I've ran out of ideas & tricks ...... Help !!

[Endeavor]

@rodabod Actually I do ....... although I never used it. How does one go about measuring the "exact" balance amplitude.

JohnR725 suggested an UV-marker and an UV-light ..... but this new (technology) idea may open up other horizons !?

I'm all ear ..... !!

[rodabod]

6 minutes ago, Endeavor said:

@rodabod Actually I do ....... although I never used it. How does one go about measuring the "exact" balance amplitude.

 

Get lots of light on the movement and take a slo-mo video. You may need to shoot through a  loupe to get a good focus. Then play back the video and watch how many degress the balance swings back and forth. You can follow once balance wheel arm as a visual cue.

[Endeavor]

@rodabod: I think that was a brilliant idea !! I stuck a 2.8x eye-loupe with Rodeco to my iPhone and the picture is stunning .

Going in real slow motion and one can see that the balance wheel is making about 180 degrees. A clear view of the escape wheel & pallet fork. Now I need to find I way to upload a part of the video. One minute is already 49mb. I have a Macbook, anybody good at editing video's and which program to use?

 

Edited September 17, 2018 by Endeavor

[Endeavor]

Here a section of the edited video. I slowed it even more down than the slo-mo of the iPhone.

 

 

Edited September 17, 2018 by Endeavor

[Endeavor]

On the iPhone video (edited with iMovie) it seems that the amplitude is slightly over 180 degrees. If one knows the amplitude, one can, with the aid of a timegrapher, determine the lift angel (again, thanks JohnR725 ):

I took a (horrible looking) raw-data graph of the movement Dial Down (as the position was in the iPhone video), and with the timegrapher lifting angle set at 56 degrees, the graph showed indeed slightly above 180 degrees. We can therefor with a reasonable certainty say that this movement has a lifting angle of 56 degrees.

Edited September 17, 2018 by Endeavor

[Endeavor]

Meanwhile I serviced the 501 movement, which suffered a bit the same fate as the 503 movement. I stripped it for the 3rd time, cleaned, re-pegged all pivot holes and pivots. This time I brushed with a fiber brush the roller impact jewel, the pallet fork jewels, the horn and dart. I straightened the main-spring and oiled & grease with the thinnest oil & grease I had (Dr.Tillwich 1-3 oil & B-52 grease for the arbor)

With only the balance wheel and pallet fork mounted on the main-plate, at rest the pallet fork sits right in the center of the banking pins, i.e. there should be no beat error. There can be no adjustment made to improve; it's spot on !

Demagnetized each part again. Inspected everything with a 10x eye-loupe before assembling. The wheel train turned by the slightest touch of the ratchet wheel. I could not see any back-spin, perhaps ones or twice an escape wheel teeth moved very slightly backwards, but that was hard to tell for 100% sure. Check the pallet jewels for lock and draw, that seemed fine for all escape wheel teeth. Installed balance and wound spring to full. After the compulsory waiting time I filmed with my iPhone in slow-mo the balance amplitude, it reached barely 180 degrees. Due to the size of the file, I'm not going to upload that video.

Put the movement on the timegrapher, adjusted the daily rate slightly down. Here are the Dial Down graphs:

For some reason the header disappeared, but the readings are the same as in the graph above.

Raw data Dial Down

The 1 Second sweep DD:

It seems that the distances between the pulses (tick & tock) are not equal. This could be due to the beat error, but in theory there shouldn't be any since statically the balance wheel / pallet fork lines up exactly in between the banking pins. Also one pulse consist of 4x peaks, whereas the other pulse seems more consolidated to 3 to 4 peaks. This can also be seen in the 0.2 second sweeps below.

5 beats per second = 1 beat per 0.2 seconds. Here are 0.2 second sweeps of one tick and the graph below the tock;

Interesting to see are the distances (time) between the  sound peaks; the total duration of the tick seems longer than the tock.

If there is indeed something wrong with the escape / pallet fork jewels, I guess with my limited amount of tools, I'm getting pretty much to the end of the road with these movements. Apart from perhaps a new main-spring, I have followed all advises, tips & tricks.

I'll give the 503 movement another full-blown treatment and will oil it with the lightest oils & greases. If the beat error and low amplitude don't improve either, I'll call it the day (with these movements that is ).

This won't be a shame; these movements have given me already a tremendous steep learning curve ...... each and everything has its purpose, perhaps that was all these movement were meant to do ?

I'll report back on the 503 movement ....

 

Edited September 19, 2018 by Endeavor

[mikepilk]

Are you sure the spring is Ok - i.e. it still has enough 'springiness' ?

I ended up stripping a Smiths movement 4 times to try to determine why I couldn't get the amplitude above 180 deg.

In desperation I took out the 'new' spring I'd fitted only to discover it was totally useless - it bent like a bit of tin and didn't and didn't spring back.

[Endeavor]

@mikepilk: thanks There were 3x more or less identical movements, same age, same N.O.S, same springs; all behaved the same. I have no doubt that the springs have lost some of their "springness". However, they are still willing to get out of the barrel once you start pealing the center out. I straightened the tail ends, that helped .... so they do give some "resistance". There is also a "reasonable" amount of tension if you wind the barrel. New springs will for sure improve the amplitude but at the moment I'm not convinced whether these movements are worth the investment? The huge beat error (where there shouldn't be any !), the wave in the 503 wheel train, the erratic pulses on one side of the 503 pallet fork jewel, the daily rate changes a lot in different positions, etc etc. If they were running smooth, but just low amplitude, no doubt I would order some new springs ..... at the moment I'm not so sure .......

Anyway, anybody who is still hanging in there, still following this thread and try to help, should be promoted to "Distinguished WRT member" 

Edited September 19, 2018 by Endeavor

[mikepilk]

6 minutes ago, Endeavor said:

@mikepilk: thanks There were 3x more or less identical movements, same age, same N.O.S, same springs; all behaved the same. I have no doubt that the springs have lost some of their "springness". However, they are still willing to get out of the barrel once you start pealing the center out. I straightened the tail ends, that helped .... so they do give some "resistance". There is also a "reasonable" amount of tension if you wind the barrel. New springs will for sure improve the amplitude but at the moment I'm not convinced whether these movements are worth the investment? The huge beat error (where there shouldn't be any !), the wave in the 503 wheel train, the erratic pulses on one side of the 503 pallet fork jewel, the daily rate changes a lot in different positions, etc etc. If they were running smooth, but just low amplitude, no doubt I would order some new springs ..... at the moment I'm not so sure .......

Anyway, anybody who is still hanging in there, still following this thread and try to help, should be promoted to "Distinguished WRT member" 

It's been a good learning experience, for you, and us watching. 

I never try to give up on these difficult movements, because you can learn so much about what could be wrong.

And I've still lots to learn.

....... but there's always a point when you've had enough ! 

Good luck

[Endeavor]

44 minutes ago, mikepilk said:

...... but there's always a point when you've had enough ! 

Yes, I'm slowly getting there; "worn-out" & exhausted are perhaps additional descriptions

But indeed, the learning curve has already been tremendous. I know now much more the in-depth workings of an escape wheel, the interaction with the pallet fork jewels; lock, unlock, draw and the sounds they create. The pallet fork, horn & dart. Also I went much further & deeper then ever before with the W.O.S timegrapher software, the raw data and trying to understand the interpretations / implications, the sweep graphs etc. How to determine the lift-angle with a timegrapher and a whole lot more. Of course, next to everybody who contributed with advice, JohnR725, the timergrapher "guru" helped and taught me a tremendous amount

Anyway, "we" haven't fully given up........ Let's wait for the final 503 results and then draw a conclusion ......

Edited September 19, 2018 by Endeavor

[JohnR725]

a couple of minor corrections and then a suggestion.

If you can bend the mainspring the way you are bending it, it's soft. So the instant I goes back in the barrel it's no longer in the shape that you just form that in. but we still should be getting better results when it's fully wound up and if the watches running 24 hours later it seems like the mainspring still has enough power we still should be seeing something better?

then minor correction on beat? Yes the pallet fork is supposed to be between the banking pins but the roller jewel needs to be centered in the slot on the pallet fork. So depending upon the size of the watch and how much play there is the roller jewel can be on either side of the slot of the fork and the fork still be centered but you're not in beat.

I run into this quite a bit with the American pocket watches I typically work on. so visually everything looks perfect timing machine disagrees. Then throw in the other complication on American pocket watches movable banking pins that seldom are where they are supposed to be. so I find on pocket watches at least a little more time and effort spent is required to put the watch in beat each time visually the fork will look perfect but the timing machine will disagree until the roller jewel is reasonably closely centered. But on a watch of this size being off a little bit is not going to be 9 ms.

Then don't consider that you're giving up your taking a break which could be hours days weeks conceivably even years. breaks are good a lot of times you just can't see what you need to see. Other times you need new skills or new comprehensions or just something different. At some point in time when you come back you might suddenly notice some glaring mistake that you just not seeing or you'll have new skill set to grasp something. Or maybe we've already looked at everything and the reason they were new old stock is they never left the factory as they have an issue.

[Endeavor]

@JohnR725: Thanks again

About the mainspring: yes, they are (too) soft. Spring steel is hardened and shouldn't take on (so easily) a different shape. However, as you said, they all powered the watch for 34+ hrs. When I take them out of the barrel, after I had straighten the tail, the retained a part of the new shape, less straight as they went in, more "open" then original.

Beat error; With the balance wheel on its own mounted on the main plate, when in rest, the roller impulse jewel is right in between the banking pins. I just repeated the same exercise, this time I included the pallet fork, using the fork as a kind of (visual) extension to the banking-pins. With all in rest, fork is right in between the banking pins and the roller jewel floats right in between the horns.

This brings me to the following thought; In theory there should be no beat error. Assume that one pallet jewel, or the interaction between of one pallet-jewel /all escape wheel teeth is not as it supposed to be, could this not result in a uneven "trust" of the balance wheel? Meaning, to take for explanation purposes the names tick & tock, that the tick drives the balance wheel further away than the tock. This would explain the apparent beat-error. This would explain why between a sets of tick&tocks the time in between looks longer than the other set. I included a short video of my computer screen, you may have to play it a few times to see it. One can take a sound spike in the middle and then observe that each time the sounds flips to the left or to the right of that "middle" spike. Clearly the "beat-error", where, in my opinion mechanically speaking, there shouldn't be any. It look to me that the beat error is introduced rather than a mechanical beat error.

If the "drive" from the escape wheel to one pallet jewel is not optimal, then next to the beat-error, the balance wheel doens't get the optimal trust either, perhaps resulting in the (persistent) low amplitude?

As said before, if there is a problem with one of the jewels, the draw, the locking depth or whatever, with my current tool and current knowledge, I'm pretty much to the end of the road.

I fully agree with your latest idea to take a brake. Idea's and thoughts do mature, so does the "operator". Perhaps one day comes, as the Germans say; "das Aha erlebnis"

We'll see whether I can make some improvements to the 503, but that these watches are going to be a present is very doubtful.

I'll report back ......

 

 

Edited September 20, 2018 by Endeavor

[nickelsilver]

If the beat error is that high with the fork centered at rest it's fairly certain the escapement is very poorly adjusted.

[Endeavor]

@nickelsilver: Yes, that could certainly be one conclusion ....... any suggestions for a cure?

Edited September 20, 2018 by Endeavor

[nickelsilver]

13 minutes ago, Endeavor said:

@nickelsilver: Yes, that could certainly be one conclusion ....... any suggestions for a cure?

Ha, yes, set up the escapement correctly!

There's lots of info above, but to distill it and give you my procedure here goes.

You'll want the watch assembled with power on the train, with little or no oil on the escapement. A microscope, ideally with bottom illumination really helps with the checks. With a finger, rotate the balance until you unlock one side of the escapement, keep going slowly until it drops on the other stone. It should land on the locking face, not the impulse face. At this point, with an oiler, check that there is some freedom at the fork (you're checking if the roller jewel is binding at all, so very small but perceptible freedom). Continue rotating the balance, and continue checking the for freedom to roller, the fork horns are first, contacting the roller jewel, then the guard pin to the safety roller. The guard pin clearance is normally slightly less than fork horn clearance. Do this on both stones, and if you want to do it right, on all the escape teeth. There must be freedom, AND-

the freedom shouldn't be so large as to let a tooth progress onto the impulse face of a pallet stone. The fork should snap back to its banking.

It's unlikely to have an issue with fork horns but bent guard pins do crop up.

Now check your locks. Same as before, rotate balance, observe how much drop lock you have on each stone. It should be a bit less than a third of the width of a pallet stone as a rule of thumb, and approximately equal on both sides. Now check that you have run to the banking. This is your total lock, and there must be at least a little run to the banking. The total lock will be about a third of a jewel width.

 

When you get used to doing the checks you'll check locks and fork horn/guard pin clearance run to the banking all sort of at the same time. But at first do them in order.

 

A cheat for getting the pallet stones in the right position, and it depends on the horns and guard pin being correct, is this: Checking fork horn clearance, with the horn against the roller, observe how much lock you still have. Try pushing in the stone until it is as small as possible while still remaining on the locking face. Both sides. Check drop lock, if it is too small (landing on the impulse face either side), pull both stones out a tiny bit. Like 0.01mm. Very tiny. Check again. Continue until you have safe drop lock. This will be as far in as the stones can be.

 

From your charts above it looks like something is touching, in the escapement, and my guess is the guard pin.

[Endeavor]

@nickelsilver: Thank you very much for your helpful & useful answer Just to be sure of the names and to avoid confusion, I've been using "the dart", which you call the guard or guard pin, those are the same I'll take it?

Just for my own education, what is the exact function of the guard or guard-pin? Knowing the exact function helps tremendously where to look for....... So far I haven't noticed anything unusual to the horns nor the dart, but did I know where to look for .....?? A microscope is not at hand, but I was thinking about getting one of those USB microscopes like Tmuir has, or a cheaper version thereof: https://www.watchrepairtalk.com/topic/9760-check-check-and-then-check-again/?tab=comments#comment-89315

These kind of in-depth explanations do give me new knowledge and "tools" to check the escapement further out. Making those tiny adjustments, if required,  to the pallet fork jewels with simple hand tools, is going to be another exciting, jet very challenging chapter.

Anyway, it seems that we haven't come to the end of the road with these movements. Before I dive deeper into the escapement rabbit-hole, I may need a few days off; to refresh, regroup, reorganize and renew my P.O.A. (Plan Of Attack).

Just when things started to look bleak, you joined the group & discussion .... talk about timing !

Thanks !

Edited September 20, 2018 by Endeavor

[nickelsilver]

Yes guard pin and dart are the same thing. The guard pin interacts with the safety roller (or, in the illustration above simply the roller, that's an old escapement), and the fork horns with the roller jewel. Their function is to prevent the escapement from unlocking at any time when the roller jewel is not in the fork slot, it's a safety action.

[Endeavor]

Thank you for your additional info ....... I guess I have to look a bit deeper into that and try to find some more study material about it. At the moment I don't grasp the full extent / depth / workings of your explanation ......

If you have some study material, or a link in which they explain the working(s), that would be highly appreciated .....

Edited September 20, 2018 by Endeavor

[vinn3]

20 hours ago, mikepilk said:

Are you sure the spring is Ok - i.e. it still has enough 'springiness' ?

I ended up stripping a Smiths movement 4 times to try to determine why I couldn't get the amplitude above 180 deg.

In desperation I took out the 'new' spring I'd fitted only to discover it was totally useless - it bent like a bit of tin and didn't and didn't spring back.

    i think some one amealed the entire spring.  only the tip needs anealing.     repairing a broken main spring is not easy.  vin

[Endeavor]

@vinn3; Thanks vin. With all respect, but best is to open a new thread with @mikepilk, unless it was mend for me?? Then I'm sorry that I didn't get it !

To come back on the 503 movement, which I now cleaned, pegged for the x-times. Different this time is that I paid additional attention and fiber brushed the escape wheel teeth, the pallet jewels, horns, dart or guard an roller impact jewel. I also lubricated everything with the thinnest oil and grease. The amplitude in the slow-mo iPhone video showed an amplitude of over 180 degrees, I guesstimated 200 or 210. The timegrapher confirmed that. Here are the graph and I suspect also with this movement that one pallet jewel isn't doing it's job, or as Nickelsilver suggested, something is up with the guard pin. There is also, in between the tick & tock, another spike which is there most of the times, but not all the times ?? Hence the threshold is set high.

Dail Down;

1 Second sweep:

0.4 second sweep, showing the tick, tock and the mysterious "piep" :

As said above, I need time to study Nickelsilvers new information and suggestions. These movements are not going to be a present anymore and will become future study material.

When ready, I probably will open another thread in the likes of "Adjusting a troublesome escapement". How far I'll come with that, needs to be seen. But interesting & educative it will be !!

For now I would like to take the opportunity to thank you all for your time, all your help, tips & advises. For sure, for me it has been a huge learning curve; going down the escapement rabbit hole, the software- and the mechanical-side of it.

Hopefully you will join me on my next "endeavor" !? I guess for the ones who made it this far, are interested in finding out how the story ends. Regardless how it does, it was & is a tremendous experience !

Thank you all for your very welcome help !!

 

 

Edited September 20, 2018 by Endeavor

[Geo]

What a cracking thread, and all credit for sticking with it to get the thing right.  No wonder you post under the name "Endeavour".  The input and help you are recieving is excellent.  I'm looking forward to the next instalment!

[Nucejoe]

Hi there, 

Guard pin dose not function, it's job is go through the hollow in the roller so to esure the impulse jewel recieves energy at the right position," if by sudden motion of the movement" the wheel picks up speed and impulse jewel get ahead of escapement, the pallets may jump or lock.

The test for the guard pin not hitting anything is to paint the impulse jewel all over( especially the tip with a magic marker) instal  balance let run for some time, uninstal the balance check for any sign of the impulse jewel  tip hitting the guard pin. 

Regards

[Endeavor]

@NucejoeDas "Aha Erlebnis"

Thanks

[Nucejoe]

I allways fix the faults before cleaning.

All cleaning dose, can be acheived by dropping a drop of lighter fluid on jewels and let train run then install balance & fork drop make it run by blowing on the balance.

[rodabod]

@Endeavor I discussed lever escapement safety action and testing a little bit here: