How do one poise Atmos 519 pallet fork ?

[watchknight]

I am a watchmaker by trade, recently started to be exposed to more clocks as it will be another good source of income. I have done limited atmos, only 1 or 2 non 519 series. Just today, I disassembled the 519 atmos which has broken chain, mangled spring that is hooked outside the barrel. A broken jewel of the second wheel ? ( the biggest wheel in the train movement ) and a scratched pivot of the wheel. I was following the step by step protocol and came to the point of poising the pallet fork. The 519 is quite different from the other series as it is a flat piece of fork at the end. I was reading other manual as they described using a round hammer to gently tap the opposite side of the flat piece you want the pallet fork to bend to. But I might have misunderstood something or did something wrong as the pallet fork is still unpoised after some gentle tap. From the picture, it can be seen the pallet fork is tilting to the right side. How should be poised ?

 

[Watchfixer]

Fantastic pictures! How did you take these pictures like these?

Cheers, Watchfixer

 

[Mike Murray]

Truthfully, that small amount it is out of poise will have NO ill effect. For this "apple" type of fork, the poising is done by gently manipulating the "throat" of the fork. By throat I mean the long "neck" of the fork.
Hitting is not something I've ever had to do and I did read that somewhere but I dismissed it as a myth or at the very least something that may cause an irreversible error.

Mike

 

[Watchfixer]

NO! Not the long part of fork. That is critical relationship between escapment wheel and the lever fork. Leave that alone. There are a reason this lever is designed this way. Remember the inventor who did that first atmos clocks long ago struggled for long time to get this working reliably with low power mainspring that runs several *months* on one wind without the bellows. Bellows winds very tiny motion to move one or two clicks at a time over temp and atmospheric pressure changes.

Cheers, Watchfixer

 

[watchknight]

NO! Not the long part of fork. That is critical relationship between escapment wheel and the lever fork. Leave that alone. There are a reason this lever is designed this way. Remember the inventor who did that first atmos clocks long ago struggled for long time to get this working reliably with low power mainspring that runs several *months* on one wind without the bellows. Bellows winds very tiny motion to move one or two clicks at a time over temp and atmospheric pressure changes.

Cheers, Watchfixer

Hi, the pictures are taken using a 20x loupe on a macro len camera. How would you poise the error ?

 

[Watchfixer]

There are some few who specialize on atmos clocks and you could question about this lever but I don't know where to get reference except one website.

Cheers, Watchfixer

 

[John Hubby]

I have to agree with Mike Murray, that the amount your fork is presently out of poise will not affect the operation of the clock, provided it is in beat. Having said that, however, it may be parallax in the photo you posted of the end view of the fork, but the long end of the fork appears to be bent slightly to the left which would explain why the fork is tilted to the left.

I have drawn a red line through the centerline of the long end of the fork and as can be seen it shows that part to be bent slightly to the left. Bending it slightly to the right (as viewed in this photo) will probably bring the fork very close to being poised.

Any time you manipulate that part of the fork you will slightly change the beat of the escapement. It may be that "as is" the fork is out of beat but the only way to know for certain is to physically check it with the movement assembled and mounted in the clock, with some power applied to the train. You can manually check the beat with an amplifier, to see where each pallet "ticks" as the fork is moved from one side to the other of center position. The ticks must be equidistant from the center for the clock for it to be in beat.

I would check the beat first, then worry about the poise. If it is in beat now, no need to do anything else. If not, follow the above to improve the poise (assuming the beat error tells you the same story) and recheck the beat after adjustment.

 

[leftarmdwn]

I agree it looks just fine.If you want to be a perfectionist add a drop of epoxy glue to one side and see if it poises and be prepared to add or remove some
Errol

 

[watchknight]

I have to agree with Mike Murray, that the amount your fork is presently out of poise will not affect the operation of the clock, provided it is in beat. Having said that, however, it may be parallax in the photo you posted of the end view of the fork, but the long end of the fork appears to be bent slightly to the left which would explain why the fork is tilted to the left.

I have drawn a red line through the centerline of the long end of the fork and as can be seen it shows that part to be bent slightly to the left. Bending it slightly to the right (as viewed in this photo) will probably bring the fork very close to being poised.

Any time you manipulate that part of the fork you will slightly change the beat of the escapement. It may be that "as is" the fork is out of beat but the only way to know for certain is to physically check it with the movement assembled and mounted in the clock, with some power applied to the train. You can manually check the beat with an amplifier, to see where each pallet "ticks" as the fork is moved from one side to the other of center position. The ticks must be equidistant from the center for the clock for it to be in beat.

I would check the beat first, then worry about the poise. If it is in beat now, no need to do anything else. If not, follow the above to improve the poise (assuming the beat error tells you the same story) and recheck the beat after adjustment.

I see, I will put the assembled and check if it is in beat first before doing anything. Thanks for the help guys.

 

[watchknight]

I have cleaned up the clock, and let it run from 1 2/3 revolution. After roughly half a day later, the balance rotate a constant revolution of 1 complete turn. I took a look at the escapement. And noticed the left side of the pallet has more drop, more run to banking and more total than the right side of the pallet. The right side of the pallet has no run to banking at all. The roller is in beat, that is when left to rest it goes to the center. Another thing I noticed is the pallet fork didn't stop at the same distance on both side. The pallet fork seemed to be much closer to the Bigger side of the banking pin. I have thought of bending the long arm of the pallet fork at the direction towards the right side to increase the run to banking on the right side. But thought I should seek some advise before doing anything. Below are the pictures, advise are greatly appreciated.

Left pallet

Right pallet

Pallet fork seemed to be closer to left bigger banking pin

 

[Watchfixer]

Leave it alone till get a service manual to set this up properly. There are too many issues.

1. The shaft of the balance is off-center (very noticeable above the pallet fork).
2. Banking pins are not in contact or is it on purpose? (that is a true question).
3. Do not worry about the pallet lever yet. Take care of the #1 and find out the question about #2.

BTW, good questions and good pictures! Once this is taken care of, clean and lube but there is specific areas that is not lubricated.
There should be a website about atmos and contact the guy about this issues.


Cheers, Watchfixer

 

[etmb61]

2. Banking pins are not in contact or is it on purpose? (that is a true question).
Watchfixer

There are also banking pins at the pallets. The two at the fork are redundant.

http://www.abbeyclock.com/cal/atmos122s.html

Eric