Watch loses minutes per day; timegrapher reads fine

[cortman]

Hi again forum!

I'm still working on the 16s Elgin Grade 211 Model 7 pocket watch. I formerly posted about this watch when I replaced the hairspring and it ran unnaturally fast while reading slow on a timegraph. I have since replaced the balance complete with a used but working balance from another movement.

I've run into a new conundrum: with the new balance the amplitude has improved to around 190-200- still low but as good as I'm expecting for this old watch. I've managed to regulate it to a fairly consistent +12-20 spd, with about 1.5 ms beat error. I've observed the watch running on the timegrapher for up to 15-20 minutes at a time with no significant deviation.

The problem is that the hands read slow- hours slow per day. No problem, I've fixed loose cannon pinion problems before. I tightened it up and it feels great when setting (it definitely was tell-tale loose before). However the watch continues to lose about two minutes per hour from observing it. The hands don't appear to get out of sync with each other, and it continues to read well on the timegrapher. The cannon pinion is tight- I re-tightened it to where it almost became too difficult to set, just as an experiment, and it continued to be slow. The hand itself doesn't appear to be loose on the pinion. The hour wheel and minute wheel both appear to be in excellent condition.

What else could be causing this issue? I'm really at a loss.  Thanks for any and all help.

[Woolshire]

Looks like you’ve addressed everything I would have thought of, except for possible wrong minute/hour wheels? Might want to count teeth and compare to known examples of Elgin model 7.

[JohnR725]

5 hours ago, cortman said:

15-20 minutes at a time with no significant deviation.

I don't suppose we can have a picture of theTiming machine results?

Out of curiosity are you running the movement in the case or out of the watch case?

[Klassiker]

What bpm is the new balance running at? What does a tooth-count tell you is the correct frequency?

[spectre6000]

Canon pinion was first thought, so you got that. Tooth count was my second thought; Woolshire got that one. How is the pallet fork engagement? What are the odds it's occasionally jumping teeth? How is the engagement and/or shape of other teeth? Maybe a wheel is skipping a tooth? 

[cortman]

Thank you very much for the replies!

10 hours ago, Woolshire said:

Looks like you’ve addressed everything I would have thought of, except for possible wrong minute/hour wheels? Might want to count teeth and compare to known examples of Elgin model 7.

I will pull the bridge off this morning and count the teeth. I have a parts movement of the same type available, so I will compare to it.

 

9 hours ago, JohnR725 said:

I don't suppose we can have a picture of theTiming machine results?

Out of curiosity are you running the movement in the case or out of the watch case?

I've attached a reading of the timing- I've watched it run both in and out of the case.

 

9 hours ago, Klassiker said:

What bpm is the new balance running at? What does a tooth-count tell you is the correct frequency?

The timegraph is saying 18000, which according to all documentation I can find online is the correct speed.

 

47 minutes ago, spectre6000 said:

Canon pinion was first thought, so you got that. Tooth count was my second thought; Woolshire got that one. How is the pallet fork engagement? What are the odds it's occasionally jumping teeth? How is the engagement and/or shape of other teeth? Maybe a wheel is skipping a tooth? 

That's an interesting thought. I didn't notice any damage to the escape wheel. The impulse pin on the pallet fork looked bent (laterally with the shaft of the fork) slightly, but I didn't know if it was intended to be that way. I'll compare to the parts movement pallet fork.

 

I'll check the teeth and the pallet fork and return back. Thanks again.

[cortman]

Well I have some news! Thanks to all you who recommended counting teeth; coming from servicing pretty standard vintage wristwatches I would have never thought to bother with it, but sure enough one wheel has 5 more teeth than the comparable one from my parts movement. So I've swapped them out and am now timing it to see how it works. Thanks a lot you all, I really appreciate the help!

[JohnR725]

8 hours ago, cortman said:

I've watched it run both in and out of the case.

That's an interesting pattern on the timing machine?  have you demagnetized the watch? Patterns are always interesting on the timing machine this could be because it's magnetized or power transmission problem can do the same thing

5 hours ago, cortman said:

but sure enough one wheel has 5 more teeth than the comparable one from my parts movement. So I've swapped them out and am now timing it to see how it works.

Which wheel did you swap out and how many teeth did the bad one Have versus the good? If you have the right parts book like I have it actually lists all the gears how many teeth and I'm curious about what's going on here. So I really frown on when people start swapping parts on vintage watches as it doesn't always end well. Usually because wheels are specifically fit or made in batches and batches can vary a little bit.

This problem actually comes up a lot with wristwatches like Rolex. This is where you have a series of watches that have the base parts in the base parts have a frequency of 18,000 beats per hour but the other ones are different frequency. So eager people who want to sell a watch will mix-and-match components and then just sell the watch is running but not necessarily specify it doesn't keep time.

9 hours ago, cortman said:

I've watched it run both in and out of the case.

The reason I ask is that American pocket watches do not typically have stems. They rely on the case for the stem aspect so If the movement is out of the case it's in setting mode. Typically the watches have a way of putting the watch into a service mode which puts it back in the winding but that would assume that you understand that know how to do that. So yes the watch will run out of the case in setting mode but driving the hands will be an issue and could be a reason a watches running slow because you're driving all of this setting parts gear train that you shouldn't be.