Dreaded Fixed stud beat correction. Trial and error?

[Fraczish]

I've got a Longines 290 back together and running in its most basic configuration. However there is substantial Beat error circa 2.5ms. No corrector unfortunately so I'm going to have to adjust the collet (yikes!) 

My understanding from mark's video is to remove the balance from its cock and turn the collet. Is this just a case of trial and error to determine the correct amount and direction? Or is there a little Pro tip that could help?

Cheers all!

[clockboy]

This, what seems a simple job can go horribly wrong. On close inspection see if there is any reference point in relation to the collet split. Before any adjustment make sure the screwdriver shaft is clean with no grease deposits. If you have a ref point move the collet a tiny amount towards that ref point. re-install the balance and see if the direction is correct. Then move in tiny increments until you are satisfied. Getting this style to a zero beat error is difficult,  personally providing I get it within 0.9 or below I leave well alone. 

[Fraczish]

Thank you Clockboy. Useful tip on the reference point. 

[watchweasol]

Hi  Once you get the balance out of the watch check on the underside of the rim there is often a mark at the point where the stud is as a reference.  This adjustment can be done with the balance on a balance tack and working through the coils but needs great care. If its somthing you have not undertaken before remove the balance from the cock. but again check for the stud point if none put a mark on the rim yourself markerpen. You might find the attached useful.

[watchweasol]

Hi This video on the kink might be interesting for you      

 

[Michael]

Great video 

[Jon]

If you find turning the hairspring collet in situ, as watchweasol said, you may find it easier to remove the balance from the cock.

[Nucejoe]

I winder if manufactureres have a special tool for this task, hairsprings on old balance completes come so perfectly in beat.

[watchweasol]

iHi  There are specific tools for doing this job by Bergeon but at £45 a go not cheap. TMuir put up a post with a tool he made and I also have made them myself using the metal insert from a winscreen wiper blade and a lot of work with the Dremel and an oil stone.  The red,black,and green ones I bought over 30 years ago.

[praezis]

Fraczish,
if you are afraid of damaging something, leave it as is!
A beat error of 2.5 ms does no harm at all, it annoys owners of Chinese TMs only.

If you want to correct this "error", you can at least determine the direction of the needed move:
Release the mainspring fully. The now loose fork will stand a bit more near to one banking pin. In the same direction the collet must be rotated a tiny bit. Then test again with balance wheel mounted. 
To get comparable values, the balance must swing with same amplitude on each test. Procedure can get time consuming like this writing!

Frank

 

[mikepilk]

When cleaning the movement, I put just the balance back on the mainplate, then get a line of sight down the centre of the escape wheel/pallet jewels and between the banking pins to the impulse jewel. If not central, rotate the balance and make a mark opposite the stud.

You can do it with the movement assembled, but I find it easier with a bare movement and no pallet.

I can usually get to about 1 m/s just by sight. Which is good enough without an adjuster - I think people worry too much about a bit of beat error. As @clockboy says above, things can easily go horribly wrong.

 

[JGrainger]

I usually place the balance on a staking block and adjust with an oiler if I don't have a purpose made tool which fits. I like to find the place the stud should be by bringing the balance wheel to the correct position when fitted and with the help of a microscope graticule, move the microscope to look down on the balance vertically, rotate the graticule to an angle to match the stud and pivot - so I can easily see where the line of the graticule sits on the balance rim. I can then use a screw or marker pen mark (sometimes there's already a scratch), then use it to help get the stud on that same line of the graticule, between the balance pivot and a pre-determined screw or mark on the balance rim. The microscope isn't necessary but makes it easier to be precise.. I used to do the same but with a screwdriver instead of the graticule/ microscope.

[HectorLooi]

I made this tool last week. Got the idea from an item on ebay. I haven't tested it on an actual watch yet, but on a scrap balance, it works very well. With this, in theory, the collet can be adjusted in situ. The possibility of getting a zero beat error is very high.

[spectre6000]

23 hours ago, praezis said:

A beat error of 2.5 ms does no harm at all, it annoys owners of Chinese TMs only.

What, to you mind, is the limit to this?

[Nucejoe]

 Fine adjustment for "in beat"  is doable by moving hairspring through stud hole, just unpin the HS , adjust and repin.

Needless to say, no need to push the pin all the way home untill you are satisfied with the beat error and call the adjustment final.

 

[Nucejoe]

54 minutes ago, Nucejoe said:

 Fine adjustment for "in beat"  is doable by moving hairspring through stud hole, just unpin the HS , adjust and repin.

Needless to say, no need to push the pin all the way home untill you are satisfied with the beat error and call the adjustment final.

 

Farnkly speaking, I am lazier to even do this. Uncut pins come aprox 2 cm long,

which I use while still adjusting, once I called it in-beat, I cut the pin in situ and push the rest home inside the stud hole. No loosing the pin. 

[praezis]

6 hours ago, spectre6000 said:

What, to you mind, is the limit to this?

Hello,
short question, but a bit difficult answer.
This error has little to no influence on performance as it is compensated by right and left swing of the balance wheel.

There is little info on the topic. Values claimed in the net are just opinions without substantiation.

The manual of a timing machine gave the answer: limit for the esc. error is 15 degrees. Here the unpowered escapement still rests in lift position.

Unfortunately most TMs do not display the real esc. error (degrees) but just milliseconds that are not directly related to the real esc. error - ms value changes with amplitude and with bph despite a constant esc. error! One exception is PCTM that displays degrees, too. 

If you convert deg to ms, 15 degrees equates 9.6 ms @ 200 deg amplitude, @ 18000. Adjustment to 1/10 of that limit = 1.0 ms will be excellent.

Frank

[Nucejoe]

All above tools are handy to turn the collet with, but its still trial and error.

NOS balance completes are so fine adjusted to near zero beat error, I was thinking manufacturers might use a mchine of some sort to bulk produce and so finely adjusted.

[Fraczish]

Thank you for everyone's thoughts on this subject, very helpful indeed! 

I decided initially that I would tackle this and took the balance out. On closer inspection the index finger (is that what it's called?) Wasn't like any I'd played with or even seen before in that it didn't have a slot to rotate it in order to release the hairspring. At this point I decided that this was not the movement to be trying this on and I can live with it for now. I will probably revisit it when a little more confident. 

Definitely going to have a go at adjustment with the spring on, just on a few scrappers first. 

Thanks again chaps. 

Edited July 8, 2020 by Fraczish

[VWatchie]

On 7/6/2020 at 6:27 PM, mikepilk said:

I put just the balance back on the mainplate

@mikepilk I'm experimenting/practicing adjusting the collet on a scrap movement to build the confidence needed to adjust the collet on a balance that matters to me. I'm trying some different techniques. So, my question is simply this; when you put the balance back on the main plate, do you leave the collet/hairspring mounted on the staff with the stud in the stud hole, or do you remove the collet/hairspring first. If you don't remove it I guess you will have to hold the balance still (as it will want to swing back to its relaxed state) when making the mark opposite the stud? Hope that explanation was understandable! Thanks!

[mikepilk]

12 hours ago, VWatchie said:

@mikepilk I'm experimenting/practicing adjusting the collet on a scrap movement to build the confidence needed to adjust the collet on a balance that matters to me. I'm trying some different techniques. So, my question is simply this; when you put the balance back on the main plate, do you leave the collet/hairspring mounted on the staff with the stud in the stud hole, or do you remove the collet/hairspring first. If you don't remove it I guess you will have to hold the balance still (as it will want to swing back to its relaxed state) when making the mark opposite the stud? Hope that explanation was understandable! Thanks!

Yes. I have the balance complete with hairspring stud fixed. Then gently rotate the balance with a finger until the jewel is between the banking pins, and note the balance position - against a balance screw, or use a marker pen.

If it needs adjustment, I'll remove the balance, hang it on a balance tack, and make an adjustment (as Mark's video shows above). I filed down the end of an old screwdriver so that it will fit in to the slot in the collet. 
This is the part to practice. It's VERY easy to damage the hairspring. Older 'blue' hairsprings especially are very easy to distort, and when hanging on the tack, I'll try to support them to stop too much stretch. 
If it's a very small hairspring, or a more valuable movement with no replacement hairsprings likely, I'll reduce the risk by taking the balance off the cock, and adjust as @Jon shows above. There's always the chance of damage either way, so if I get with say, 2ms, I consider that acceptable and stop fiddling.

I wanted to know how much I needed to turn the collet to correct the beat error, so did a calculation - I can't find it at the moment, but from what I remember (for a 18,000 bph movement, and 250 deg amplitude) it's about 4 - 5 deg.

[mikepilk]

I found the calculation for balance velocity, thanks to David at https://www.vintagewatchstraps.com/shm.php :

 

So if the beat error is 1ms, the angular error is 4.2°.

 

 

[JohnR725]

2 hours ago, mikepilk said:

Yes. I have the balance complete with hairspring stud fixed. Then gently rotate the balance with a finger until the jewel is between the banking pins, and note the balance position - against a balance screw, or use a marker pen.

technically this is correct but it's also sort of incorrect? The problem is there's a tiny bit a play of the roller jewel in the fork slot. This means you can have the fork absolute perfect between the banking pins and the timing machine may disagree with your assessment.

2 hours ago, mikepilk said:

If it needs adjustment, I'll remove the balance, hang it on a balance tack,

personally I never hang balance wheels on the balance tack. In the past I found it stretches the hairspring and if you have just the balance with hairspring it makes it easier to see what you're doing as you can see where the stud is in relationship to whatever you are using as a reference.

 

[VWatchie]

2 hours ago, JohnR725 said:

technically this is correct but it's also sort of incorrect? The problem is there's a tiny bit a play of the roller jewel in the fork slot. This means you can have the fork absolute perfect between the banking pins and the timing machine may disagree with your assessment.

 

On 7/6/2020 at 6:27 PM, mikepilk said:

I put just the balance back on the mainplate, then get a line of sight down the centre of the escape wheel/pallet jewels and between the banking pins to the impulse jewel.

I believe there's been a slight misunderstanding!? As I understand it @mikepilk doesn't have the fork installed when making the assessment. Just the balance assembly.

[mikepilk]

3 minutes ago, VWatchie said:

 

I believe there's been a slight misunderstanding!? As I understand it @mikepilk doesn't have the fork installed when making the assessment. Just the balance assembly.

That is correct, I don't usually have the fork installed. As @JohnR725 mentions, it's won't be exact, but it's close enough for a starting point.

After cleaning the main plate + installed balance in the ultrasonic cleaner, I fit the balance jewels, then do the check.