|Over Christmas 2009 I started on a new timepiece with a proper going train. I started with some plans which had gear templates which you were supposed to photocopy and paste to high quality plywood, then cut around with a scroll saw. I tried this, and couldn't get workable gears. With timepiece gears the precise form of the gear is not too critical, you only have to look at some of Harrison's work to see what you can get away with, however, the gear forms from the plans were too poor to work at all.
Here it is, finally hanging on the wall in Summer 2011. I still have problems keeping it accurate. It tends to loose a few minutes a day, so still some work to do to it.
|The item from the original plans which did work was the escapement. It is a Brocot with slightly odd teeth on the wheel. These are simple in conception, and easy to make, but horribly inefficient at delivering their impulse. The bearings for the escapement started being a double small ball bearing, but as with the second wheel, ball bearings are not suited to these very low force stop/start applications, and they tend to detract far too much from the turning effort given by the great wheel, so were replaced by plane bearings runing on shafts caterlevered from the single clock plate.|
|This is a clear photograph of the wheels and pinions of the motion works. With wooden timepieces the main problem is getting the gear forms to work. The first system I tried was a very clever way of automatically making the epicyclic by forms using the geometry of circles passing along a straight edge. It was invented by Steve Garrison, and can be found here. The method was admirably ingenious, but blade defection on my scroll saw was too great to get accurate tooth form, and the other way of using this system meant getting your fingers too close for comfort to a table saw blade, so I abandoned attempts using this.
The way I eventually found would work was to make up paper templates using a gear generator program from a Canadian chap called Matthias Wandel. These I could make to any specification, including the large diametrical pitched gears I wanted to make.
|This is the click. The pulley is toothed, allowing a decent purchase on the braided line used to suspend the weights from. The ratchet paw is clearly visible, with a spring made from a bit of junior hacksaw blade to keep it nicely engaged with the ratchet wheel.|
|The pendulum is suspended directly from the shaft upon which the anchor is mounted. This is an area where a double ball bearing system works well. It is capable of moving with little friction, and simutaniously supporting the weight of the pendulum. Having said that for the MkII it might be better to adopt a more traditional frictionless bearing with a crutch. There is a frictionless bearing here, but is simply to let the pendulum have a bit of give in relation to the anchor.
The aluminium plate you can see slightly below the pendulum suspension is a plate. This is where the second wheel shaft is mounted. This plate is adjustable so the deapthing between the great wheel, and second wheel pinion, can be made exactly right. One of the problems with a wooden clock is that the wood will move, and of course this throws the deapthing out. Any MkII will have the cantilevered shafts on a steel frame which will in itself be conceled within the clock plate.
|The rear of the clock plate is hinged to allow access to the click and pendulum suspension. The whole cabinet is made from elm which my uncle, my father, and I cut down in the mid-1970s when Dutch elm disease started to infest all of Britains elm trees. Despite the timber being really well seasoned it still moves a lot.
Any MkII would be drastically redesigned using the lessons learned from this one. I also fancy a strike, so may start on a new clock as soon as I've worked out how strikes work.