Need help with Unitas 6325 amplitude problem

[VWatchie]

 

I serviced this watch housing a Unitas 6325 just a month ago, and although the amplitude was borderline acceptable, 250/260 degrees dial up/dial down and about 220/230 in all other positions I decided to call it day and simply wear it and enjoy it. It definitely is one of my favourite watches as it so clean looking and so easy to read, and I paid less than $30 for it, which doesn’t hurt.

Anyway, yesterday I decided to tighten the cannon pinion because I had some intermittent problems with the timekeeping (full story here). After this I put the movement on my timing machine and the readings were bad, about 200 degrees dial up/dial down. Totally unacceptable.

So, I decided I needed to investigate. I checked the pallet stone locking (not good but not too bad), the oiling (looked just fine), the balance jewel settings (looked straight), end-shake of all wheels including the balance (I have limited experience of what is acceptable but it felt fine, tight but still moving freely with a very fine play), looked around for any dirt, dust, hair, debris under my stereo microscope but couldn’t see any, and I made sure the train was spinning freely. So, remaining to investigate, and the place where I probably should have started, the dreaded balance and hairspring. Dreaded because in this field my skill level is near nil.

What I saw looking at the balance and the hairspring is what you can see in the two above videos. As I remember it, it should be possible to move the regulator arm along the terminal curve without disturbing the hairspring. As you can see, this is certainly not the case. Also, to me it looks like the hairspring is almost pinched between the index pins of the regulator arm. To me it looks like it need more play, but I need your opinion.

When I first looked at the index pins I noticed that the outer and inner index pins weren’t exactly opposite one another, so I adjusted the outer index pin by placing a screwdriver in the slot of the boot of the outer index pin and turned it until the index pins were opposite one another (video made after this adjustment). However, this only seemed to aggravate the amplitude problem, and after this adjustment the hairspring looked even more pinched between the index pins.

So, I have a few questions; how likely is it that the low amplitude is the effect of the status of the hairspring and index pins? What can I do and what tools do I need to remedy this problem? I guess I could try to find a watch repairer but that would sort of defeat the purpose of this hobby of mine  Nevertheless, trying to form a new terminal curve (if necessary) would likely end in disaster, so I just don’t know… I’d hate to ruin this movement. Perhaps, that wouldn’t be needed, and I could just focus on the index pins (which I never did either but have some info about from @HSL, thank you), no?

Well as you can probably notice, I feel pretty bewildered about this, so I need you support. What’s your take on this?
 

 

[watchweasol]

Hi    according to convention when the balance is at rest the terminal curve should rest in the center between the boot and the pin. As you said you should be able to run the regulator arm  from + to - without .interfering with the spring    If the spring rests on one pin or the other it will affect the swing of balance and the rate Itshould beat evenly between the pins..

[watchweasol]

Hi  Watchie  you might find this interesting  its an old book by FW Brittan on springing and adjusting watches        cheers

practical springing.pdf

[nickelsilver]

I doubt the hairspring being out of true is affecting your amplitude, unless it has also thrown the beat error way out. If you look at the pic the first thing I would try would be to push the hs in at the green arrow. You can see it is bunched up near the blue arrow and spread out at the orange arrow- this should correct this (mostly) as well as bring the regulating curve more in line. Then you will want to push it one way or the other right at the stud, just a little bit, so that it bounces evenly between the boot and pin.

 

 

[Graziano]

In the video of the hairspring it looks to me like the hs needs an adjustment ,this will affect the amplitude .What I suggest is thake the balance off complete turn it upside down on bench and move the regulator from side to side .You will see if the hairspring is pulling or not .You can also make adjustments to the boot .If the hairspring needs adjustment watch this video , note the amplitude on this video.

no time better than the present to upgrade your skill.

 

Edited February 25, 2020 by Graziano

[CaptCalvin]

It more relates to positional error. What you saw can sometimes be done deliberately as part of the adjustment process. I dont think it's best practice but still a technique that exists in an adjuster's toolbox: to bear the hairspring hard against one pin which as you can imagine would cause the hairspring to move every which way as you move the index. I don't know if your hairspring is like that for this reason or just bad manufacturing, but it's unlikely that your amplitude problem stems from here.

Edited February 25, 2020 by CaptCalvin

[VWatchie]

17 hours ago, nickelsilver said:

I doubt the hairspring being out of true is affecting your amplitude, unless it has also thrown the beat error way out. If you look at the pic the first thing I would try would be to push the hs in at the green arrow. You can see it is bunched up near the blue arrow and spread out at the orange arrow- this should correct this (mostly) as well as bring the regulating curve more in line. Then you will want to push it one way or the other right at the stud, just a little bit, so that it bounces evenly between the boot and pin.

 

 

Thanks @nickelsilver! The illustration is extremely appreciated!

I'll try this. What tool can/should I use to push on the HS (I've never touched a HS before)?

Do I push the spring little by little until it looks good or do I use some other technique?

I suppose I can leave the balance in the watch (mainspring wound down of course), no?

Again, thanks!

[VWatchie]

@nickelsilver This is what the HS looks like after having manipulated it, to the best of my ability, according to your instructions. This was my very first attempt ever to manipulate a HS and it was a pretty intense, or should I say a tense, experience. To push the HS I used the tip of my Dumont no 2 after having thoroughly cleaned it, and I pushed slowly and gently several times. However at some point I probably over did it and I mangled the HS slightly where I pushed it (your green arrow), but I don’t think, or rather I hope it doesn’t have any significantly negative effect!?

This is the current position of HS between the boot and the pin after having pushed near the stud. Probably not perfect but better than it used to be. It seems to take an extremely small amount of manipulation to shift the position of the HS between the boot and the pin.

After a full wind, the graph on my timing machine (dial down) now looks much, much better! Before the HS manipulation the dots were scattered, the amplitude was below 190 degrees and it was very difficult to set the beat and the rate as they seemed to affect one another. After the manipulation it only took me a couple of minutes to set the beat and the rate. So, huge improvement but still the amplitude is of course unacceptable.

Moving the regulator arm along the terminal curve now have little effect on the HS. Not perfect perhaps, but definitely an improvement (apologies for the Darth Vader sound effect)!

Not sure this video brings anything to the table, but this is what it now looks like when ticking along.

I should mention that with lite power in the mainspring, 10 winds or so, the amplitude gradually diminishes until the balance comes to a complete stop after I’ve started it giving the movement a rotational twist. From start to stop it takes only about 30 seconds.

So, any suggestion about where to go from here?

[nickelsilver]

That hairspring looks great, good job!

 

As for the amplitude (and stopping in 30 seconds?), I suppose you've checked all the usual things, train freedom, endshakes everywhere, peg your holes especially for the fork pivots, check that the barrel has shake on its arbor and the lid is well seated.

[jdm]

At this point I suggest to record the noise in different positions. For that it would be expedient to build the simple adapter detailed in the latest postings of the topic below. Then you, or others here, can analyze it aurally and graphysically. That will reveal is anything is rubbing, very useful in strange cases, and an important feature of the best timegraphers.

 

 

[VWatchie]

3 hours ago, nickelsilver said:

That hairspring looks great, good job!

 

As for the amplitude (and stopping in 30 seconds?), I suppose you've checked all the usual things, train freedom, endshakes everywhere, peg your holes especially for the fork pivots, check that the barrel has shake on its arbor and the lid is well seated.

Thanks!  Gathering the courage to manipulate the HS was definitely a milestone in my development. Very proud of myself! 

Being completely wound down, when I wind the watch approximately 10 turns on the crown (and I definitely start feeling resistance from the mainspring) and give the movement a twist to kick start it, it takes about 30 seconds before it gradually comes to a complete stop. It just doesn’t want to run with too little power on the mainspring.

I haven’t checked that “the barrel has shake on its arbor and the lid is well seated”. To make sure, if I remove the barrel from the movement and manipulate the arbor there should be some shake, no? Or, do you mean shake in the space between the main plate and the barrel bridge, or both? By the way, I didn’t replace the mainspring as the existing looked OK (I guess). Maybe I should have replaced the mainspring with a new? Even if the mainspring isn’t perfect shouldn’t it be able to deliver enough power when fully wound to reach an acceptable amplitude, or could this possibly be the source of the problem?

Well, I usually don't peg the bearings as I clean them manually using warm water, detergent and a very soft, dense and fine bristled toothbrush (my wife is a dentist nurse and she brings me these and other "nice tools"). After rinsing (dipping) in IPA, when I check the pivot holes under my stereo microscope (40X) they look spotless, but I often wondered if they really are spotless or if there's some "invisible residue" of IPA and/or detergent still inside the holes. Although I have a good scalpel for the job, shaping the tip of peg wood is such a pain that I skip over this step when the bearings look spotless. Perhaps I shouldn't? What's your recommendation?

While on the topic of cleaning. I just got myself a 68 year old ELMA watch cleaning machine with replaced wiring (€160 on Swedish eBay www.tradera.se). It seems to be working well enough, but I haven't used it yet as I'm waiting for some ELMA Red and SUPROL PRO, and I intend to follow the cleaning procedure as described by ELMA in this PDF document. If the jewel holes look spotless after having used this machine would you still recommend pegging?

I have another similar calibre Unitas 6380 also suffering from too low amplitude (thread about it here). The pallet stone locking on both movements look pretty much the same, i.e. too heavy. So maybe that is the source of the problem, or perhaps that’s the normal locking depths for these calibres? The status of the 6380 is still the same but I decided to prioritize this 6325 movement.

Well, so many questions, but any input, ideas or comments are appreciated!

Maybe I should re-strip it, clean it, replace the mainspring and meticulously check everything as I assemble and lubricate… In that case I could video record and take pictures of anything suspicious and you guys could tell me what you think. Hmm…
 

[nickelsilver]

I would definitely recommend pegging at least the pallet fork jewels (the hole jewels). They can look clean but still be manky, and it doesn't take much at all there to zap power. I doubt the mainspring is the source of your problem. Especially if it is a modern one. Yes the barrel should have shake on its arbor, there was a thread on that in the last week or so. 10 turns of the crown should equate to at least 1 or more turns of the barrel arbor; this should run the watch for hours, not seconds. You have a power loss somewhere in the train. If the train runs free with the fork out and just one or two clicks of wind, then you have an issue at the fork, could be the adjustment of the locks but more likely something in the jewels or on the pivots or lack of endshake.

[Endeavor]

Perhaps you did, but I haven't read it. Things changed after you installed the modified cannon-pinion. I guess first thing to do is to pull it again and re-checked the amplitude ?

Perhaps the intermediate keyless wheels are running heavy, causing the pinion to slip in the first place. Now with the pinion tightened, you see the (power-drain) effects ?

I'm sure you did .....

 

Edited February 27, 2020 by Endeavor

[VWatchie]

6 minutes ago, Endeavor said:

Perhaps you did, but I haven't read it. Things changed after you installed the modified cannon-pinion. I guess first thing to do is to pull it again and re-checked the amplitude ?

Perhaps the keyless is running heavy, causing the pinion to slip in the first place. Now with the pinion tightened, you see the (power-drain) effects ?

Sure you did .....

 

Indeed, things changed after I had tightened the cannon pinion (I guess...) so I first removed the minute wheel and then the cannon pinion but unfortunately it made no difference at all. And you're right, I never mentioned it, but thanks a bunch for reminding me! These things are so easy to overlook. I've made sure the entire dial train (motion works) is running smoothly.

[VWatchie]

27 minutes ago, nickelsilver said:

I would definitely recommend pegging at least the pallet fork jewels (the hole jewels). They can look clean but still be manky, and it doesn't take much at all there to zap power. I doubt the mainspring is the source of your problem. Especially if it is a modern one. Yes the barrel should have shake on its arbor, there was a thread on that in the last week or so. 10 turns of the crown should equate to at least 1 or more turns of the barrel arbor; this should run the watch for hours, not seconds. You have a power loss somewhere in the train. If the train runs free with the fork out and just one or two clicks of wind, then you have an issue at the fork, could be the adjustment of the locks but more likely something in the jewels or on the pivots or lack of endshake.

I'll check all of these things and will report back! Thank you very, very much for your support! 

[Nucejoe]

Canon pinion could be rubbing, I just pull it out, eleminate it.

That lack of end shake may be the fault!  I loosen bridge screws or you should just feel end shakes if you can reach gears in the train.

 

[VWatchie]

Here I’ve removed the balance and the pallet to test the free running of the train. It looks and feels smooth enough to me, but please have a look to form your own opinion.

Here I have reinstalled the balance without the pallet to see if the balance is oscillating freely. Looks fine to me but please have a look.

Before shooting this video, I have meticulously pegged the pallet jewel holes in the main plate and on the pallet bridge to an absolutely spotless state. I’ve also pegged (polished) the pallet pivots to ensure they are absolutely spotless and shining. I also checked the shake and the pallet is moving freely. Still, the balance stops when there’s just a small amount of power in the mainspring. Looking at the video it seems the balance stops in the exact same position each time. I wonder if any conclusion can be drawn from that?

The pegging seemed to have some effect as the amplitude increased (dial down).

It looked as if the escape teeth and pallet stone impulse faces were dry. As I remember it, I had applied a tiny bit of Moebius 9415 pallet stone oil on the let-off (exit) pallet jewel, but as I wasn’t perfectly sure I decided to apply a minuscule amount again. After this the amplitude increased even a bit more.

I’ve let it run overnight and the amplitude has increased again. It now moves between about 240 and 260 degrees, but most of the time it’s around 250 degrees dial down and around 240 degrees dial up, give or take a few degrees, and around 195 degrees in all other positions.

So, I guess an improvement, but till not quite acceptable. I’d like to reach at least 270 degrees which I know isn’t too much to ask from this movement.

I have another Unitas calibre 6325 doing around 300 degrees dial up and dial down and around 270 degrees in all other positions. So, what else can be done or investigated?
 

[VWatchie]

Well, as this thread doesn’t seem to get a whole lot of love I’ll see if I can give you some ideas VWatchie.

My experience is limited, but as far as I can see the train doesn’t seem to suffer from any drag, so I really don’t think there’s any need to check the barrel arbor shake. Also, I noticed it took about a minute or so for the balance to stop oscillating (without the pallet installed) after you started it with that piece of peg wood, so I think that’s fine too. So, train and balance seem to be in good shape now that you have adjusted the HS with the help of nickelsilver (well done btw!)

The way I see it, things start to go bad once you replace the pallet. Now that you’ve pegged the pallet jewel holes and the pallet pivots did you check its shake and that the pallet is swinging freely? Also, have you checked the status of the pallet stones? Use your tweezers to make sure they aren’t loose or angled in some strange direction. While you’re at it, take a look and feel at the impulse jewel and the roller just to make sure everything seems OK. I know this calibre has fixed banking so it wouldn’t be easy to make any adjustments if needed. Nevertheless, test the pallet banking just to make sure the guard pin doesn’t lock up when banking.

Of course, it could be a matter of the pallet stones locking too deeply into the escape wheel teeth. I did see in the video that the balance locks up in the same position each time it stops. As already mentioned, my experience is very limited in this field, but I still think that it is plausible explanation. I’m sure more experienced members can confirm or refute this idea as needed. Anyway, you mentioned that you have another Unitas 6325 with a near perfect amplitude so why not test your movement with its pallet. If it solves the problem, you’ll know for sure it is that pallet that’s causing the low amplitude and locking the balance.

Good luck and keep us updated!
 

[VWatchie]

Thanks VWatchie, much appreciated! I'll look into the things you're suggesting and we'll see where it goes!

[Endeavor]

Yep, often you are your best own teacher .....

[Graziano]

43 minutes ago, VWatchie said:

Thanks VWatchie, much appreciated! I'll look into the things you're suggesting and we'll see where it goes!

Your funny 

[Endeavor]

Looking at the first video, showing the train without a pallet-fork, just the escape-wheel rotating; I can't be impressed with the way it rotates. Not sure either what you are doing; tapping or rotating the crown. If it was tapping and that caused the train to continue to run again, that seems suspicious, i.e some additional friction/drag or additional power-release by the mainspring(?).  I can't see any escape-wheel recoil either ?

The timegrapher lines seems pretty straight.

Edited March 1, 2020 by Endeavor

[VWatchie]

4 minutes ago, Endeavor said:

Looking at the first video, showing the train without a pallet-fork, just the escape-wheel rotating; I can't be impressed with the way it rotates. Not sure either what you are doing; tapping or rotating the crown. If it was tapping and that caused the train to continue to run again, that seems suspicious, i.e some friction/drag.  I can't see any escape-wheel recoil either ?

The timegrapher lines seems pretty straight.

Interesting input indeed! I’ll take it into account when investigating further.

[Endeavor]

I edited my remark with "or additional power release by the main-spring (?)"

Also, what was shown, was that just the train or with a cannon-pinion driving the rest of it ?

[VWatchie]

With cannon pinion (entire motion works) installed.

Edited March 1, 2020 by VWatchie
additional info added