There were types of pendulum driven clocks using mechanical switch contacts since the late 1800s, such as the factory "Master clock", and gadgets such as the previous electromagnetic pendulum system I posted going back to the 1950s or earlier - but none I am aware of that used sensing from the power coil?
Nice work! Following Harrison's line of thought, the shorter the period of the oscillator, the finer you can tune the clock. So your present setup has a ?half second period and quite a large swing. Once you have the electriks sorted, I would be inclined to go for a much faster period and much shorter swing, and use a knife edge on a glass slide as the pivot. Your setup lends itself well to being fitted into say a 100 mm dia PVC pipe with end caps. You could then evacuate the container as much as you can (a bike pump with the valve reversed would do) to reduce friction. Then don't mount it on the same wall as any other pendulum as they will synchronise in spite of best efforts at tuning (as noted at Greenwich). Great job tho!
The driver circuit should be adaptable to suit any pendulum period and mass, by changing the drive pulse & blanking durations and the drive current. My interest is mainly the electronic side, that other people can copy to use in their projects. Also, I think a nice slow, visible, moving pendulum is part of the attraction of older style clocks! Re. pendulums in general; as I understand it, the swing angle should be less that 15 degrees to maintain accuracy? - as well as being thermally compensated, as with properly made clock pendulums. I believe it should be adjustable to self-regulate the swing angle, if the pulse duration is set to match the time it takes the magnet to move slightly clear of the coil with the correct swing, & the drive current is then also adjusted for ideal swing. If the pendulum is then moving too fast, it will have left the coil before the pulse ends, so receive less energy?
That's something I mentioned as a control method, and what I tried in the next prototype - but it does not give stable control, which is why I've not done another video on this yet. Prototype: robert-jenkins.co.uk/site/images/Pendulum_Driver/prototype_small.jpg What I discovered from experimenting with this is that it needs an adjustable delay between the trigger and the start of the pulse, in addition to the variable pulse length: With the pulse starting as the magnet passes dead centre in the coil, there is very little energy transfer to the pendulum until later in the pulse, when the magnet is some distance away from the centre. At the exact mid position, there is no direction to the repulsion between the fields. The faster the movement, the more energy it picks up, in other words. The most effective energy transfer will probably be when the magnet is around half way out of the coil, so a short pulse at a point around there should give a self-stabilising effect. ie. If the pendulum is moving too fast (too wide a swing) it would already be beyond the peak energy point when the pulse occurred. The next version is in progress, I've just been busy with work so not had much time to do anything on these. (The other added features are constant current drive, so battery voltage variations should not affect the pendulum drive power - that's set by the wedged-in resistor; and the option to separate the sense and power coils, for some other applications).
brilliant! looking for just such a device to help out a >> 100 year old clock movement that gets cranky every so often...
Has it been cleaned and oiled recently? It made a world of difference for me.
We had a clock with this system in1940. Must have been twenty years old then.
There were types of pendulum driven clocks using mechanical switch contacts since the late 1800s, such as the factory "Master clock", and gadgets such as the previous electromagnetic pendulum system I posted going back to the 1950s or earlier - but none I am aware of that used sensing from the power coil?
Nice work! Following Harrison's line of thought, the shorter the period of the oscillator, the finer you can tune the clock. So your present setup has a ?half second period and quite a large swing. Once you have the electriks sorted, I would be inclined to go for a much faster period and much shorter swing, and use a knife edge on a glass slide as the pivot. Your setup lends itself well to being fitted into say a 100 mm dia PVC pipe with end caps. You could then evacuate the container as much as you can (a bike pump with the valve reversed would do) to reduce friction.
Then don't mount it on the same wall as any other pendulum as they will synchronise in spite of best efforts at tuning (as noted at Greenwich).
Great job tho!
The driver circuit should be adaptable to suit any pendulum period and mass, by changing the drive pulse & blanking durations and the drive current.
My interest is mainly the electronic side, that other people can copy to use in their projects. Also, I think a nice slow, visible, moving pendulum is part of the attraction of older style clocks!
Re. pendulums in general; as I understand it, the swing angle should be less that 15 degrees to maintain accuracy? - as well as being thermally compensated, as with properly made clock pendulums.
I believe it should be adjustable to self-regulate the swing angle, if the pulse duration is set to match the time it takes the magnet to move slightly clear of the coil with the correct swing, & the drive current is then also adjusted for ideal swing. If the pendulum is then moving too fast, it will have left the coil before the pulse ends, so receive less energy?
Shorten the pulse time! Adding a resistor is just lost power. Shortening the pulse is better for battery life!
That's something I mentioned as a control method, and what I tried in the next prototype - but it does not give stable control, which is why I've not done another video on this yet.
Prototype:
robert-jenkins.co.uk/site/images/Pendulum_Driver/prototype_small.jpg
What I discovered from experimenting with this is that it needs an adjustable delay between the trigger and the start of the pulse, in addition to the variable pulse length:
With the pulse starting as the magnet passes dead centre in the coil, there is very little energy transfer to the pendulum until later in the pulse, when the magnet is some distance away from the centre. At the exact mid position, there is no direction to the repulsion between the fields. The faster the movement, the more energy it picks up, in other words.
The most effective energy transfer will probably be when the magnet is around half way out of the coil, so a short pulse at a point around there should give a self-stabilising effect. ie. If the pendulum is moving too fast (too wide a swing) it would already be beyond the peak energy point when the pulse occurred.
The next version is in progress, I've just been busy with work so not had much time to do anything on these.
(The other added features are constant current drive, so battery voltage variations should not affect the pendulum drive power - that's set by the wedged-in resistor; and the option to separate the sense and power coils, for some other applications).
bedini
i like this! we could make a 3d printed pendule clock