I was working on an erratic movement recently, and trying to figure out why it kept stopping, here are a couple of tg-timer graphs that illustrate the issue very nicely.
Just for the record, the issue was a tiny spec of dirt on the edge of one of the teeth of the pallet wheel of a Timex model 32 movement, but the movement in question is not really relevant.
Note how on the vertical chart (the time axis) the rate speeds up as the dirt comes in to contact with the first pin of the fork, then speeds up further and becomes a little erratic when it hits the second pin, only to then slow back down again as that tooth passes and the mainspring manages to return some energy to the balance. Also note that this is a gradual increase, the recovery is not instant, since the balance needs several cycles to get back up to full amplitude, and back in to beat.
Here is the inevitable result when the dirt eventually causes sufficient disruption to get the thing to stop.
Note how the ballance continues to bounce around, but since it has insufficient energy to allow the pins to clear the tooth, its amplitude just drops off a cliff and it stops.
One final point. Although I like tg-timer, it has one glaring failure, in that it doesn't put time marks on that vertical chart, which would be useful, as it would allow you to see that the disruptive event takes place at one minute intervals, which is also how long it takes for this particular palette wheel to make one complete revolution.
If you are set up, a video replayed slow, would reveal what is going on.
If you are thinking, Pallet jewel position a hairy edge, you can lower or raise the escape wheel.
I keep in mind, the fork interfaces functionally at both ends, pallets and horn. , if it is happy at impulse jewel side, I wouldn,t touch the fork.
Good luck Pal.
eBay has a plethora os movements, with case still attached, for next to nothing. Horological schools usually recommend a particular movement for class work and practice. However, if you simply want to pull it apart to see how it works and then try to put it back together, buy anything that ticks. You can pick up a “dollar watch”, times, etc or other vintage unmarked movements for next to nothing. Seeing things move together is far more important, IMHO, than a specific movement.
do be careful! You may spark a new expensive hobby! :-)
cheers from Texas
Cleaned, de-maged and checked h/s for stickyness. Nothing to be observed from that. Did another test you may comment on. With balance removed, I gently nudged the pallet fork side to side to start the release-lock operation. At a couple of close to each other positions on the escape wheel circumference this is not functioning as I believe it should. It all happens fast and is hard to observe but I think there is a ”skip over” where the locking doesn’t happen as it should but locking is happening on the next tooth. As escapement teeth look decent, could this be a case of pallet jewel position on the hairy edge? Sent from my iPad using Tapatalk