the rather small wind vane compared to the established gears. The ratio of the area of the vane to the area of the immersed paddle if trimmed to about 60cm long were probably reasonablel. When seen in practice the coupling from the gear quadrant to the rudder quadrant was a ridiculous mechanical disaster, and the 'back driving' of the wheel to rudder coupling even worse. Even if the gear had steered the boat, the wheel to rudder coupling would have died from excess wear driving the inertia of the wheel, so it would have been a totally impractical set up so live with.
The gear should at least have been able to follow the manual inclinations of the vane at a water speed of 3 – 4 kts when uncoupled from the rudder, but as it couldn't then the thought had to be that the paddle was inclined forward, but even if that was the case the paddle still should have tried to drive to the side it was turned towards. It made no sense.
For the designer/ manufacturer to install it with this result beggars belief. My friend completed a circumnavigation using electric autopilots. He even banged his head on Alan Martienssen's yuloh on Zebedee.
Cheers, Slieve.