Hi
I have had the clock running for over 6 months now but have struggled to get it to keep to good time , it regularly gaining between 5-7 minutes a day , but then running to time for a week or so
I was sitting quietly watching the clock and it suddenly pushed the second hand ahead 2 seconds ahead for several ticks before going back to normal
it has happened twice in 15 mins which accounts for the time gain , but not sure how to cure it ,
the static and active pawls are set as best I can with only about 1mm backlash . but this would not account for the sudden double ticks
ghe Second hand is not always in the same position when it happens as I thought it may be unbalanced and pulling it round
any ideas ??
Hi Hoppy
I have completed your mod acouple of weeks ago , the clock now runs very reliably , however the extra weight of the magnets have made it very noisy when the arm drops down on the ratchet gear , to the point it where was banned from the living room .
My solution was to super glue a very thin piece of felt (a piece of chair foot protector sliced with a sharp knife) ) approx 1/2 mm thick to underneath of the arm .
This has taken it back to its original noise level and is accurate to +- 1-2 secs a day ,which I am more than happy with
Thank you!
Take a look at a new post that I made. Watch the video and see if this would help you.
Leon
I always felt like there could be a better pawl configuration that was more tolerant of overswing. Suppose the clock needs 2 degrees of pendulum swing for a single tick but won't double tick even with 10 degrees of swing. I couldn't quite get something like that working.
My suspicion is that many issues with double ticking are caused by unequal weight or friction in the ratchet gear. Sometimes the pendulum amplitude seems to change as the ratchet rotates. Possible causes might include a bent second hand arbor or unequal friction on some of the gears.
Early prototypes of the clock would double tick when the second hand was around 3 o'clock and the pendulum would lose energy around 9 o'clock. This was caused by the extra weight on the long pointer. The counter weight was re-designed and needs to be printed with 14 perimeters or 100% infill to be nearly perfectly balanced.