ETA 2804-2. Pallet fork won't swing freely?

[VWatchie]

So, I'm in the process of troubleshooting a manual would ETA calibre 2804-2 (background info in this thread).

I'm trying to apply the "function group method" shown on the Chronoglide Youtube channel (video here) which I think is a rather brilliant idea, and I've come to the pallet fork.

So, with only the pallet fork and the pallet fork bridge mounted on the otherwise bare main plate (actually unlike in the video where the pallet fork is tested with power on the train), I am trying to make the pallet fork swing back and forth by simply tilting the movement holder left and right.

Previously I have performed this kind of test on a couple of Raketa movements (calibres 2609.HA and 2623.H) and on those occasions I could just tilt the movement holder about 5 to 10 degrees or even less to make the pallet fork freely swing from one banking pin to the next. However, not so with this ETA 2804-2. It's only when I've tilted the movement holder about 70 degrees that the pallet fork swings over to the other side, and sometimes I even have to gently tap the movement holder to make the pallet fork move.

The pallet fork jewels have been meticulously cleaned, pegged, and checked at 40X magnification on my stereo microscope and I can detect no problems. Also, the pallet fork staff pivots and pivot shoulders look perfect. The end-shake has been adjusted to make sure there's a healthy gap (with a bit of a margin to make absolutely sure) between the pivot shoulders and the flat surfaces of the jewels. Still "the little bugger" won't move when I gently tilt the movement holder  and I can't think of anything else to check 

My only thought is that perhaps the pallet fork of these ETA 28XX movements is constructed in such a way that the weight on both sides of the pallet fork staff is equal. That is, the pallet fork has no incentive to swing when the movement holder is tilted because the pallet fork has the same weight on both sides of the staff.

So, what's your thoughts!?

Edited July 9, 2023 by VWatchie

[RichardHarris123]

If evenly ballanced, the fork should stay still if the plate is rotated.

[Neverenoughwatches]

1 hour ago, VWatchie said:

So, I'm in the process of troubleshooting a manual would ETA calibre 2804-2 (background info in this thread).

I'm trying to apply the "function group method" shown on the Chronoglide Youtube channel (video here) which I think is a rather brilliant idea, and I've come to the pallet fork.

So, with only the pallet fork and the pallet fork bridge mounted on the otherwise bare main plate (actually unlike in the video where the pallet fork is tested with power on the train), I am trying to make the pallet fork swing back and forth by simply tilting the movement holder left and right.

Previously I have performed this kind of test on a couple of Raketa movements (calibres 2609.HA and 2623.H) and on those occasions I could just tilt the movement holder about 5 to 10 degrees or even less to make the pallet fork freely swing from one banking pin to the next. However, not so with this ETA 2804-2. It's only when I've tilted the movement holder about 70 degrees that the pallet fork swings over to the other side, and sometimes I even have to gently tap the movement holder to make the pallet fork move.

The pallet fork jewels have been meticulously cleaned, pegged, and checked at 40X magnification on my stereo microscope and I can detect no problems. Also, the pallet fork staff pivots and pivot shoulders look perfect. The end-shake has been adjusted to make sure there's a healthy gap (with a bit of a margin to make absolutely sure) between the pivot shoulders and the flat surfaces of the jewels. Still "the little bugger" won't move when I gently tilt the movement holder  and I can't think of anything else to check 

My only thought is that perhaps the pallet fork of these ETA 28XX movements is constructed in such a way that the weight on both sides of the pallet fork staff is equal. That is, the pallet fork has no incentive to swing when the movement holder is tilted because the pallet fork isn't heavier on the side that interacts with the roller pin.

So, what's your thoughts!?

Eyup Watchie, I guess it can be either or both of two things, the weight of the pallet fork doesn't shift over until its reached a particular inclined angle ie. more than you are anticipating. Or friction, in one or many places. Even though you have  endshake could a pivot still be resting on a jewel as you tilt until it releases from the jewel ? Or could it be a binding of the two pivots across their jewels while at an angle, one reason that olive shaped jewels are used in quality watches as opposed to straight sided jewels. I think there is an illustration in one of de carle's books, i will try find it and pop back with a picture. De Carle practical watch repair page 30. That might just create enough friction. I suppose there is no set ruling as to when the lever tips over with gravity as long as it does move.

Edited July 8, 2023 by Neverenoughwatches

[Endeavor]

2 hours ago, VWatchie said:

The end-shake has been adjusted

If it was okay before, I always go by; "what has changed" ?

[nickelsilver]

If it's clean and you have endshake, it could be balanced enough that it needs a nudge to move, like tapping the movement holder on the bench. If it really is free, then the smallest puff from a blower should move it.

[VWatchie]

4 hours ago, nickelsilver said:

If it really is free, then the smallest puff from a blower should move it.

Ah, that I've never tried. Amazingly useful tip as always! 

5 hours ago, Endeavor said:

If it was okay before, I always go by; "what has changed" ?

That seems like a very good rule of thumb. However, in this case, the movement has never run well. Not before service and not after.

[JohnR725]

8 hours ago, VWatchie said:

My only thought is that perhaps the pallet fork of these ETA 28XX movements is constructed in such a way that the weight on both sides of the pallet fork staff is equal.

I do remember somebody saying once about how pallet forks should move effortlessly exactly as you're trying to test for. But I've also seen examples of poised pallet forks. then pictures I had to go to the Internet to find a picture of your pallet fork. The arm not sure what part it's called that goes towards the balance wheel doesn't look very long. Also it's a really tiny pallet fork it may just not have enough mass to move whether it's poised are not.

43 minutes ago, VWatchie said:

That seems like a very good rule of thumb. However, in this case, the movement has never run well. Not before service and not after.

so basically what you're telling us is were searching for the original problem that was never found. Always nice when looking for a problem to understand what it is were looking for. this is why typically ask people who just service the watch that's not running I ask the question what was it doing before. It helps isolate whether the problem was caused by servicing or not.

with the pallet fork out if you slowly started to wind the watch as the gear train start to spin effortlessly? then when you stop turning do you get back spin?

[Neverenoughwatches]

1 hour ago, VWatchie said:

Ah, that I've never tried. Amazingly useful tip as always! 

That seems like a very good rule of thumb. However, in this case, the movement has never run well. Not before service and not after.

With the answer of the pallet being evenly balanced by you and a few others earlier. If that were so then consider when the movement holder is tilted then the plate and its jewel is moving about the lever's pivot as it stays balanced. Then the lever will still touch from one banking pin to the other if it's pivot were completely free within the jewel. Which it isn't exactly as the pivot still has some friction as it rests lengthways within the jewel. You would think the weight of the lever would be enough to break that resting friction. 

[Neverenoughwatches]

19 hours ago, RichardHarris123 said:

If evenly ballanced, the fork should stay still if the plate is rotated.

if the lever stays at its balanced or poised point while the plate moves around it. That would show that the pivot is free within the jewel. If the lever is rotating along with the plate then there is obviously some friction between the jewel and the pivot. 

[nickelsilver]

There's always friction. It a matter of how much.

[Neverenoughwatches]

8 minutes ago, nickelsilver said:

There's always friction. It a matter of how much.

the pivot just resting in the jewel has friction. Nothing is entirely free floating, unless the watch is in space.

[VWatchie]

14 hours ago, JohnR725 said:

I do remember somebody saying once about how pallet forks should move effortlessly exactly as you're trying to test for. But I've also seen examples of poised pallet forks. then pictures I had to go to the Internet to find a picture of your pallet fork. The arm not sure what part it's called that goes towards the balance wheel doesn't look very long. Also it's a really tiny pallet fork it may just not have enough mass to move whether it's poised are not.

It was very interesting to read especially as I had never heard of it before and it is in line with what I suspected.

14 hours ago, JohnR725 said:

with the pallet fork out if you slowly started to wind the watch as the gear train start to spin effortlessly? then when you stop turning do you get back spin?

I haven't gotten that far yet, but I've made it a habit to check for backspin. Thanks for reminding me though. It's a very good and important check.