Positional error

[BrianKehoe]

I have been working on an 8 day travel clock/watch, 16 size, swiss, with a seven jewel lever escapement, non-jewelled train. Fully disassembled, cleaned reassembled, lubricated. No notable wear or issues found.
The timegrapher machine shows a 4 min difference positioanally. Pendant down(normal running position) is 4mins slower than DD (fastest running position) , 3:45 mins slower than DU and 2 mins slower tha PL and PR. After demagnetizing,
I checked the poise of the balance wheel, in jaws, it is basically perfect. Pivots not bent. Balance wheel runs nicely free in the movement with hairspring off and the lever out, (slightly less free in positions other than DD and DU, as one would expect). The hairspring is not interfering with anything in any position. Of interest, on the timegraph, the entry and exit signal run straight/parallel in all positions other than the slow one PD, so I suspected sideshake at the lever, balance or escape wheel was creating the problem, with a too deep or inconsistent lock, draw and release. I removed all the escape components again, read through DeCarle's description of lever/escapewheel action and checking lock and draw, found no issue. I did note the banking pins are not perfectly vertical as he describes they should be, but the endshake of the lever is minimal so I bypassed that, also the position of the lecer when locked is equidistant from the balance hole, both sides. I then removed the balance endstones again and checked all the jewels, balance, lever and roller, for cracks, chips and looseness. Beat error when assembled and tested has been variable, but even when it is as low as .4ms the error is the same as when it is up around 2ms. I installed the hairspring in the cock alone and adjusted it to center. It was off a bit, but not dramatically. The next thing I did was burnish the balance pivots regardless that there was no VISIBLE issue with them, more to see them turn under the microscope than anything else. The pivot ends were a bit flatter than I am used to seeing so I rounded them a little as well. I read through the other related poising and position error posts, where I learned to test if the error reverses at low amplitude, it does not, which supports my assessment it is not a balance poise problem. I am hoping someone here will have some helpful advice, as I am facing returning the piece to my client in this unsolved condition.

 

[Chris Radek]

Hi, welcome to the forum! First, excellent job stopping and thinking, instead of madly filing away on your balance. It's a situation like yours that can really trick a botchmaker into making things worse.

I think you have already seen the problem, which is being shown by your bad traces in the problem position. What you don't know yet is the solution. What does the bad trace look like? Is one line still ok but the other one is variable? Could you take a picture of it and share it here?

This is a guess, but I wonder if there is insufficient draw to reliably pull the lever up against the top banking pin (against gravity) when in the bad position. Did you use the right amount of 9415 on the pallets? If there's adequate power on the train (amplitude in the good positions is fine) and the pallets are lubed properly, I think you're right to suspect there's a geometry or wear problem in the escapement. Take the balance out, put the movement in the problem position, and see if there's definite draw to both banking pins (push the lever slightly away from each pin and see that it smartly moves right back). If it doesn't, it will cause all sorts of inconsistency.

Here's an example of something you might not expect, and might miss at first, that I found causing unreliable/insufficient draw:

 

[BrianKehoe]

Thank you Roughbarked. I did not test the draw in the fault position. Great microscopic pic. The line on the machine looks random, no cycling pattern, like when you have a single bad tooth or a not-round wheel.

Attachments

• 0 bytes · Views: 0

 

[BrianKehoe]

Hi, welcome to the forum! First, excellent job stopping and thinking, instead of madly filing away on your balance. It's a situation like yours that can really trick a botchmaker into making things worse.

I think you have already seen the problem, which is being shown by your bad traces in the problem position. What you don't know yet is the solution. What does the bad trace look like? Is one line still ok but the other one is variable? Could you take a picture of it and share it here?

This is a guess, but I wonder if there is insufficient draw to reliably pull the lever up against the top banking pin (against gravity) when in the bad position. Did you use the right amount of 9415 on the pallets? If there's adequate power on the train (amplitude in the good positions is fine) and the pallets are lubed properly, I think you're right to suspect there's a geometry or wear problem in the escapement. Take the balance out, put the movement in the problem position, and see if there's definite draw to both banking pins (push the lever slightly away from each pin and see that it smartly moves right back). If it doesn't, it will cause all sorts of inconsistency.

Here's an example of something you might not expect, and might miss at first, that I found causing unreliable/insufficient draw:

View attachment 755718

Sorry, thank you Chris.

 

[Skutt50]

What figures do you have for the swing in various positions?

 

[BrianKehoe]

What figures do you have for the swing in various positions?

260 DD AND DU
220 PD AND PU~

 

[Skutt50]

OK that is consistent enough but could be a bit higher..........

 

[roughbarked]

Thank you Roughbarked. I did not test the draw in the fault position. Great microscopic pic. The line on the machine looks random, no cycling pattern, like when you have a single bad tooth or a not-round wheel.

Attachments

• 0 bytes · Views: 0

The photo and comment were "Chris Radek"

 

[watchwinder]

I have been working on an 8 day travel clock/watch, 16 size, swiss, with a seven jewel lever escapement, non-jewelled train. Fully disassembled, cleaned reassembled, lubricated. No notable wear or issues found.
The timegrapher machine shows a 4 min difference positioanally. Pendant down(normal running position) is 4mins slower than DD (fastest running position) , 3:45 mins slower than DU and 2 mins slower tha PL and PR. After demagnetizing,
I checked the poise of the balance wheel, in jaws, it is basically perfect. Pivots not bent. Balance wheel runs nicely free in the movement with hairspring off and the lever out, (slightly less free in positions other than DD and DU, as one would expect). The hairspring is not interfering with anything in any position. Of interest, on the timegraph, the entry and exit signal run straight/parallel in all positions other than the slow one PD, so I suspected sideshake at the lever, balance or escape wheel was creating the problem, with a too deep or inconsistent lock, draw and release. I removed all the escape components again, read through DeCarle's description of lever/escapewheel action and checking lock and draw, found no issue. I did note the banking pins are not perfectly vertical as he describes they should be, but the endshake of the lever is minimal so I bypassed that, also the position of the lecer when locked is equidistant from the balance hole, both sides. I then removed the balance endstones again and checked all the jewels, balance, lever and roller, for cracks, chips and looseness. Beat error when assembled and tested has been variable, but even when it is as low as .4ms the error is the same as when it is up around 2ms. I installed the hairspring in the cock alone and adjusted it to center. It was off a bit, but not dramatically. The next thing I did was burnish the balance pivots regardless that there was no VISIBLE issue with them, more to see them turn under the microscope than anything else. The pivot ends were a bit flatter than I am used to seeing so I rounded them a little as well. I read through the other related poising and position error posts, where I learned to test if the error reverses at low amplitude, it does not, which supports my assessment it is not a balance poise problem. I am hoping someone here will have some helpful advice, as I am facing returning the piece to my client in this unsolved condition.

With the watch running slower in the DD or DU i would look for problems with the pivots of the balance staff. Mushrooming , Bent pivot , Pivot to short so that the pivot shoulder may be rubbing on the jewel hole .
end stone rough , Pivot end rough

 

[BrianKehoe]

The photo and comment were "Chris Radek"

I saw that, and corrected it.
As in "sorry, thank you Chris"

 

[BrianKehoe]

Hi, welcome to the forum! First, excellent job stopping and thinking, instead of madly filing away on your balance. It's a situation like yours that can really trick a botchmaker into making things worse.

I think you have already seen the problem, which is being shown by your bad traces in the problem position. What you don't know yet is the solution. What does the bad trace look like? Is one line still ok but the other one is variable? Could you take a picture of it and share it here?

This is a guess, but I wonder if there is insufficient draw to reliably pull the lever up against the top banking pin (against gravity) when in the bad position. Did you use the right amount of 9415 on the pallets? If there's adequate power on the train (amplitude in the good positions is fine) and the pallets are lubed properly, I think you're right to suspect there's a geometry or wear problem in the escapement. Take the balance out, put the movement in the problem position, and see if there's definite draw to both banking pins (push the lever slightly away from each pin and see that it smartly moves right back). If it doesn't, it will cause all sorts of inconsistency.

Here's an example of something you might not expect, and might miss at first, that I found causing unreliable/insufficient draw:

View attachment 755718

Thank you Chris.
I intend to test the lock and draw in the fault position as you describe. I will respond here with results but it will be a few days away. The jewel wear in your photo would indeed be easy to miss and I can see how it would express a problem more dramatically in different positions. I Appreciate the insight.

 

[praezis]

the pivot ends were a bit flatter than I am used to seeing so I rounded them a little as well.

Not a good idea. I see too round pivot ends often. They will give nice amplitude values in the flat, but large difference between flat and vertical (much lower).
Your rate values would be more meaningful if you added the corresponding amplitude values. You really had consistant 220 degrees in all 4 vertical positions?

Frank

 

[penjunky]

I wonder if there is insufficient draw to reliably pull the lever up against the top banking pin (against gravity)

Hi Chris

I don't recall reading about that draw anywhere. Is there any literature on it available, can't find nothing scouring the net.

Thanks...Roger

 

[BrianKehoe]

First of all, thank you to the folks who have offered thoughts and ideas.
I tested the lock and draw of the lever as Chris Radek suggested. With one full turn of load on the mainspring, the draw is "smart" in the dial down position, but notably sluggish in the PD position, both at the entry pallet and the exit pallet, but more so at the entry side, where the fork needs to be "lifted" against gravity. When more power is added to the spring the effect decreases, but does not disappear.
Interestingly, the wandering signal line has disappeared. Also the amplitude has decreased in all positions. The watch was left to run in the fault position, and had stopped after a few 3+ days. There is an amplitude difference and a timing difference between fully wound and partly wound indicating to me a power problem. Remember, this is an 8 day movement=extra wheel, so perhaps the power loss due to a tired mainspring is exacerbating an issue at the pallet jewels.
This will be my approach, but I am posting pics of the timegrapher in positions DU, PD, PU and DD for interest and discussion.

 

[Skutt50]

Just to make sure, check the pallet fork. If a pivot has some damage, is bent or is dirty it may work fine in one position but jam in the other positions. If everything looks fine, don't oil the pivots and check if the pallet fork will move from one end position to the other if you tilt the movement.

 

[BrianKehoe]

Just to make sure, check the pallet fork. If a pivot has some damage, is bent or is dirty it may work fine in one position but jam in the other positions. If everything looks fine, don't oil the pivots and check if the pallet fork will move from one end position to the other if you tilt the movement.

Thanks Skutt50.
One of the 1st things I did was check the pivots and and sideshake, and endshake, at the balance, the fork and the escape wheel.
I like the idea of a freefall test though.
Will check it later.