I've become convinced that continuous-motion escapements are especially suited for wooden works clocks. The main reason is that the escapements can occasionally pull on the gear train, in addition to being pushed by it. This should make the clocks naturally tolerant of sticky spots on the teeth. It should also allow the clocks to work with much lighter driving weights.
Let me start by saying that conical pendulum clocks of the type made by Farcot seem like the best solution. They also happen to be things of beauty:http://youtu.be/QO_oUpGRSGc
First, though, I want to mention a couple other devices.Fly Fans
Nobody will argue that fly-fans are the ideal heart for a precision timepiece, but they are a good proof-of-concept. I've done a lot of looking, and find that fly fans have been used to control the speed of various mechanical devices, including lighthouse lenses. In clocks, they are commonly used to control the chimes, where the repeated start-stop action can cause wear and breakage. But if the fans are allowed to turn continuously, that liability becomes an asset. The fan's mass can pull on the train, making poorly balanced and rough wheels functional.http://www.youtube.com/watch?v=bJhIGWiMwwg
If my memory is right, air drag is proportional to the square of speed, which means that if you want to double the speed of a fly fan, you'd make the weights four times heavier.Brake-type Centrifugal Governors
Also known as Centrifugal Brakes, these devices control various light-duty machines, such as movie cameras, rotary telephone dials and wind-up phonographs: http://www.youtube.com/watch?v=7bfzLoHiXTkhttp://www.youtube.com/watch?v=aB2AWckXCgEhttp://www.youtube.com/watch?v=8ye2Bj6reEc
I've also found these in lighthouse mechanisms. But they are also used in clock drives for telescopes, which need to be exquisitely precise to make time exposures. It's therefore puzzling that they are seldom used in plain old clocks.
I can think of one possible disadvantage: the speed of these devices is much greater than the escape wheels in clocks with pallets. This implies that you'd need to have an extra wheel and pinion. Ordinarily, you'd expect an extra wheel to multiply the friction load, but as I've been hammering here, continuous-motion clocks should have naturally low friction. Another benefit: the weights don't have to accelerate the gear train after every pendulum swing.Conical pendulum clocks
Besides the Farcot clocks, a well-known brand is the Briggs Rotary Clock. The Farcot ones correctly keep the pendulum at a small angle to the vertical, while the Briggs ones seem to have an uncomfortably large swing. There is one variation which is critical to putting them in a wooden works clock. Usually, the bottom end of the pendulum rod rests against a horizontal arm which rotates, but sometimes this arm is replaced with a two-tined fork. The two-tine arrangement will allow the weight of the pendulum to occasionally pull on the train.
I'm not a woodworker, but I'm eager to read what people think about these mechanisms. (They're not even usually called escapements, but I can't see a better term.)
I'd also like to know if I'm right that replacing the pallets with a heavy fly-fan would allow much lighter driving weights. If it works, it would allow beginners to make a clock that actually works, and would free experts to spend their efforts making interesting complications.