My skeleton clock

Thank you Martin, I appreciate your comments.

To answer your question - No, I haven’t tried a threaded mandrel, although I’m tempted to drive over to a friends toolroom and see if he will let me use one of his lathes.

Although wrapping the wire around a threaded mandrel was suggested by several members in another ME thread (Balance Springs II - under the topic - Clocks), I have to say that while I loved screw-cutting, for this purpose the Hobbymat doesn’t impress me one bit. I miss my Myford too much!!!

In his ME series c.1972, John Stevens, the clock designer describes in a fair amount of detail, how he made the (flat-section) spring. It involves more than just winding however, because he takes us through the heat-treatment process.

Norman, Thanks for your ideas. I’ve responded to your PM, as you may know by now.

Regards to all,

Sam

The new pictures show the three essential elements which are a part of the skeleton clock `coming alive’.

The fusee and barrel can be seen connected together via an 80lb (36kg) breaking strength braided fishing line. It was a nice stroke of luck that the line was yellow. It seems to go well with the brass.

During winding, the line is transferred (upwards) from the barrel to the fusee at a decreasing rate, thus providing a type of torque compensation as the main spring approaches being fully wound. I’m not convinced that the profile of the fusee, generated by a 2" radius, accurately compensates for the increasing main-spring torque. But it’s going to have to do, cos I won’t be making another. However, while I was winding it up completely for the first time, I could feel the compensation taking place.

Also visible in this picture, although not easy to determine, is the stop-work mechanism designed to prevent over winding, and thus avoiding having the line drop off the end of the fusee. It is identified by the notched steel arm pointing away from the camera. As the line is wound onto the fusee and travels towards the front of the clock, the spring-loaded steel arm is deflected into alignment with the snail-shaped hook, just visible under the small end of the fusee.

Although concealed within the `great wheel’, there is a blued-steel circlip-shaped spring called the maintaining spring. This provides continuing power to the clock during winding. While it is not obvious, there are 29 items (including screws) which make up the entire fusee mechanism.

The next picture shows each of the important parts of the escapement.

I deliberately chose a long exposure time to display the movement of the escapement. Hence the reason for the blurred image. There are already pictures in this album showing this part of the clock before the clock came to life. The oscillation of the lever is just visible as a double image above the fifteen-toothed ‘scape wheel. At the top of the picture is the over-stiff spring, responsible for the fast beat. It is made from 0.011" (0.28mm) diameter wire, whereas my calculations suggest it should be 0.008" (0.2mm) diameter. Other calculations reveal that this thicker wire produces a spring which is approximately twice as stiff as it should be.

My last picture to be added shows the bell and hammer in place. This `once-on-the-hour’ strike is just about to take place in the picture. However, because the clock is running at more than twice the correct speed, it currently strikes every 28 minutes. The hammer is lifted via a pin on the `minute’ wheel and `lifting piece’ positioned at the front of the clock. It was necessary to introduce a spring which would stop the hammer from bouncing more than once on the bell. This seems to work OK.

The rough-looking screw just about the hammer head provides end adjustment for the `fourth’ wheel (contrate), who’s teeth are turn at right angles to engage with the 14 toothed escapement wheel pinion.

Hi Dick,

Thank you for your comments, and for following the clock’s progress. Once again, your comments fit into the picture perfectly.

Now here’s an update on my headway, and a couple of responses.

The clock may require some more tweaking of the balance wheel weights, but I’m pleased to note that it is running very well indeed.

Since I have no data on the validity of the main-spring/fusee `compensation’, I suspect that there could be `patches’ where the `going’ train is over powered. However, for various reasons, I’m compelled to accept that the main spring is here to stay, and changing the fusee profile is not an option either.

For more information with lots of friendly responses, please see five other threads via :- http://www.modelengineer.co.uk/forums/searchresults.aspSearch=spring&SearchType=1&t=133

As you can see here

the balance wheel has tiny screws with large heads. These are positioned around the rim, the intention being that they provide a means for time adjustment. To increase the mass of the balance wheel, I slid and glued a brass bush onto the heads of both of the two main adjustment screws, and began a more serious approach to getting the speed correct. With more mass and tweaking of the screws several times per day, I finally achieved an accuracy within three seconds per hour. It ran for about five weeks, with occasional but not yet a full winding.

I was therefore satisfied that the clock could be pulled apart for a final scrub up.

Now the clock was running too fast, but gradual unscrewing of the main screws has produced an accuracy equal to or even better than the previous results.

Sam