I have access to (OK I have wound it twice a week for 20+ years) a flat bed tower clock in the local church. It occurred to me recently that creating a model of it it might make a good build project (once I get the current list out of the way in a year or so!).

It is a three train flat bed clock made in 1889 by Smiths of Derby, and still maintained by them.

The movement is fairly simple, pinwheel escapement, count wheel chime and strike. Each train has a limited number of wheels. All the pinions are lantern, and the bearings are brass bushes, some in iron castings individually mounted on the top and bottom of the bed so depthing should be fairly easy. The bed is one piece cast iron – but I would probably fabricate it from steel.

The real thing is about six feet wide. It beats 3150/hour, and trips the chime train four times for each revolution of the output shaft (closest to the camera on the image above) . Thinking it through I t seems to me that if I made a 1/4 scale version it would beat twice as fast, and if I tripped the chime train twice per rev I would get a clock that looked OK, and kept to time. The motion work behind the dial ( about 20ft above the clock) is driven by a pair of bevel gears.

By halving the ratio of these gears I should be able to get the hands to turn at the right speed.

I can't see anything too difficult (apart from having to learn how to cut wheels!!) except possibly the pair of bevel gears that send the output up to the dial and those on the motion work, and the latter could probably use a standard commercial set as they will be non scale and hidden behind the dial.

The chime barrel will be fairly fiddly to get right, but in principle should be doable.

Before I set to measuring the real thing, counting teeth and planning a bit harder, I thought I would ask on here to see if anyone has done something similar – or if anyone can think of any major gotcha's I might run into.

Edited By Peter Cook 6 on 30/10/2021 16:23:31

Thanks Michael, I suspect that as a clock it is actually less complex than the Wilding design in many ways. The trains are simpler, and as far as I can see the clock was a "mass production" design, so the bearing carriers are identical front to back. The great wheels on all three trains are the same, as are all the lantern pinions and the wheels for the chime and strike – so once I have sussed out how to make one – I will be into production, and the second one of anything I make is usually better than the first.

And it has the overriding advantage that I can simply have a look at the real thing whenever I get stuck.

Thanks Martin, more issues to muse upon. I had not really considered the weight issue. The real clock has 185Kg on the chime train, 110 on the strike and a mere 27Kg on the time train a grand total of 322 Kg hauled up 10 metres twice a week. The Chime and strike are three pulley systems, the time is only two.

I think I was hoping to get away with rather less than the 80Kg that comes from dividing by four, but I think some serious math on run times and weights is going to be required. I have a couple of longcase clocks with barrels about 2" in diameter that run for eight days, so if I am careful with pivot quality on the time train I might be OK.

A quick calculate suggests that with scale barrels I will need about  2metre of cable for a 24 hour run if I allow it to run at double speed. But getting 80 pins on the pinwheel might be a challenge! Lots to think about. It looks like this will take all the time until my backlog is cleared just thinking about the issues.

Thanks again – its the things you didn't know you didn't know that are the most helpful in these forums.

PS Michael – While browsing the Burton blogs up popped a link to my clock. Most surprised!!

Edited By Peter Cook 6 on 31/10/2021 17:25:17

Thanks Martin. More useful thoughts. At twice speed the great wheel turns once every 1.5 hours. Scale drum is 36mm diameter and 90mm long so 113mm of cable per rev and 16 revs per 24 hours = 1.8metres per day.

With 96mm of drum length, 8 days allows 12mm/day of width so max cable thickness is 12/16 or 0.75mm.

Thin steel cable @0.6mm has 20Kg breaking strain.

100lb breaking strain fishing line is 0.5mm diameter.

Hopefully that sort of pull will do it. However I will as you suggest start with the escapement and experiment.

The real pendulum is 1.5m long, so lengthening at scale it is a possibility.

I think I am beginning to understand why people build to existing sets of drawings – scaling reality is complicated!