Positional deviation, balance problems

[Endeavor]

Dear All;

I encountered a problem which I haven't been able to solve yet. The movement is a Poljot 3133, 21,600BPH, lift angle 50 degrees. The movement has been serviced and runs reasonable healthy (very slight wave from the train) in both horizontal positions, DU & DD.

However, in all the vertical positions the beat-error stays good, the amplitude drops and daily rate literally "takes-off".

What have I done so far;

- Checked and adjusted the hairspring; took the hair-spring off the balance wheel, made sure that the arc of the hairspring matched the arc of the regulate pin & boot and that the collet is central above the jewel. It runs & oscillate precisely between the narrowly adjusted pin&boot. All coils are nicely concentric and the hairspring runs nicely centric.

- Changed out both anti-shocks, Chaton & cap-stone, no improvement.

- Changed out balance assembly for an assembly of a good running 3133 and the movement runs fine . Inserted the "troubled" balance assembly in a healthy running movement and again, it runs fine in horizontal positions, but takes off in all vertical positions.

- Changed out the entire balance cock + anti-shock; no improvement.

- Impulse jewel appears to be mounted solid.

- Balance end-shake seems very fine and pegged both balance-staff pivots multiple times.

- The pallet-fork end-shake seems fine and the "healthy balance" runs good with it.

Since the "healthy balance" runs fine in the movement, and the "troubled balance-assembly" runs also poor in the "healthy movement" I suspect that the problem lays in the balance assembly but I just can figure out the why

Has anybody any idea's what the problem is and suggestions where I could look next?

 

Dial-UP;

Dial-Down;

Crown-down;

Crown-Left;

CR & CU are yielding the same pictures.

Hope to hear ........

 

Edited May 26, 2021 by Endeavor

[Graziano]

Hi there,looks like friction from center wheel,excessive side shake maybe 

Edited May 26, 2021 by Graziano

[Endeavor]

At the moment I can't see anything wrong with the center wheel, but perhaps the problem lays in the train?

Could you please explain how you suggestion works and why it would not show up in the horizontal, but only in the vertical positions?

In the meantime I inserted another good running balance (not finely adjusted for the movement under investigation) and I does show the same behavior, but slightly less severe. Here again about 100 degrees are lost.

DD;

CD;

CR;

I will strip the movement and check. If the problem sits in the train, there are options left like changing the center-wheel (great-wheel or the first off the main-spring barrel), changing the whole main-plate, barrel-bridge etc.

What I don't understand is that when the trouble balance is placed in a good running 3133, it also runs poor?

Edited May 26, 2021 by Endeavor

[Marc]

The first step I would take would be to scrupulously clean the hair spring and demagnetize.

I would suggest that there is a slight stickiness to the H/S which is enough such that as the H/S sags in the vertical it brings coils close enough together to stick, causing the increase in rate, but in the horizontal positions the stickiness isn't strong enough to overcome the natural shape of the H/S without the help of gravitational sag, so no sticking together and the rate behaves.

It may be that if you were to inspect the H/S in the vetical positions you can see if this is an issue, but as you didn't mention cleaning the H/S or demagnetizing as steps you have already taked then this would certainly be where I would start.

[Endeavor]

Thanks Marc; even though not mentioned, but the hair-spring has been thoroughly cleaned. However not demagnetized. In vertical, while running, all coils appear to run free, but I will give your suggestion a try and report back.

Edited May 26, 2021 by Endeavor

[Endeavor]

Disassembled the balance wheel+hair-spring. It didn't show any magnetism on the Lepsi App, but nonetheless demagnetized the set. While disassembled, I took a picture of the hair-spring.

Assembled and installed, but sadly the same problem, be it with an improved amplitude;

DD;

 

CD and CL, same with CU & CR;

 

 

Hmmm ......

[nickelsilver]

The amplitudes are OK now. Check that the hairspring is bouncing evenly between the regulator pins. Sometimes is can be touching one pin all the time except at higher amplitude, where it pulls away; this will give a slower rate in high amplitude positions (horizontal), and speed it up at lower amplitudes. If all 4 vertical positions are running at +100 or so that is a likely culprit.

[Endeavor]

Yes, I spend quite some time making sure that the hair-spring curvature was "perfect" and that it bounces in horizontal positions equally between the boot & pin. The gap between boot & pin has also been reduced till a practical minimum. The 3133 has a normal pin & boot, not like the adjustable stud as with an ETA7750.

In vertical position with the balance facing up (CR) the hairspring seems to lean more towards the boot and with an amplitude 258. Balance down (CL) it is bouncing equally between pin & boot (244). CU it's leaning towards the boot(257), CD towards the pin (253)

In all vertical positions the daily rate is as shown in the pictures above.

Edited May 26, 2021 by Endeavor

[Endeavor]

To give some slightly more back-ground about this movement; I acquired it a number of years back and most likely, with another balance, I experienced a problem. Couldn't solve it and stored it in my Poljot 3133 spare-parts box as an "complete" movement but without the balance wheel.

The current balance + hairspring are new as supplied on eBay but again, as I suppose it was back then, I now experience some odd behavior.

This time I'm more determined to solve the problem as I need the movement for a new Junker 3133 dial + case.

 

 

Edited May 26, 2021 by Endeavor

[JohnR725]

7 hours ago, nickelsilver said:

The amplitudes are OK now.

can't quite put a finger on it but I don't like the timing machine results at all I don't think they're right? How easy it is it to adjust the beat on this watch? In other words can you just move the beat adjustment arm? The reason is any time a Chinese machine I see zeros I get suspicious that maybe it's not correct. So if you deliberately take the watch out of beat and put it back in the positions where it has zeros does it show that the beat is whatever you set it to?

The other thing happens is visually look at the balance when it's in the lower amplitudes how bad does it look? There are the examples of the timing machine thinking super low amplitudes are higher than they really are and that changes things. So I just don't like what I see in the timing machine but I can't quite put a finger on it. Which unfortunately if I'm right changes everything you're seeing.

So how would a normal functioning watch of this type look like on the timing machine?

6 hours ago, Endeavor said:

The current balance + hairspring are new as supplied on eBay but again, as I suppose it was back then, I now experience some odd behavior.

as the balance wheels been changed and that shouldn't be the issue what about the balance jewels themselves? Especially the whole jewels if one of those has a bad spot L show up in the crown positions and probably won't show up either dial up or dial down. Otherwise I would've suggested it could be the balance pivots but since it's a different balance wheel as long as the hole size is right for the size of the pivots that shouldn't be an issue.

[Endeavor]

Hi John;

thanks for jumping in too

With regards to the timing machine I've been thinking the same; rubbish in = rubbish out. Last night I was thinking to put some hands on it to see if it indeed runs that fast in the vertical positions. With those indicated daily-rates I should be able to see that in a short period of time.

I've been changing out the both Chatons & capstones and the balance-bridge; no difference. Last thing, even if it was just for the sake of it, is changing out the main-plate for a new (which I have in my spare-parts box).

Today, time permitting, I'll check the beat-error regulating point you brought up and also (again) swap the balance assembly to some good running 3133's and insert good running balances in this movement, trying to figure out whether I'm dealing with just the balance assembly or perhaps with more simultaneous problems.

Please give me some time to do all these investigations and I'll report back ......

[AlexanderToerzs]

Very much appreciated if you keep us updated on your findings. Cheers Alex.

[Endeavor]

Started this morning with throwing the beat-error way off and slowly returned to the same starting point, concluding that adjusting the beat-error is a consistent/repeatable point of the lever.

Leaving deliberately a beat-error to check the timegrapher performance, here are the DD, DU, CD, CL, CU & CR. The worst performer seems to be CU, but it is "remarkable" that all vertical daily-rates are about in a range between +73s/d and +83s/d.

Next plan of action is to change out only the main-plate, transferring everything ever as currently is.

Here are the timegrapher readings;

DD;

 

DU (as mentioned before, there is a small wave from the power-train)

CD;

CL;

CU

CR;

Edited May 27, 2021 by Endeavor