Hermle 340-020 slowing down as days go by

[DirkP121]

Have a 1976 Hermle 340-020. It ran for maybe 10 years back then and then stopped. It sat unused until a few months ago when I decided to look into getting it running again. I cleaned the pivot points and oiled them. Then cleaned the escapement with a artist brush and a solvent. Then oiled it. The clock is now running. 

 However, I have been tracking its accuracy. What I see is that after winding it runs fast but slows as the days go by. For example.

1 day after winding- gained 30 sec in 24 hours. 

2 days after winding- gained 20sec in 24 hours.

3 days after winding- lost 28sec in 24 hours

4 days after winding- lost 1min 45 sec in 24 hours

5 days after winding-  lost 2min 30 sec in 24 hours.

Is this normal? If not, what could be causing this drop off in regulation?

Thanks

[clockboy]

Hi welcome to the forum. The issue you are experiencing with the clock is typical of a clock that needs servicing with a high probability that bushes need replacing. When fully wound this issue is overcome but as the mainspring unwinds there is less power and the clock slows.Looking at the barrel barrel bushes would be a good starting point 

Edited March 26 by clockboy

[oldhippy]

Thank you for your introduction and welcome to this friendly forum.

We all look forward to your contributions and continued involvement. 

We like new members to make an introduction before posting. 

Is it a pendulum movement or floating balance? As clockboy has said what the problems could be I totally agree. You will need to remove the springs from the barrels so you will need a clock mainspring winder, springs can become week and will not work properly due to lack of oil and dirt. Photos would be good we might be able to spot something. Please make sure they are clear , both plates and the inside of the movement. 

[DirkP121]

This movement has a floating balance. It is a mantel clock. I took some pictures as I did my work just to make sure I could put it back together. They are not extensive or the best quality. I don't want to take the movement out again to get more pictures, at this time, because it seems every time I do, it acts different when put back in the cabinet. I want to study its behavior for a while.

I have considered buying a new movement, but it bothers me that although it was made 38 years ago, it only ran for maybe 10 of those years. In terms of run time, its not that old. And then there is the concern that I might pay $400 for a new one and it has the same accuracy limitations.

[oldhippy]

The floating balance should be oiled, just a tiny drop of watch oil at the top jewel so it runs down the wire. You really need to get to the mainsprings as they are the source of the power.  

[DirkP121]

I have a copy of the Hermle service manual. It lists the lubrication point with a note saying the bearings on the old floating balance unit should not be lubricated. So I did not do it.

This is part of the discussion of the main spring in the manual. I get the feeling you are right about the main spring needing attention. Although I am a little (a lot) afraid to take down the movement that far. Will likely end up buying new movement to get a functional clock and then experiment with the old movement just for the fun and challenge of it. 

[oldhippy]

Well I have around 30 years experience restoring clocks and watches. I have serviced many floating balance movements and always added a tiny drop of oil to the floating balance,not only Hermle but Smiths it improves the rotation of the balance and I have never had any come backs. 

[nevenbekriev]

On 3/25/2024 at 7:59 PM, DirkP121 said:

1 day after winding- gained 30 sec in 24 hours. 

2 days after winding- gained 20sec in 24 hours.

3 days after winding- lost 28sec in 24 hours

4 days after winding- lost 1min 45 sec in 24 hours

5 days after winding-  lost 2min 30 sec in 24 hours.

Is this normal? If not, what could be causing this drop off in regulation?

Thanks

This is a nice table, but it misses the main parameter - the amplitude of balance oscillations.

As You say the movement behaves differently every time when something done to it - I will suggess the amplitude is not good and this may be because the balance/escapament interaction is not well adjusted. The 'guard pin' there is different compared to 'normal' movements. It is a little capricious to adjust, it has not to touch the balance in any of it's position

Edited March 27 by nevenbekriev

[DirkP121]

nevenbekriev-

Thank you for commenting. Every time I get a comment, it triggers new thoughts and insight.

You "liked" the table but asked about amplitude. Are you wondering what a table with amplitude verse days might look like? How would you measure amplitude? In degrees of rotation of the floating balance wheel?

You mentioned the guard pin. Is this the pin you are referring too? If so, are you concerned about it hitting what is the safety roller in the diagram 2?

These comments have got me doing more research. I found a document.

Hermle Floating Balance Summary of Inspection, Diagnosis, Cleaning and Repair Procedures … ..prepared by David Robertson

It states

Check for equal gaps between top and bottom tube ends and frame. I am now thinking this is the gaps shown below. (floating)

I noticed when I started this project there was not gap at the top but did not know enough to question that.

Am I correct that this is a problem?

Thanks to everyone reading this.

[nevenbekriev]

When a newbie starts to put His fingers in a clock movement, if the clock starts to tick, then the newbie's heart is full of joy, He is inpatient and wakes at night to check if the clock is still ticking... A ticking clock or watch is like a man that breaths and has pulse. And this doesn't mean that that man is healthy and able to do work. The same with the clock - if it ticks and even to the end of power reserve, this still means nothing and doesn't guarantee timekeeping at all. And the main parameter of a clock or watch health is the amplitude. As You didn't mention anything about amplitude, I guessed that it may be You case too...

Yes, the amplitude is measured by degrees of rotation of the balance wheel against it's 'zero' or neutral position. And a table with the amplitude will have one additional column, for the amplitude. Actually, I don't need this table, but just to know if the amplitude is good enough or bad, at least when the main spring is fully wound. Good will mean above 250 degr., and all less than 180 will mean too bad.

Yes, the guard pin and safety roller are what is shown and they would never touch when the balance swings. If they touch, this leads to friction and amplitude loss.

Yes, the gap with the arrow is 0 and it is a real problem, but it is true only when the string is in horizontal position, and I believe that the picture has been taken exactly in such position. In vertical position, which is normal for the balance, the mass of the balance wheel must straighten the hairspring in a manner that the balance will hang on the spring and neither the bottom of the tube, nor it's top will touch the limits. This is easy to check - try to lift or lower the balance along the string and You must see that it is possible and actually it hangs on the spring...

As I understand, You didn't disassemble the movement and cleaned it in whole. No, this is wrong, as all the abrasive particles that got in the bearings with the dust and as wear product will stay there, and the old hardened oil can not be fully removed. The main springs can not be cleaned at all this way too. Sometimes such cleaning leads to stop of the movement again in about one month, and sometimes leads to very fast wear of the bearings.

[DirkP121]

nevenbekriev-

You nailed it with your description of me and my reaction when the clock started ticking again. I am a newbie.  I love the sound and idea of mechanical clocks but the idea of owning one and trying to keep them running has never appealed to me. My wife bought this one and an antique German wall clock.  When I looked into having someone repair them for me, the universal response was "it's really expensive to work on them, you should just replace the movement". So, I had nothing to lose, I started researching them and opened them up. The wife is happy because she hears the sound of the clocks again. But I have gone down the "accuracy" rabbit hole.

In the vertical position, the balance wheel was not floating. It was sitting on the bottom of the frame. I adjusted the lower spring collet and got it floating. It easily passed the 270 degree 3 to 5 minute oscillation test. It took 8 minutes for the wheel to completely stop moving. 

I put it the unit back in the movement and checked the safety pin. It does not touch the safety roller anywhere in +/-270 degrees rotation from neutral position.

But the amplitude of the rotation with the spring fully wound is weak based on what you are saying. It rotates +/-90 degrees from the neutral position. 

No, I did not take the movement completely apart.  That seemed way outside my skill set at the time. There is a reason I became an electrical engineer and not a mechanical engineer. I am much more comfortable with moving electrons than tiny moving metal parts. Will I do it in the long run? Anything can happen. I don't seem to be able to let it go.

[DirkP121]

An update. After getting the balance wheel floating correctly and passing the 270 degree 3 to 5 minute oscillation test, I monitored the speed/accuracy for 5 more days. It gained 20 to 25 sec every 24 hours each of the 5 days. That is a big change from prior to getting the balance wheel floating. The speed is now consistent, although too fast. The bad news is the speed control on the balance wheel is already set to minimum. So I don't have any easy way to correct it. 

Also, 2.5 days into the 5 day period, I decided to record a video of the balance wheel and escapement to analyze at slow speed (frame by frame). Prior to starting the 5 day test period, I thought the amplitude was only 180 degrees. But when I viewed the video 2.5 days later, I was getting a full 360 degree amplitude. Not sure if my initial naked eye interpretation of the amplitude was wrong or if it really did increase over time. After winding, I get an amplitude of just under 1 1/4 turns.

The lever pin engagement with the pallets on the escapement wheel looks just fine. They are landing in the lock section of the pallets cleanly.

Not sure where I am going from here.

[nevenbekriev]

Why don't You show the video that You recorded?

Amplitude can not be more than 350 degr, as the rotation of the balance is limited. You rather have 180 than 360. But You have good inprovement.

If the balance is set to max moment of inertia and still the clock is faster, this should be because someone has changed/modified something in balance itself. The fact that the balance was not floating points to such thought. But it is possible to do something - put small equal weights (e.g. small watch screws...) in two opposite holes on the rim. Then 'floating' can change a little, so new adjustment of it must be done

[DirkP121]

I wanted to attach the video but It does not appear the forum allows videos. Just still photos. 

Note the excerpts from Daivd Robertson's Document

 

I currently meet these criteria. 

I know I can add weights. I just don't know where to get them yet. I have been looking for a reasonably priced used/broken one to take them from.

Thanks again for your thoughts on this. It helps me think through this process.