Help with diagnosis: Dial up vs. dial down performance after service

[pubudeux]

Hey all,

Last night I completed a service of a Seiko 6309A movement. While the performance is much better now, I notice a troubling issue on the timegrapher.

Here is the reading dial down:

Here is the reading dial up:

 

Because I am seeing all that noise when the watch is dial-up, I'm assuming it is something gravity-related, which makes me think it might be the shock absorber on the balance pivot.

Does this make sense? Are there any other areas of the movement that might be effected like that by gravity in between two Dial Up and Dial Down?

 

[Klassiker]

Could be gravity-related, in the sense that too much end-shake allows parts to move to different positions. Then they interact differently with one-another.

What is noticeable here, apart from the "dirty" traces, is the poor amplitude (I am assuming almost fully-wound) in both positions, the difference in amplitude, and the difference in beat. That suggests a lot of different problems to me, which could include dirt, too much end-shake (pallet, balance, escape wheel), too little end-shake (train), bent banking pins (if this movement has them), worn jewels, and no doubt a number of other possibilities.

I would strip down the escapement, reclean and check all the parts for contamination and wear. When you have the balance and pallet out, check the train is running freely.

Let us know what you find.

[pubudeux]

Hey, thanks for that.

The train is running freely (with a puff of air and comes to a stop cleanly, same when barrel is installed and mainspring wound a bit) and the movement was assembled after a cleaning and visual inspection.

Im trying to find out if this seems like a common issue due to one of the shock absorbers being problematic (whether it's dirt, lubrication, etc.). But from what you're saying there's nothing you see that points to that over any other issue?

[Nucejoe]

Most common cause is staff pivot, excess end shake also lets different part of pivot come in contact with the jewel which exacrebates the retarding effect of a bent pivot. So before strip I check end and side shakes of staff, fork arbour and escape. What do you get in pendent up and down positions?

 

Edited July 25, 2020 by Nucejoe
shock spring can cause problem if doesn't keep the jewels down in place.

[Klassiker]

I'm no professional and certainly no expert either, but I don't think it's that simple. If it was just one shock absorber then the performance would tend to be better with that side of the movement uppermost. That's not happening here. You are losing 30-odd degrees of amplitude.

Assuming you cleaned, inspected and lubricated both shock absorber assemblies, and you didn't see any major issues, then I can't see it being e.g. a stray speck of dirt on one end-stone. If you didn't unspect them though, then they could both be cruddy, worn or cracked.

Otherwise, if the balance is running true and without excessive end-shake, then I'd be looking at the pallet as the next most likely culprit.

[HSL]

There could be a bunch of reasons for this but what one can see from those graphs the fault probably is in the escapement.
I would check the escape wheel pivots for ware and tare, and the escape wheel faces for ware and dirt. 
Next I would check the pallet fork pivots and the faces of the jewels.
The last point to look at is if the hairspring is flat, it could hit the regulator pin if not.

[pubudeux]

1 hour ago, Nucejoe said:

Most common cause is staff pivot, excess end shake also lets different part of pivot come in contact with the jewel which exacrebates the retarding effect of a bent pivot. So before strip I check end and side shakes of staff, fork arbour and escape. What do you get in pendent up and down positions?

 

Pendent up and down is "in between" these two kind of. Much better than the dial up reading.

I'll start with trying to re-set the shock absorber

[pubudeux]

1 hour ago, Klassiker said:

I'm no professional and certainly no expert either, but I don't think it's that simple. If it was just one shock absorber then the performance would tend to be better with that side of the movement uppermost. That's not happening here. You are losing 30-odd degrees of amplitude.

I don't know if I trust the amplitude reading on that timegrapher when the reading is so dirty. Just like the beat error shows as ~5 ms, when it is clearly around .1-.2ms as evidenced by the DD reading. 

[Nucejoe]

22 minutes ago, pubudeux said:

Pendent up and down is "in between" these two kind of. Much better than the dial up reading.

I'll start with trying to re-set the shock absorber

How can shock absorber affect amplitude or positional variations? Unless loose and fail to keep the jewel-chaton assembly down and let it move.

[Nucejoe]

Just now, Nucejoe said:

How can shock absorber affect amplitude or positional variations? Unless loose and fail to keep the jewel-chaton assembly down and let it move.

or you mean to switch upper and lower jewels around.

[pubudeux]

Exactly: its one thing I wasnt too sure about when putting it in whether it was securely placed. I'll report on the results here. If its no good, I'll investigate further.

[pubudeux]

1 minute ago, Nucejoe said:

or you mean to switch upper and lower jewels around.

One thing I wasnt sure about: is there an up side and down side for these jewels or are they completely flat on both sides? I assumed i could choose one side as the "down" side to oil after cleaning. These are Seiko Diashock settings.

[JohnR725]

 

1 minute ago, pubudeux said:

I don't know if I trust the amplitude reading on that timegrapher when the reading is so dirty. Just like the beat error shows as ~5 ms, when it is clearly around .1-.2ms as evidenced by the DD reading

several things need to be in agreement before the timing machine is telling you the correct information. Otherwise it's telling you you have a problem. The graphical display and numeric display should indicate something similar. then visually looking at the balance wheel it should resemble what you're seeing on the timing machine. For instance timing machine tells you have a 300° aptitude the balance wheel looks like it's barely moving you have a problem. this particular example is a common problem with all timing machines. If the amplitude is too low the machine reads the wrong part of the waveform gives you happy numbers it just aren't there.

then if the graphical display is showing totally random dots then the numerical display is totally worthless also. So many times people look at the numbers and and totally ignore the graphical display because they like the numbers better.

Then there's a problem you're having that 4.9 ms the line spacing is not right. I'm giving a link below you'll have to scroll down to the pictures and you'll see that 4 ms lines are easily twice as far apart as yours.

then when giving us timing results it's always nice if you would give us one crown position like crown down for instance.

2 hours ago, pubudeux said:

The train is running freely (with a puff of air and comes to a stop cleanly, same when barrel is installed and mainspring wound a bit) and the movement was assembled after a cleaning and visual inspection.

another test to try for train freedom is backspin? This is where with the pallet fork out you start to wind the barrel and stop. As you start to wind the mainspring will supply energy to the train the train will spin when you stop winding the trains momentum will carry it past the stopping point it will actually back wind the mainspring just a little bit. as you're watching the escape wheel spin and your winding when you stop the momentum of the train will continue forward you can see that then it will ask he stopped for just a second sometimes you can see it sometimes you can't and then the escape wheel will spin backwards. So if you see the backspin that's usually a sign of a train this properly free. Doesn't occur on all watches for variety of reasons though usually watches with sweep second hands it has Springs associated with them.

1 hour ago, Klassiker said:

I'm no professional and certainly no expert either, but I don't think it's that simple. If it was just one shock absorber then the performance would tend to be better with that side of the movement uppermost. That's not happening here. You are losing 30-odd degrees of amplitude.

usually for dial-up and dial down you will see an amplitude drop if the pivot is bad but? But here we have a graphical display indicating a problem in both images. There is a problem beyond dial-up and dial down plus having numbers that we do not believe on the timing machine at least I don't believe them.

then there is the other problem I seriously doubt that that's a 4.9 ms beat error because it just isn't.with a small graphical display of the Chinese 1000 machine the graphical display cannot indicate the same thing the numerical display indicates. So in a certain points you'll get a rollover effect where the lines are closer together than what they should be for the same numeric results. At 4.9 ms you should be just about on the extreme of the graphical display and that looks much closer to one or 2 ms so your graphical display is not agreeing with numeric results this is come up before you can see it in the link below scroll down you can see images.

https://www.watchrepairtalk.com/topic/9683-hi-strange-timegrapher-reading/

I am curious about how do you lubricate the escapement?

then visually or definitely having an escapement issue. I'm also suspicious of the numeric results for the amplitude. If you're amplitude is below 200° then you getting interesting affect on the timing machine everything gets magnified as the amplitude drops all your problems get much much worse. So as others of set up above you seem to be having more than one issue perhaps

1 minute ago, pubudeux said:

One thing I wasnt sure about: is there an up side and down side for these jewels or are they completely flat on both sides? I assumed i could choose one side as the "down" side to oil after cleaning. These are Seiko Diashock settings.

classically with the end stones one side is flat and one side is curved. The curved side goes out the flat side goes in. It's a really easy mistake for new people to make. If you not sure which side is which look for a reflection off the top of the jewel. Like a fluorescent lamp reflection is much more noticeable you'll see the curvature the jewel. They were definitely design with the flat side going in.

[Nucejoe]

4 minutes ago, pubudeux said:

One thing I wasnt sure about: is there an up side and down side for these jewels or are they completely flat on both sides? I assumed i could choose one side as the "down" side to oil after cleaning. These are Seiko Diashock settings.

One side is flat which is to face the pivot the other is domed

Edited July 25, 2020 by Nucejoe
i mean the end stones. jewels are same on both side

[Nucejoe]

JohnR is very right 4.9ms BE seems invalid.

[pubudeux]

10 minutes ago, Nucejoe said:

One side is flat which is to face the pivot the other is domed

This is very good to know. Since there is a 50% chance I put the wrong side facing the pivot, I will attack this first and then move on.

11 minutes ago, JohnR725 said:

I am curious about how do you lubricate the escapement?

Per the Seiko 6309 tech sheet I lubricated the shock absorber jewels and the pallet fork jewels with Moebius 9010.

[clockboy]

Dial Up & Dial down variances I have found is usually is either too much end shake or the the H/S is not flat . I would would check the end shake and if OK re-clean the balance assembly and re-lube the cap jewels. 

[Nucejoe]

36 minutes ago, pubudeux said:

This is very good to know. Since there is a 50% chance I put the wrong side facing the pivot, I will attack this first and then move on.

Per the Seiko 6309 tech sheet I lubricated the shock absorber jewels and the pallet fork jewels with Moebius 9010.

That doesn't explain the variation even if you did put the end stones upside down. Rarely a case of end stone just barely pushing on pivot can drop amplitude but the drop would be the same in FU and FD.

If the jewel- chaton sits incorrectly ( hard to acomplish) then we might say the shock springs are holding the jewels down in wrong position. 

I wouldn't tear the movement down, but check the endshakes,  proper seating of the jewels and examine staff and pivot under high magnification.

Good luck.

[JohnR725]

1 hour ago, pubudeux said:

Per the Seiko 6309 tech sheet I lubricated the shock absorber jewels and the pallet fork jewels with Moebius 9010.

I snipped an image out  of these 6309 tech sheet. So I assume you're oiling like it shows one drop on each of the stones correct? Because if you are you're doing it incorrectly even if that's what the tech sheet shows. Tech sheets are interesting in that they don't always explain techniques were quantity is they just show you where. If you have uneven lubrication on the escape wheel you can actually see that on the timing machine.

[JohnR725]

32 minutes ago, Nucejoe said:

That doesn't explain the variation even if you did put the end stones upside down. Rarely a case of end stone just barely pushing on pivot can drop amplitude but the drop would be the same in FU and FD

conceivably will probably turn out to be something simple but? It just looks like too much is going on here something just isn't right

ignoring the timing machine numbers visually looking at the beats of the watch is the watch in beat?

Then how exactly are you winding this watch up?

 

[pubudeux]

Hey all,

Thanks for all the replies and suggestions. To begin, I took out, wiped off oil with Rodico and re-lubricated cap jewel, making sure to lubricate the flat side, which would contact the pivot. I then set it and took care make sure the spring was secured in place.

Here are the readings:

Pendant Down:

Dial Down:

Dial Up:

Pendant Up:

 

So I know there is low amplitude, but it does seem like the escapement itself is working decently well. I wound the watch by winding the ratchet wheel with a screwdriver.

 

I think the weird readings were caused by my positioning of the cased watch on the timegrapher, as when I have the crown facing right, the readings come out dirty, and when facing left (away from spring handle as pictured), the readings come out clean. Pretty strange to me, but I guess there was some kind of interference or shaking with the crown?

 

I think my amplitude issues might be mainspring related, as this was my first time servicing a mainspring and I definitely didn't give it the best clean possible. The positive thing is, the amplitude is about 50-70 degrees better than what it was before the overhaul

Edited July 25, 2020 by pubudeux

[jdrichard]

 
classically with the end stones one side is flat and one side is curved. The curved side goes out the flat side goes in. It's a really easy mistake for new people to make. If you not sure which side is which look for a reflection off the top of the jewel. Like a fluorescent lamp reflection is much more noticeable you'll see the curvature the jewel. They were definitely design with the flat side going in.

Take a slo-mo video of the balance with your iPhone and determine the amplitude in both positions. If there is a significant difference, then go to all the other advice.

Sent from my GT-N5110 using Tapatalk

[Klassiker]

22 hours ago, pubudeux said:

Hey all,

Thanks for all the replies and suggestions. To begin, I took out, wiped off oil with Rodico and re-lubricated cap jewel, making sure to lubricate the flat side, which would contact the pivot. I then set it and took care make sure the spring was secured in place.

...

I think the weird readings were caused by my positioning of the cased watch on the timegrapher, as when I have the crown facing right, the readings come out dirty, and when facing left (away from spring handle as pictured), the readings come out clean. Pretty strange to me, but I guess there was some kind of interference or shaking with the crown?

...

I think my amplitude issues might be mainspring related, as this was my first time servicing a mainspring and I definitely didn't give it the best clean possible. The positive thing is, the amplitude is about 50-70 degrees better than what it was before the overhaul

Good to see you are making progress. That's looking a lot more healthy and truthful. Were either of the two end stones in the wrong way up, or is the "improvement" just down to having a better contact to the microphone? The microphone is situated under the metal "claw", so it is a good idea to put the crown up against it as you found out for yourself already. The crown transmits the vibrations better than the casing, especially if there's a plastic movement holder in between. Also, adjusting the gain to maximum I find often works to clean up dirty traces.

As it's an automatic, there's a good chance the mainspring is slipping too soon, and you aren't getting a full wind. You say you did't clean the mainspring too well, but (whisper it) they are not really that sensitive to a bit of dirt. There is a lot of energy in there, so contamination will increase wear and tear, but it is unlikely to cost you much amplitude. Did you lubricate the spring and barrel, and if so what with, how much, and where did you apply it? Another possibility is, you are applying too much oil at the escapement end, where a little bit of drag has a big effect.

[pubudeux]

2 minutes ago, Klassiker said:

Good to see you are making progress. That's looking a lot more healthy and truthful. Were either of the two end stones in the wrong way up, or is the "improvement" just down to having a better contact to the microphone? The microphone is situated under the metal "claw", so it is a good idea to put the crown up against it as you found out for yourself already. The crown transmits the vibrations better than the casing, especially if there's a plastic movement holder in between. Also, adjusting the gain to maximum I find often works to clean up dirty traces.

As it's an automatic, there's a good chance the mainspring is slipping too soon, and you aren't getting a full wind. You say you did't clean the mainspring too well, but (whisper it) they are not really that sensitive to a bit of dirt. There is a lot of energy in there, so contamination will increase wear and tear, but it is unlikely to cost you much amplitude. Did you lubricate the spring and barrel, and if so what with, how much, and where did you apply it? Another possibility is, you are applying too much oil at the escapement end, where a little bit of drag has a big effect.

Thanks for your reply.

Yea for some reason the crown on the claw side caused a ton of noise. It could be a quirk with the design of the 6309/casings as it happened on another similar Seiko as well.

I think the cap jewel fix I did improved amplitude about 10-20 degrees on average, but not sure. Now it is "settled in" at about 220 amplitude even after running for 12 hrs plus. Not the best but not too bad I don't think.

I lubricated the mainspring like so: 

1) run watchpaper with tweezers to get gunk off

2) watchpaper with tweezers dipped in iso to clean

3) dry watchpaper

4) watchpaper with 9010

5) dry watchpaper

I just didnt get alll the gunk off from the entire mainspring

I greased the barrel by applying 8301 to the walls at 4-5 points around the circumference, making sure not to get any on the top of the barrel

To your point about the escapement: I did get a little trigger happy with lubing the pallet jewels. I'm going to take care to be more restrained in my next service. I also will have a closer look at the mechanics of swiss lever escapement to find the precise friction points. I kind of just slathered it on the ends of the jewels imprecisely.

[JohnR725]

2 hours ago, pubudeux said:

I greased the barrel by applying 8301 to the walls at 4-5 points around the circumference, making sure not to get any on the top of the barrel

lubrication always such a interesting subject with lots of opinions and views. So 8301 is equaled 8300 which is considered a standard grease not a breaking grease.

2 hours ago, pubudeux said:

watchpaper with 9010

then 9010 is considered a light oil. If the mainsprings renew their considered to be prelubricated no lubrication is required. If you lubricating the mainspring 8200 which is a lighter version of your 8300 is the recommended lubrication. But I have seen people use 8300 for the mainspring itself. Other times they will use a clock oil as it's kind of a compromise between the oil and a grease. Probably better if he used to heavier oil on your mainspring as 9010 is just to light almost like not being there at all.

On the other hand if this works fine for you don't worry about its the rules on lubrication art entirely written in stone and are subject to some interpretation and variations most places.

2 hours ago, pubudeux said:

To your point about the escapement: I did get a little trigger happy with lubing the pallet jewels

escape but lubrication is one of those places where things get more critical. Ideally if you're following Swiss recommendations you're supposed to surface treat some of the components to keep the lubrication from spreading. Especially if you're using an oil. If you use a 9415