- Harrison claims to have achieved an accuracy of one second a month with his regulators, something I will try to match with this wooden regulator.
- Use no lubrication in the movement.
- Segmented wooden wheel construction, including the escapement. Use Acacia Rhodoxylon, a very dense and stable hardwood from outback Queensland. The pallets of the grasshopper escapement to be of Lignum Vitae.
- Use the cycloidal tooth form on the wheels instead of Harrison’s chordal pitch gearing. Similarly, I used cycloidal brass pinions, instead of the lignum vitae roller pinions. Ideally the escape wheel pinion should be a roller pinion to absorb the recoil.
- Use hybrid ceramic and stainless steel ball races are used throughout, instead of anti-friction wheels. The pendulum suspension also uses ball race suspension.
- The Grasshopper Escapement, Peter Hastings produced two designs for me, the constant force escapement, and the 2:3 ratio variable force escapement. I selected the constant force type for this clock. Pallet tooth form was modified to create safety action and prevent run-away.
- The remontoire is used to eliminate any variation in torque force delivered to the escape wheel through the drive train. It also frees the escapement from the effects of engaging friction when the pendulum is driving the escape wheel backwards during recoil. The remontoire layout for this clock is derived largely from H4, and as with the RAS regulator, is powered by a small spiral spring. I will leave the constant force remontoire used on H2 and H3 for another time. Ideally the remontoire should power the escape wheel as in the RAS, but this would necessitate a 120 tooth escape wheel, which would be difficult to construct reliably in wood.
- Temperature compensation: The temporary pendulum is made of a length of well seasoned rosewood. Likewise the bob is temporary. When time permits I would like to construct a grid-iron pendulum, as well as experiment with fused silica rod.
- Circular error compensation: Harrison used Huygens’ solution of cycloidal cheeks, but there is evidence that he modified it on the RAS regulator to a circular form. Either way, very fine tolerances are required to get the system to work as intended. Instead, I have deviated significantly from Harrison’s solution, and used a system developed by Dallas Cain and Pierre Boucheron using twin pendulums (Horological Science Newsletter 1993-1, 1994-1). I am not aware of any other clocks using this system.
- Barometric compensation: I doubt if this clock in its current form will be accurate enough to warrant any attempt at it its inclusion.
- The clock is currently running in a temporary stand. To get the rate to within I second a month, it would have to be housed in a cabinet, and securely mounted onto a masonry wall.
