Advice on a less powerful mainspring

[DanB]

Hi all,

I completed a clean/adjust/lube on an FEF 380 from an old Doxa. I installed a new mainspring bought from O. Frei and got the one specified on their movement database. Based on the extremely high amplitude and its erratic performance, I'm guessing the mainspring may be too strong for the movement. I have re-cleaned and re-lubed the balance jewels but that didn't have an impact on the amplitude. I can use some advice on which less powerful mainspring to get, and/or if I am overlooking any other possible causes. I'm thinking the width, thickness, and length of the mainspring should be constants, so I guess I'm looking at less Dennison strength and/or less metric strength? I have attached two amplitude photos, one at 1/3 wind, one at nearly full wind, abd a photo of the O. Frei catalog description of what I installed. Ignore the rate in the timegrapher - I adjusted the rate to spot on at first but as the movement ran overnight it sped up. Any advice will be greatly appreciated.

Thanks!

[KOwatch]

If you're looking to depower, go thinner, like .105 or .10.

I think Dennison is a 1950s era, less-precise way of measuring springs: the thickness ("strength") only steps in .01mm increments rather than current .005mm increments.

I've heard of using a thicker oil in your going train to accomplish depowering, might be easier method, at least temporarily, to see if it resolves your issue than fitting a new mainspring.

Older spring materials were less powerful for the same thickness of spring, so it's not unheard of having to reduce the specified movement spring thickness when repowering a watch. The big change was going from blued steel to the newer white alloys, my impression was that transition happened in the mid-1940s , but who knows what it came with, FEF380 movement is of that era. Ranfft db lists your movement mainspring as .105 thickness; so by this, you've up-powered.

Is it a 380,380-1 or 380-2? Generale Resorts lists it as the  GR 3967 (or the longer GR 3969 if it's the 380-1), that's a listed .105 thickness there also

 

[DanB]

Hi KO,

Thanks for the reply. The movement is a straight up FEF 380, no date. So if I understand you correctly, metric strength of .11 is too high, so if I go to .10 I should be in the ballpark? I found one at O. Frei but although the metric strength was lower, the Dennison strength was higher at 11. Would that matter? Pasted below are the numbers for that mainspring.

[Nucejoe]

 Oil fork jewel holes and thicker oil on pallets.

Good luck.

[Ronp]

Hi Nuceyou,

nog to be unfriendly, but Wrong answer, never oil pivots from pallet. Resolve the issue. 
Why? F.e if the oil gets hot in the sun and gets thin watch will overbank again, it will respond less well to sudden changes in orientation for the pallet will follow much dampened. And offcourse additional wear on the pivots from the wheel after the mainspring and all the bearings in the barrel and barrel pivot bushings.
 

Edited October 28, 2020 by Ronp

[Nucejoe]

45 minutes ago, Ronp said:

Hi Nuceyou,

nog to be unfriendly, but Wrong answer, never oil pivots from pallet. Resolve the issue. 
Why? F.e if the oil gets hot in the sun and gets thin watch will overbank again, it will respond less well to sudden changes in orientation for the pallet will follow much dampened. And offcourse additional wear on the pivots from the wheel after the mainspring and all the bearings in the barrel and barrel pivot bushings.
 

Thanks for proving me wrong. 

Regards  joe.

[rodabod]

That doesn’t look like re-banking to me. 
 

How is the timing in pendant down and pendant left orientations?

[DanB]

Per Rodobod's suggestion, here is the performance in those positions plus pendant up. I included a dial up for reference and adjusted the rate only in that position at nearly full wind. Given the variation in rates, could it be a balance issue? Maybe something wrong with the pallet pivot jewel?

 

[nickelsilver]

You're not rebanking, but that is a lot of amplitude. What is the amplitude when you let the spring down about 1/2 turn?

 

As for the rate, that is not that bad for such a movment. If you close the regulator pins a bit it should bring the vertical rates up  some.

[DanB]

The amplitude at 1/2 wind is still at about 300 degrees (even at 1/3 wind). I'll try tightening the regulator pins some. I've never done that before.

My concern about the rate is, I adjusted it to that level of performance before, let the movement run about 12 hours while in the movement holder, and then when I checked the rate again it had shot up to about +90 secs a day. I gave it a wind and it was still that fast. So something is happening while the movement is running that impacts the rate. I'm guessing it could be because it's overpowered, but I'm not sure.

[KOwatch]

Some folks are much more into the thick oil trick than others. Prob a useful trick to help narrow down problems. Good point about the heat/viscosity changes.

Dennison "strength" just means thickness, although it's similar to wire gauges, lower number is thicker, higher is thinner. Just ignore the Dennison numbers and use metric if possible.  If it is overpowered (unclear if that is the case here), the metric .10 thickness might be underpowered (might be preferable to overpowered), as the GR specs show a .105mm mainspring. Good to go with the spec mainspring if it's available? I don't have enough experience to say if +.05 or -.05 thickness should make that much of a difference, prob case-by-case.

[DanB]

Hi all,

Thanks for all of the suggestions. I decided not to close the regulator pins. On closer inspection it looked like the hairspring was sitting just right between the pins and my fear was that any fussing could cause more harm than good. In other previous timegrapher readings, I did see some strange patterns that does suggest there has been some re-banking (I didn't know what those wild readings indicated, but now I understand what they likely meant). I bought another FEF 380 that should arrive soon. It was described as running, so assuming that is the case my plan is to clean/lube/adjust that movement and will reuse its own mainspring. If I get that movement to run at a solid standard, I'll temporarily sub that mainspring into the movement I'm now working on to see if that solves the problem. If it does, then I'll order a less powerful mainspring.

I'll keep you all informed of how things go. So good to be the beneficiary of all your combined knowledge and experience.

[Nucejoe]

On 10/28/2020 at 12:27 PM, Ronp said:

Hi Nuceyou,

nog to be unfriendly, but Wrong answer, never oil pivots from pallet. Resolve the issue. 
Why? F.e if the oil gets hot in the sun and gets thin watch will overbank again, it will respond less well to sudden changes in orientation for the pallet will follow much dampened. And offcourse additional wear on the pivots from the wheel after the mainspring and all the bearings in the barrel and barrel pivot bushings.
 

 

On 5/9/2020 at 7:44 PM, EatPeach said:

Hey guys. I was killing my time and reading the technical sheet of 7s26. I found seiko suggest watchmaker to oil both two pallet fork pivots. Then I check other models as well, the results are the same. The technical sheets of 7s26,7s36,7009 all recommend  watchmaker to oil the two pallet fork pivots with MobiusA( presumably 9010). I think this is not a very common practise. What do you think? What's the purpose of this？Will you oil it?

Have a read.

[Ronp]

As far as I know in my 15 years as professional watchmaker they are the only only ones advising this for some Seiko family calibers.

I have not seen a single Swiss calibre service manual with instruction.

[Nucejoe]

1 hour ago, Ronp said:

As far as I know in my 15 years as professional watchmaker they are the only only ones advising this for some Seiko family calibers.

I have not seen a single Swiss calibre service manual with instruction.

True, This is one exception, I neither knew nor beleived it until I read master nickelsilver's statement. And did you read the section that seiko later abandoned the practice. 

Regards 

joe

[Ronp]

I double checked tonight at the Seiko internal side and the manual still indicates oiling the pallet pivots.
Weird. I have no direct logical explanation for this.

The forces are so small so very little wear and tear on the pivots. Only thing I could think off is that the oil is compensating for a not so well polished surface of the pivots. It are after all mass produced at low cost parts. 

[Nucejoe]

28 minutes ago, Ronp said:

I double checked tonight at the Seiko internal side and the manual still indicates oiling the pallet pivots.
Weird. I have no direct logical explanation for this.

The forces are so small so very little wear and tear on the pivots. Only thing I could think off is that the oil is compensating for a not so well polished surface of the pivots. It are after all mass produced at low cost parts. 

Nickelsilver explains that the practice was to improve long term stability of time keeping. I hope I am saying right.

Seiko didn't pariticularely care about high amplitude with seiko5  and I dont recall any mention of epilame to keep oil stay put on fork pivots and jewels. 

Personally I don't oil fork pivots.

 

[rodabod]

The pendant-down and pendant-left rates are pretty bloody close at least. The high amplitude affects the isochronism negatively, however it depends on where the amplitude lands over 24hrs.

If the watch is really not timing well on the wrist compared to the rough average of your timegrapher readings then I’d look into other issues such as sticking hairsprings or rebanking when the balance is excited through motion - that’s something that some physically active people here have discussed recently and not something I’ve experienced myself, but it’s definitely a real thing. 

[DanB]

An update. I did install a less powerful new mainspring - metric strength .1 instead of .11 - and that reduced the amplitude about 20 degrees in both dial up and pendant down positions to something more reasonable. But what was more troubling was that as the watch wound down to about mid-strength, the rate dropped significantly to about -120 seconds per day. This was happening before I put in the new weaker mainspring. I swapped the balance assembly from another FEF 380 (cleaned and oiled) and that made the performance more stable. That tells me that maybe something is up with the original balance assembly, possibly in the balance complete, maybe the hairspring. I don't know. 

Attached is the performance today dial up and pendant down, pendant left. Big variation in rate and beat error and pendant left is a mess. I noticed the regulator pins are not as perfectly aligned as on the original balance assembly. I wonder if a slight nudge of the pin to make it tighter will improve the rate is down and left positions? I'm reluctant to mess with this since it's such an imprecise method, the fragility of the pin, and the possibility of disaster. 

Any other suggestions are appreciated.

[Ronp]

Please verify pendant left the hairspring is expanding and contracting free, no windings touching each other or rubber the the balance cock or wheel. Just to be sure.

second make sure the compass pins ar perfectly parellel to each other.

third, do not bent the regulator pins unless hou have seen by inspection under higher magnification (microscope)  the drop or locks are not equal.

4e: if all above is fine, check the height play of the pallet fork, this should be small be pallet needs to move freely. You can simply test this with the escape wheel and balance out of the watch when turning the movevent the pallet should easily drop to the other regulator pin.

5th: swap the pallet fork itself.
 

Please  let us know the results.

 

ps I do not like the traces, they are wobly, not straight lines, this is often due to issues with the escape wheel, dirty teeth, worn pivots, small particle in the jewelholes, not 100% clean pallet stones, not freely moving pallet etc., much more likely than an issue with the balance. I assume you used epilame for the ancher wheel and pallet stones and oiled them with a very small drop of 941 oil?

Edited October 31, 2020 by Ronp

[DanB]

I didn't like the traces as well and wondered what could be the cause. I did use 941 oil on the pallet stones and after the movement ran a bit I cleaned the pallet stones so hopefully there was only oil on the escape wheel teeth. I did not use epilame - that has not been part of my practice as of yet. I guess it's time to invest in it. I'll go through some of your other troubleshooting suggestions as well. Thanks!

[Ronp]

Yes, epilame, applied in the right way is very important.

[rodabod]

I agree with Ron’s advice. Something fairly significant is happening in the pendant-left position so try to figure out why the amplitude has dropped off so far. 

[DanB]

Following Ron's and other's advice, I went through the list. 

-- Hairspring coils were not touching in any position. That said, the hairspring looked a little odd when I held the balance complete horizontally at eye level. The coils at the edge of their circumference sagged more in one direction then when I flipped it over. Could be nothing, but I swapped it for another balance complete.

-- Banking pins. Yes, one of the pins was not perpendicular to the movement plate so I moved it into position so that they were both parallel to each other

-- Regulator pin. It looked fine through a 10x, so I left it as is

-- Pallet fork end shake. This seemed fine as well

-- Epilame. Yes, I sprung for a bottle ($$$) and at the rate I work on watches I should have enough for 10 lifetimes. I treated the escape wheel (teeth only) and the pallet stones. Gave them a quick hair dryer treatment. Installed the escape wheel and pallet fork and let the watch run for 10 minutes, then oiled the pallets with 9415.

-- Swapping the pallet fork. I tried that but no difference.

Attached are the current timegrapher readings. I adjusted the rate for pendant down and pendant left since those are the most common positions the watch will be in (like right now when I'm at the keyboard). It's running fast dial up, but I figure I'll wear the watch a few days, see how it does, and will adjust if needed. The traces are not as smooth as I would like, but I'll take them. My sense is this FEF 380 was at best an average quality movement in its day, and this one doesn't even have incabloc shock protection. 

Again, thanks to all for your suggestions. 

[rodabod]

Is this the new balance? The pendant down/left positions are well matched. 
 

The timing differences will likely improve if the amplitude gains a bit.