I`ve not posted an update for a few days and the truth is, I`d made all the parts required to test the mechanism and I must have spent near 10 hours just tweeking the setup to get it to run. Finally, just as I was giving up again for the evening, I had a brain wave based on the advice and suggestions given to me by John/Gadgetbuilder; I made some changes and off it went!

I`ve many corrections to make to the working mechanism and you will spot a huge error in the working of the backstop – notice when the backstop engages the large tooth,it jumps/catches on the count wheel. I think this is just the position of the backstop pawl but I wanted to leave it and let it have some time running. It didn`t effect the physical mechanism but it will cause errors in the time, so it will need correcting. I have a much much better idea now of how this all fits together so I`m now very confident I can get it working more smoothly!

Here goes, I just did my best to video it in bad light at this time of night.


1. The pendulum swings 40 periods each minute, gathering a tooth on the 40 tooth count wheel. The count wheel therefore rotates once every minute and so the seconds hand could be placed on this wheel but it would turn in reverse. Consequently, Wilding didn`t have any seconds indication but Woodward placed numbers on his count wheel to approximate the seconds.

2. The count wheel has a single deeper tooth than the rest. When the count wheel pawl (the top wire on the pendulum) drops into this deep tooth, it engages the vertical wire on the deflector piece. At 56 seconds in the video, you can see the count wheel pawl going over the top of the deflector and imagine how it interacts when it hits the deep tooth.

3. The deflector therefore tips anticlockwise and this causes the diagonal piece of the deflector piece to obviously tip down as well. You can see the deflector mounted on the backplate below (the other bit you can see at the bottom of the deflector is a stop to stop it tipping all the way back at rest).

4. This bottom piece of the deflector bears down on the lower wire from the pendulum (the impulse pawl) which engages a tooth on the pin wheel.

5. You can`t see this on the video but there is a small weight (a large allan key!) hooked directly onto a tooth (at about 3 oclock position) on the pin wheel. The tooth on the pin wheel is engaged with the bottom lip of the deflector piece. When the impulse pawl pulls the pin wheel back by a fraction, this releases the pin from the detent and the weight on the pin wheel, which in future will be a proper weight and chord, pulls the pendulum back to the right and gives it the impulse!

6. There are 60 pins on the pin wheel and so this rotates once an hour giving you the arbor for the minute hand.   A fancy “daisy motion works” will then need building to gearlessly reduce down this motion to 1 turn per 12 revolution of the minute hand/pin wheel arbour.

Hope that in some way explains how it runs!
Some more pictures for you of bits I`ve taken as I did a quick polish ready for this test assembly.

Back plate.

All the bits ready for a quick polish. This isn`t a final polish, I just needed everything clean to get it to run.

Count wheel

Back stop

I made another count wheel with shallower teeth a few days ago but managed to get the original one working so didn`t bother using it yet. I might try this one as I expect that the back stop is jumping because the tooth is overly deep on my initial wheel I`m using.

Pin wheel and deflector piece:

same as above, with the gated detent:

Well I can now go to bed much more positive than I have been doing for the last few weeks!

Good night!