What does this Timegrapher Trace Mean?

[JohnR725]

2 hours ago, Klassiker said:

Look closely at the order in which the dots appear on the screen. I suspect something like this:

One of the problems with a smaller LCD screen is rollover error for the lack of a better term. Basically with the LCD screen we only see a small window versus the numbers. We've already discussed that somewhere in the group related to the beat error. If you go over around 4 ms I don't remember the exact number the numbers will continue to rise but the graphical display wool over and might actually look perfect and the numbers will not agree. This is where the numbers always have to agree with the graphical display and of course visually that needs to agree with whatever you're seeing on the machine.

Then it's a shame the Chinese didn't exactly clone the witschi machines as they have other features that come in handy for stuff like this. Such as frequency mode where the machine will tell you how many beats per hour the watches going. It also has the added bonus when it's in frequency mode the incoming frequency is used to run the graphical display. Or simplistically the display is somewhat flat and then you can look at the display for errors in the display aspect. So in other words if you want to figure out like this watch frequency mode you could see that it's actually running super super slow and you can look at the graphical display and it's probably going to look reasonably decent in know it's just the timekeeping issue.

 

[Woolshire]

Well, looks like Klassiker nailed it on that trace overlap. I removed just one single balance screw (good tip John) and now the rate is -347 s/d. Amplitude and beat error still look pretty decent.

Also, and this seems odd to me, the balance IS original to the watch. The serial number is scratched on the underside. So, not sure how it got soooo heavy, but it is. It’s like someone decided to put a couple extra screws in it! Can anybody think of anything else that could have happened here? 

Edited November 25, 2022 by Woolshire

[JohnR725]

1 hour ago, Woolshire said:

So, not sure how it got soooo heavy, but it is. It’s like someone decided to put a couple extra screws in it! Can anybody think of anything else that could have happened here? 

I think I commented up above what the problem could be? What if somebody swapped a hairspring? Hairsprings don't have serial numbers on them and as I pointed out up above somewhere if you swap a hairspring you have to match the balance wheel to that hairspring. That would be the most likely scenario as balance wheels typically do not get heavy all by themselves.

Now that one screw confirmed what we suspected you can remove the opposite Screw to keep the watch poised and see how much better the timekeeping is.

[Woolshire]

Okay. I ended up (for now) removing mean time screws, the ones not near the arms. That did it for rate. It is currently +20 s/d dial down and -20 s/d dial up. I will likely end up looking for some smaller gold balance screws to use so I can leave all four mean time screws on the balance. I could reduce weight by filing some of the bigger screws but really don’t want to since they are gold.

Thank you for everyone’s help! I may have eventually stumbled upon the cause but would it would have taken substantially longer. This is an excellent example of the importance of stopping when baffled and seeking advice on this forum…what a great gift to both get help and provide it.

[Neverenoughwatches]

1 hour ago, Woolshire said:

Okay. I ended up (for now) removing mean time screws, the ones not near the arms. That did it for rate. It is currently +20 s/d dial down and -20 s/d dial up. I will likely end up looking for some smaller gold balance screws to use so I can leave all four mean time screws on the balance. I could reduce weight by filing some of the bigger screws but really don’t want to since they are gold.

Thank you for everyone’s help! I may have eventually stumbled upon the cause but would it would have taken substantially longer. This is an excellent example of the importance of stopping when baffled and seeking advice on this forum…what a great gift to both get help and provide it.

Well done wool, if the balance wheel is original then a weak torqued spring fitted at some point or the original become weak like a mainspring can ?  Or could the spring have been too long ? Not vibrated to match the balance wheel. Love to hear John's thoughts to the causes as well.

[JohnR725]

7 hours ago, Woolshire said:

I ended up (for now) removing mean time screws, the ones not near the arms. That did it for rate. It is currently +20 s/d dial down and -20 s/d dial up. I will likely end up looking for some smaller gold balance screws to use so I can leave all four mean time screws on the balance. I could reduce weight by filing some of the bigger screws but really don’t want to since they are gold

Then just a clarification Because is not clear in my mind. You removed one of the screws and verified your running super slow. Did you remove the opposite screw on the other side? Then Found out still slow and Removed the mean time screws or did you just remove the mean time screws?

6 hours ago, Neverenoughwatches said:

if the balance wheel is original then a weak torqued spring fitted at some point or the original become weak like a mainspring can ?

 

On 11/24/2022 at 6:04 PM, Woolshire said:

The serial number is scratched on the underside.

Then yes the balance wheel is original as the quote above indicates.

II snipped out an image and circles some things. Notice the dark spots like the one screw that looks very bad the regulator arm and even the balance wheel what exactly are those dark spots? Classically to me that looks like rust damage Or perhaps a corrosion caused by rust were water that's been cleaned up. As a reminder this watch has been serviced before who knows what's happened to it before but I've seen this before it always reminds me of rust. 

One of the problems with vintage watches are they been worked on a lot. We don't know their prior history. We do know unfortunately from Our own experiences that damaging or destroying a hairspring is actually quite easy to do. Then somewhere in the discussion was the suggestion that the balance staff might've been changed that's another opportunity for an accident to occur. It might not have even been the hairspring damaged some times when people restaffed they break the collet of the hairspring. Like wrong sized staff will do that sometimes or just careless handling.

Then to the best of my knowledge I've never actually heard of hairsprings getting weak like a mainspring. Otherwise all the watches of very old age would all be failing now and they're not

6 hours ago, Neverenoughwatches said:

Or could the spring have been too long ? Not vibrated to match the balance wheel.

At the time the watch was made hairsprings were available as a separate item. These premade hairsprings are not vibrated to match the balance wheel. There Made to very exacting specifications for timing reasons The stud is exactly at the end of the hairspring there is no room to adjust as that would screw up the specifications of the over coil hairspring.

If you pre-make your hairspring typically with over coils you then have to match the balance wheel to the hairspring. Easy to do in the factory because they have lots of timing screws harder to do in the field as we don't have typically those timing screws. But at the time hairsprings could've been purchased you would've been able to purchase the timing screws.

So basically we have a slow running watch with everything that's perfect the balance wheels the right balance wheel for the watch. Hairsprings don't have serial numbers but they're easy to destroy we see it all the time on this group newbies want to swap hairsprings. So that is a very likely scenario hairsprings been replaced. Somebody just didn't understand that you can't just replace you have timing consequences unless you have enough screws to fix that problem.

 

[mikepilk]

13 hours ago, Woolshire said:

Okay. I ended up (for now) removing mean time screws, the ones not near the arms. That did it for rate. It is currently +20 s/d dial down and -20 s/d dial up. I will likely end up looking for some smaller gold balance screws to use so I can leave all four mean time screws on the balance. I could reduce weight by filing some of the bigger screws but really don’t want to since they are gold.

I thought "gold timing screws", surely not, seems extravagant and what's the point?
But I just had a look at a couple of my "better quality" pocket watches, a 23J Elgin Veritas Grade 453 and a 21J Hamilton Grade 992 - and they do look like rose gold.

So then checking 3x Waltham Riverside (17J and 19J), 2x PS Bartlett - all look rose gold. Only on the lower grade 620 do they look like brass.    

 

[Neverenoughwatches]

2 hours ago, mikepilk said:

I thought "gold timing screws", surely not, seems extravagant and what's the point?
But I just had a look at a couple of my "better quality" pocket watches, a 23J Elgin Veritas Grade 453 and a 21J Hamilton Grade 992 - and they do look like rose gold.

So then checking 3x Waltham Riverside (17J and 19J), 2x PS Bartlett - all look rose gold. Only on the lower grade 620 do they look like brass.    

 

That's interesting mike,  gold is more than double the weight of brass. Thats an extra option for timing.

[JohnR725]

3 hours ago, mikepilk said:

"gold timing screws", surely not, seems extravagant and what's the point?

In higher grade of American pocket watches you'll find all sorts of gold. Gold timing screws found on some balance wheels. Gold settings for the jewels found in the watch versus using brass on the lower end watches. Some of the wheels are made out of a gold alloy never really nice pretty shiny look to them definitely look different than perhaps. Yes a lot a gold found on some watches. Even the gold colored watches are usually plated with gold.

 

[mikepilk]

17 minutes ago, JohnR725 said:

In higher grade of American pocket watches you'll find all sorts of gold. 

I'd forgotten about that. It's been many years since I've serviced my pocket watches. My Elgin Veritas has rose gold timing screws, jewel settings and the centre wheel is gold plated. I'd forgotten just how nice it is 
(jewels at both ends of the barrel, ah !.)

 

[Woolshire]

John asked if I removed the other normal balance screw= Yes. That made watch 700 s/d fast. So, looks like one screw made ~ 1100 sec/d difference!
The “off” mean time screws were removed to see if that was about the correct weight to remove. it was.

As far as rust, hardly any. You pointed out a tiny bit of discoloration on the reg arm, but certainly no rust on the hairspring. 
Gold; this watch has an entire solid gold train, Gold raised jewel settings,  (including for barrel), and gold balance screws. Also, the plates are white-gold flashed with fishscale damakeening….Pretty elaborate for a $200 watch today!

[mikepilk]

I have replaced hairsprings on pocket watches, and had to add/remove timing screws. As @JohnR725 mentions "At the time the watch was made hairsprings were available as a separate item. These premade hairsprings are not vibrated to match the balance wheel. "

[JohnR725]

7 hours ago, Woolshire said:

but certainly no rust on the hairspring

Early on in the discussion the reason I was asking was it looks nice and clean it probably isn't the original spring if the watch itself had water damage. Because often times you will see hairsprings with a little bit of rust damage and yours looks brand-new. As I said early on one were doing diagnostics it would be a clue that this is not the original hairspring.

I snipped out some images from the Elgin 1915 parts guide. As you can see an example of jewels in settings. They come in gold and basically brass or unset ones. Also came in a variety of materials different colors sapphire because there were probably still using natural sapphire they had fully perfected synthetic. So the pretty sapphire with gold settings would go and expensive watches.

Then you can even see they had gold plated screws at one time I guess you could adopt those. And then some information on the timing screw assortment.

 

[Woolshire]

I think I’m about done. In the end, I removed two regular balance screws and added two much smaller screws that I found in my “Junque”. Mean time screws remain on the balance. Dial down is 0 s/d and dial up is +7 s/d. The amplitude is impressive for an old pocket watch. Side positions are similar, with amplitude staying above 260.  I would like to get the beat error down to <0.5 but think I’ll leave it it for now.
Thank you all, again!

[Klassiker]

5 hours ago, Woolshire said:

Side positions are similar

Similar to one another, or also all a few seconds above zero? If so, you have got a much better result than I was expecting. Really well done and thanks for the updates.

I would be interested to what happens to the timing as the amplitude drops.

[JohnR725]

1 minute ago, Klassiker said:

I would be interested to what happens to the timing as the amplitude drops.

Normally it shouldn't be a problem unless somebody's opened up the regulator pins. On over coil hairspring is the regulator pins are supposed be as tight as possible as long as they hairspring can slide through the pins. As soon as I get opened up because somebody thinks it's post be like a flat hairspring then you'll get timing discrepancies with amplitude. Often times you will show up as you do multiposition timekeeping in other words as soon as you go to a pendant position you lose amplitude and you'll see it in the timekeeping.

Oh and I see from the pictures up above we are only doing to position timekeeping now? This is where if somebody would rotate the microphone to crown up with get an answer to the timekeeping versus amplitude plus would see if you have any poising errors after playing with the balance wheel.

[Neverenoughwatches]

8 hours ago, Woolshire said:

I think I’m about done. In the end, I removed two regular balance screws and added two much smaller screws that I found in my “Junque”. Mean time screws remain on the balance. Dial down is 0 s/d and dial up is +7 s/d. The amplitude is impressive for an old pocket watch. Side positions are similar, with amplitude staying above 260.  I would like to get the beat error down to <0.5 but think I’ll leave it it for now.
Thank you all, again!

Has been an interesting discussion woolshire, thank you .

[Woolshire]

I now have the watch cased and here is a photo of crown up after running for 12 hours. Not quite as good with losing 21 s/d in this position with still decent amplitude, but lower. Crown down, it is +26 s/d. Still near zero rate error in the flat positions.

I got lucky to find two screws that just happened to be correct for bringing to time. I am still surprised that one balance screw = 1100 s/d difference. Wow!

[Klassiker]

Winners make their own luck.