Surprising overbanking with correct mainspring. What size should I use ? GP Cal 27 (AS 1149)

[mikepilk]

I don't normally service other people's watches (for the reasons which follow!), but said I'd service a Girard-Perregaux Cal 27 manual wind.
It's basically an A Schild 1149, which is the same as an 1130 apart from barrel and train wheel bridges. So at least parts shouldn't be too much of a problem.

The watch is a disaster e.g crown wheel screw has had thread reversed, wrong screws sticking through where they shouldn't, worn jewels (including impulse), one balance pivot slightly bent etc. etc.

I finally got it running OK (surprisingly steady trace), but anything over half wind and I get overbanking (or should I call it 'rebanking'?).
Which, to say the least, was surprising given the poor state of the watch.

I'm happy I got the correct mainspring as listed by Cousins, Ranfft and a couple of other sources  : 1.70 x 0.13 x 360 

I could try oiling the pallet arbor jewels, but I don't think that will be enough.

My question is, how much should I drop the mainspring thickness from 0.13 ?  To 0.12 or 0.11 ?

 

There's also the question as to why it's overbanking. The balance looks too new to be original. I wonder if the wrong balance has been fitted ?

I couldn't get a timegrapher reading until I demagnetized it. Looking under the microscope, and in slow motion on recorded video, I can't see any coils sticking.

Edited June 1, 2021 by mikepilk

[JohnR725]

4 hours ago, mikepilk said:

rebanking'

rebanking is the correct term.

too much energy for the mainspring is with an issue for some time and a lot of watch companies have weaker mainsprings for specific calibers that have this problem. Omega actually has a working instruction that covers this. First they start off with replacing the balance pivot oil on the dial side with HP 1300 they claim may 10° to 15° drop in amplitude. then that doesn't work you're supposed to do the movement side balance pivot. Then if that doesn't work they recommend replacing the entire mainspring barrel. That is because their procedure of replacing the mainspring at least in the service center is just to replace the entire mainspring barrel. In your case I would drop to a much weaker mainspring if the heavier lubricant doesn't work.

 

 

http://www.ranfft.de/cgi-bin/bidfun-db.cgi?10&ranfft&0&2uswk&Girard-Perregaux_27

http://www.ranfft.de/cgi-bin/bidfun-db.cgi?10&ranfft&0&2uswk&AS_1130

 

[mikepilk]

24 minutes ago, JohnR725 said:

rebanking is the correct term.

too much energy for the mainspring is with an issue for some time and a lot of watch companies have weaker mainsprings for specific calibers that have this problem. Omega actually has a working instruction that covers this. First they start off with replacing the balance pivot oil on the dial side with HP 1300 they claim may 10° to 15° drop in amplitude. then that doesn't work you're supposed to do the movement side balance pivot. Then if that doesn't work they recommend replacing the entire mainspring barrel. That is because their procedure of replacing the mainspring at least in the service center is just to replace the entire mainspring barrel. In your case I would drop to a much weaker mainspring if the heavier lubricant doesn't work.

I have a copy of that Omega working instruction.

I was trying to get an idea as to how much effect reducing the hairspring thickness by 0.01 would have on amplitude.

I believe you can consider the spring as made up of sections of a beam - straining the brain cells, I remember the moment of inertia of a rectangular beam as bd³/12. So changing the thickness from 0.13 to 0.12, factors the strength by (0.12/0.13)³ = 78%.   Using a thickness of 0.10 factors the strength by 60%

 

[Nucejoe]

Have you tried oiling fork pivots? 

  I ponder if mainspring for GP27 has the same strength as that of AS1194 ?   

 I have a mainsprings for FF30 also know as FHF30 which Rolex59 is a modified varient of. All say on the package "Specially made for west end watch "    and it actually wont work in a Rolex59.  

 Mainsprings for AS1194 and GP27 might have different strength.

 

[mikepilk]

6 hours ago, Nucejoe said:

Have you tried oiling fork pivots? 

  I ponder if mainspring for GP27 has the same strength as that of AS1194 ?   

 I have a mainsprings for FF30 also know as FHF30 which Rolex59 is a modified varient of. All say on the package "Specially made for west end watch "    and it actually wont work in a Rolex59.  

 Mainsprings for AS1194 and GP27 might have different strength.

 

I wondered the same, but Ranfft suggests it's the same spring.

[nickelsilver]

You say the impulse jewel was worn, I imagine you mean a pallet jewel? Did you replace the pallet jewel(s), or the fork complete? If so it could be that you have a smaller than necessary lock, which can boost the amplitude quite a bit. It can be too small and still run fine on the bench, but would possibly have problems on the wrist. Generale Ressorts lists a 0.135mm thick spring for both the AS and GP version of this, haha.

 

Did you measure the replacement spring? I have gotten springs in the past in a little very old envelope for example marked Valjoux 72, watch rebanked like hell, measured spring and it was a couple hundredths oversize.

 

If you want to change the spring and are rebanking at half wind I would say drop 0.02mm off the bat, but do measure your replacement that's in there first.

[mikepilk]

1 hour ago, nickelsilver said:

You say the impulse jewel was worn, I imagine you mean a pallet jewel? Did you replace the pallet jewel(s), or the fork complete? If so it could be that you have a smaller than necessary lock, which can boost the amplitude quite a bit. It can be too small and still run fine on the bench, but would possibly have problems on the wrist. Generale Ressorts lists a 0.135mm thick spring for both the AS and GP version of this, haha.

Did you measure the replacement spring? I have gotten springs in the past in a little very old envelope for example marked Valjoux 72, watch rebanked like hell, measured spring and it was a couple hundredths oversize.

 

I did mean the impulse jewel. You can see the wear on the right.

The pallet jewels and fork look good, which considering the wear to the jewel, is surprising. I think the complete pallet fork nust have been replaced.

The mainspring was from Cousins, in a sealed packet, labelled correctly.  But as you said, doesn't mean that's what's in the packet. I'll measure it.

 

[spectre6000]

The beam analogy mathematics look good (though from a similar exercise about a year ago, I recall there is a formula specifically for doing this exact calculation... unfortunately, I don't have it written down anywhere and don't recall where I saw it save a recorded lecture on YouTube), but unless you know how much less energy needs to be transmitted to the balance through the gear train, it doesn't help much. 

Mainsprings aren't that expensive or difficult to replace. Why not try the one, and if that doesn't work try the other? Worst case scenario you or your buddy are out $20 (or however much for that particular spring) and/or you end up with a spare for your stash.

A similarly brute force approach would be to thicken oils all the way down the line in sequence like in the Omega lit until you reduce the amplitude to a good range.

[mikepilk]

1 hour ago, spectre6000 said:

The beam analogy mathematics look good (though from a similar exercise about a year ago, I recall there is a formula specifically for doing this exact calculation... unfortunately, I don't have it written down anywhere and don't recall where I saw it save a recorded lecture on YouTube), but unless you know how much less energy needs to be transmitted to the balance through the gear train, it doesn't help much. 

Mainsprings aren't that expensive or difficult to replace. Why not try the one, and if that doesn't work try the other? Worst case scenario you or your buddy are out $20 (or however much for that particular spring) and/or you end up with a spare for your stash.

A similarly brute force approach would be to thicken oils all the way down the line in sequence like in the Omega lit until you reduce the amplitude to a good range.

I'll try the oil, then go down 0.02 in thickness if I have to.

I'm intrigued as to how a watch, in such poor condition, can rebank. As the balance and pallet fork look to have been replaced, I wonder if a wrong part has been fitted.

I just noticed I have an AS 1951 scrap movement, which has a mainspring 1.70 x 0.10 (instead of 1.70 x 1.30)
If the spring looks OK, I'll try that to see what amplitude it gives.

Edited June 2, 2021 by mikepilk

[JohnR725]

4 hours ago, mikepilk said:

The mainspring was from Cousins, in a sealed packet, labelled correctly.  But as you said, doesn't mean that's what's in the packet. I'll measure it.

one of the problems with a brand-new mainspring  sealed in the package seems to be variations in quality? I see this with American pocket watch mainsprings which are packaged in Swiss packages and I recognize their Swiss part number. There does seem to be variations in quality from batch to batch even though you would think there shouldn't be..

 

1 hour ago, spectre6000 said:

Mainsprings aren't that expensive or difficult to replace. Why not try the one, and if that doesn't work try the other? Worst case scenario you or your buddy are out $20 (or however much for that particular spring) and/or you end up with a spare for your stash.

it's what the watch companies have been doing for years. Look at some of the Rolex parts lists they have different mainsprings for the same watch. Omega recommends changing the mainspring barrel which of course is really just changing the mainspring.. Elgin had in their final assembly verification of the amplitude. It was too low what went back to somebody to adjust it. It was too high it went back to where they had the mainspring barrels and they would swap barrels for a weaker spring. In their catalog they even indicate they used to have weaker mainsprings. Says not like this is a new problem it's an old problem that some watches just run better than others which is an interesting concept.?

2 minutes ago, mikepilk said:

I'll try the oil, then go down 0.02 in thickness if I have to.

I'm intrigued as to how a watch, in such poor condition, can rebank. As the balance and pallet fork look to have been replaced, I wonder if a wrong part has been fitted.

when you say the movement was in poor condition you didn't say it was rusted or the pivots were bands or? So other words wrong screws other stuff don't change the physical running of the watch. Plus as others I think I mentioned depending upon how the escapement was adjusted that can have a dramatic effect on amplitude.

[rodabod]

I would inspect the balance wheel to see if it looks like it is either incorrect, or has had a lot of screws/material removed. Lighter balances require less energy for a given amplitude. 
 

I’m less of a fan of the plan to reduce amplitude using heavier oil as the increased resistance causes the “Q factor” of the resonator (the balance) to decrease. It may not make a great difference in practice, but in layman’s terms, it makes the resonant frequency less precise. 
 

The only means I have for gauging how excessively thick a spring is, is by monitoring how the amplitude rises when winding from fully-unwound. It may reach a respectable amplitude after three turns of the mainspring, or it may start rebanking after a single turn, for example. There is a scientific way to gauge it, which is to measure the torque you are delivering into the mainspring while winding. This would be a similar, but far simplified method of what is done to determine the cut of a clock fusee cone. 

[JohnR725]

1 hour ago, rodabod said:

I would inspect the balance wheel to see if it looks like it is either incorrect, or has had a lot of screws/material removed. Lighter balances require less energy for a given amplitude

.when I was reading this realize I forgot to ask a question what happens when you turn the watch to a crown position maybe crown down how much is the amplitude?

I was thinking there is another thing that affects amplitude bro wait until I get an answer for this.

1 hour ago, rodabod said:

The only means I have for gauging how excessively thick a spring is, is by monitoring how the amplitude rises when winding from fully-unwound. It may reach a respectable amplitude after three turns of the mainspring, or it may start rebanking after a single turn, for example. There is a scientific way to gauge it, which is to measure the torque you are delivering into the mainspring while winding.

I think somewhere in the universe I've seen a picture of it the Swiss make a machine that measures the torque of the mainspring. But when we have several watch companies recommending replacing the mainspring or swapping barrels it indicates that every single mainspring may not be identical I know that's impossible that are made without much precision but maybe the watch companies think that the barrel's can need to be swapped may be changing the mainspring would be the answer maybe this one's just too strong

[mikepilk]

I suspect @rodabodthat the balance has been replaced by lighter one. All the screws are there, and none have been touched - which is a bit suspicious on such an old watch.

Oiling the pallet arbor pivots and removing lubrication from the escapement/pallet jewels didn't do enough.

I took the new spring out and measured it, should be 0.13 - it varies from 0.137 to 0.141. I guess the average at least 0.138

Out of interest I put in the old spring, which looks completely set and worn out. I try to re-use old mainsprings, but this one looked gone.

..... and I immediately get 270 deg DU/DD !   Problem solved.

I guess older movements were designed with mainsprings which did 'set' and not keep their shape like modern ones?

When I say 'poor condition' @JohnR725, it needs a new mainplate and barrel bridge : the stem slot on the mainplate is badly worn (someone has added some epoxy resin to get the winding pinion to engage the crown wheel). The Crown wheel screw has been re-threaded to use a clockwise screw - which was too long and prevented the clutch sliding. I used some threadlock to hopefully keep the crown wheel screw in.
The gear train isn't too bad, just a few chips around jewel holes, and the worn impulse jewel. Apart from a bent pivot :

Round the other side, the dial is very nice 

Assuming it's running OK tomorrow, I'll case it up and get it in the post a.s.a.p ! 

The only other time I've had rebanking problems @JohnR725 is with American pocket watches, especially the higher jewel count ones.  I have a lovely Waltham Riverside 19J that just wants to fly !

 

[JohnR725]

1 hour ago, mikepilk said:

Out of interest I put in the old spring, which looks completely set and worn out. I try to re-use old mainsprings, but this one looked gone.

..... and I immediately get 270 deg DU/DD !   Problem solved.

the problem unfortunately it may not be solved? A set mainspring will run a watch. The problem will become what is your amplitude at 24 hours and is a watch still keeping time if it is the problems solved if not you still have to deal with getting a weaker mainspring

1 hour ago, mikepilk said:

The only other time I've had rebanking problems @JohnR725 is with American pocket watches, especially the higher jewel count ones.  I have a lovely Waltham Riverside 19J that just wants to fly !

this is because if you look at the ancient books they would have one main spring that came at a variety of strengths. Unfortunately the American watch companies like to use the same part number for different parts variations in balance staffs for instance with the same part number different pivots sizes with the same part number and conceivably mainspring is with the same part number but different strengths. So in the case of a pocket watch a strong strength for 7 J and the week strength for 21 J. Then comes along the modern Swiss replacements a lot of times one strength only typically a stronger spring and you can have issues.

As I pointed out Elgin even had mainsprings outside of the normal range to take care of amplitude issues but modern replacements do not take that into account

[mikepilk]

1 hour ago, JohnR725 said:

the problem unfortunately it may not be solved? A set mainspring will run a watch. The problem will become what is your amplitude at 24 hours and is a watch still keeping time if it is the problems solved if not you still have to deal with getting a weaker mainspring

If it's over 200 deg after 24 hours it's going back!  I'm not charging for the service, only the cost price for parts.  I only offered to do it as I bought a Longines Admiral 5 star Cal 506 from him (which looks absolutely pristine)
I'm a big fan of vintage Longines' quality.

[mikepilk]

The good news is that after 24 hours running, I have a nice steady trace on the timegrapher and 255° amplitude.

As the last part arrived today - setting lever spring, I was expecting to get it finished. 

I was surprised to find the spring too big, and the wrong shape. It's only when I measured the movement I realised I don't know what it is !  Definitely not an AS 1130 

Hence a new thread : what-i-thought-was-girard-perregaux-cal-27-as-1149-isnt-anyone-tell-me-what-it-is/

[JohnR725]

2 hours ago, mikepilk said:

The good news is that after 24 hours running, I have a nice steady trace on the timegrapher and 255° amplitude.

just as a reminder your amplitudes are always going to look great dial up or dial down. If you really insist on one position timekeeping of the timing machine a 45° angle crown down would work otherwise dial down and crown down to positions because the amplitude is golf falloff in crown down or any of the crown positions.

But at 255° it's not good drop that much so you're fine.

[mikepilk]

6 minutes ago, JohnR725 said:

just as a reminder your amplitudes are always going to look great dial up or dial down. If you really insist on one position timekeeping of the timing machine a 45° angle crown down would work otherwise dial down and crown down to positions because the amplitude is golf falloff in crown down or any of the crown positions.

But at 255° it's not good drop that much so you're fine.

If it is an AS 1203, that explains the rebanking: mainspring size should be 1.7 x 0.115 and not 1.7 x 1.3 as for the AS 1130

The 'old' spring measured 1.7 x 1.3, which is probably why it's working so well despite looking 'tired'.

There had to be a logical solution to the rebanking, and the solution is : I'm an idiot   (for not spotting the correct movement).