• The Bulletins and Marts are again available online. The network connectivity problem has been fixed. Thank you all very much for your patience.

Kundo flutter

clksmyhobby

Registered User
Jan 29, 2011
111
6
18
Central Florida
Country
Good day,

I have a 1950-ish Kundo standard, plate 1371 with suspension unit 1, given to me as a restoration project. That went well and the clock runs consistently. The escape wheel peninsula does not appear to have been tinkered with in the past. However, the mechanism is fluttering though I can't catch it in the act. Pendulum timing is perfect, but time jumps forward in large blocks.

The fork is very near the top of the anchor pin, fork tine spacing is a snug paper thickness and top block is a good fit in saddle bracket. I have read other threads about fluttering but thought it would be good to have a new one with specific information about this particular problem.

What other suspension related adjustments can be made?
Is it possible to be a pallet issue?

Thanks in advance,
Mike

99 dome.JPG
 

KurtinSA

NAWCC Member
Nov 24, 2014
4,843
355
83
San Antonio, TX
Country
Region
Mike -

I have a diamond dial Kundo likely the same plate. My fork is sitting maybe 3-4mm from the top of the pin. Normally one would raise the fork to stop flutter, but you're at the limit. Possibly the escapement is not correct, either drops or locks. Since the peninsula appears to be untouched, you might have to look at the pallets. But certainly verify the locks and drops and then decide if they need attention.

Kurt
 

clksmyhobby

Registered User
Jan 29, 2011
111
6
18
Central Florida
Country
Hi Kurt,
Locks and drops looked good at the start and clock operated without hesitation. I was thinking maybe the locks need to be a little deeper?
 

KurtinSA

NAWCC Member
Nov 24, 2014
4,843
355
83
San Antonio, TX
Country
Region
They should only be as deep as they need to be! I would restudy the locks and drops to confirm your initial assessment. Usually that is done by manipulating the anchor manually and then confirm under operation. If you think the locks need to be deeper, then that's something you can do.

When an escape wheel tooth drops off the lock face on the pallet and gets to about 1/3 the way down the impulse face, what is the position of the anchor pin? Ideally, it should be pointed straight up.

You may need to catch the flutter in the act. And see if the flutter happens on a specific position of the escape wheel. Could be there's an issue with one or more teeth.

Kurt
 

clksmyhobby

Registered User
Jan 29, 2011
111
6
18
Central Florida
Country
Good points. I started the clock again and reset the time when I made the initial post. It has not gained any time at all in 3 hours!

When I observe the action from above, the anchor pin moves the same amount left and right of center. A side view shows the anchor pin bent slightly to the rear, into the fork tines - not much, but noticeable now.

Rotation is just over 180 (typical of my other Kundo standards), but overswing looks kind of short. Will take it back to the shop and measure it after it runs through the rest of the day.
 

Wayne A

NAWCC Member
Sep 24, 2019
834
133
43
Country
Region
You may need to watch it for a week or more to be sure flutter is gone.
 

clksmyhobby

Registered User
Jan 29, 2011
111
6
18
Central Florida
Country
The clock is sitting in a visible area (dining room) with dome removed, where I hoped to observe or hear the flutter. It gained 10 minutes overnight after running normally since the original post.

Next step is tear down and recheck everything, looking at anchor pin/fork interface and pin alignment. Then I do a careful comparison of locks and drops in the clock compared to my other Kundos in the same period. Maybe this afternoon.

I used a new suspension spring supplied with the clock when I received it, the original was twisted beyond recovery. It was in a Horolovar packet and I did not measure it. As I look at it now, it appears wider and of different material in comparison with my other Horolovar springs.

Has anyone had to make pallet or eccentric adjustments to fix flutter on this type of mechanism? I have worked on more than a dozen Kundo standards without this difficulty. A challenge, for sure.

Best regards,
Mike
 

Wayne A

NAWCC Member
Sep 24, 2019
834
133
43
Country
Region
Has anyone had to make pallet or eccentric adjustments to fix flutter on this type of mechanism?
in a word, yes. The eccentric just to even up the drops but I have adjusted pallets longer to stop flutter that raising the fork did not stop. Also raised pallets to get clocks to run that were very low powered, not allot of adjustment range but a little goes a long way.
 

Berry Greene

Registered User
Oct 2, 2017
473
31
28
Chichester
Country
Region
This subject will come up for s long as we have mechanical torsion clocks. There are many weird issues. I cannot come to the table with many answers - BUT I do have many fulfilled and unfulfilled questions. Its knowing how to phrase them and in what order that's the problem.
Starting with the early taller clocks such as a Gustav Becker - which it would seem do work quite well, and them moving in one jump to the majestic (and accordingly priced) ATMOS by Jaeger LaCoultre which also works well given some environmental considerations.
From the other eye we see the so called "anniversary" and "400 day" models produced mainly in Germany in a period after WW2 from say 1947 to 1967. These are spring driven and increasingly of much smaller scale size. In my book herein lies a part of the flutter problem.
Let me admit to the "novelty value" of this group. Sold to the many servicemen posted in Germany to take home. Don't expect miracles. This is a novelty timepiece!
The usual variation of power associated with a spring drive are present. Coupled with the slow beat times of 10 or 12/sec means the averaging in errors takes a lot longer than a 18000bph watch. {Samples over time}.
Flutter is an aside from these issues which you can read about anywhere. The problem - it seems to me, - is that there are two resonances in the torsion mechanism. These arise because it must be impulsed in order keep running. The introduction of the fork splits the torsion wire into two sections. That below the fork; and that above the fork. The one below is ponderous and slow but the one above is much faster. All it needs is a sympathetic drive that can match its speed. As we scale down in size the weight of the required impulse also declines along with the EW and verge that provides it. They become so light that they can easily follow this high speed second resonance and thus is born a tendency to "flutter" which is not so prominent in the taller heavier models.
Defeating this is almost impossible. Its waiting there to pounce. We can make the motive power more equal by altering to a weight drive. However to defeat the flutter needs some smart thinking.
My problem is that I'm not that smart guy. However, I do have one promising idea that I will pass on to you. It simply consists of fitting an air damper to the verge. This inhibits fast movements and the idea is to remove any or so much sympathy with the flutter frequency.
This solution shows great promise with the smaller Midget, Bijou, variety. They say a picture is worth 1000 words and so I will attach an example.
Of course what I really want is someone to progress the idea or at least attack it from the same position. That this is a secondary resonance we do not want.
I have also thought about some form of better latching. Maybe, (forgive me for this), an electronic latch keyed by the actual position of the main pendulum. However it seems clumsy and surely will lead us to a quartz unit with the loss of true torsion timing. What we need is something more simple.
My attempts so far involve a piece of clear stiff plastic 12 x 20 x 0.45mm. It is held in place on the verge shaft by a small piece of Rodico. (The green square!). You might well prefer something more robust - but please ensure it is working first. You might need to trim yours down a little before you firm up. If you're careful there will be no permanent change to the clockwork.
Going back to the theory - it seems that you need that fork to be as high as possible so that the resonance is also as high as possible. You also need to make it difficult to impart the necessary energy to sustain it. This would mean as high up the pin as you can get before the impulses asked for become too much and stall. (Become insufficient to maintain the main rotation). Don't be afraid to tip the fork at an angle too. This suggests that you should not be aiming for, or expect, too much rotation. 180 degrees for a Midget (Kundo, Schatz Jum/7 K&S - all about 16cmH). Perhaps I should have said not less than 180 and not more than 275 deg.
Or just maybe you should avoid these small scale versions altogether......!
Variation in the spring power over the course of a year is never a recipe for consistency. You could minimize this effect by more frequent winding if substituting with a weight has no appeal for you! Find the mid point and work around that counting the clicks per month. Its not many!
Here are a few givens: A firm shelf without vibration - even ground tremors - consistent temperature, and draught free. Not much to ask is it? The mechanism must be clean and the pivots clean, loose, but true. The correct temperature compensated suspension wire/spring.
Remember this is as much a novelty clock for viewing its hypnotic mechanical motion, as it is a timepiece. That said the quartz has changed our conception of accuracy and what we expect.
I hope that helps. BerryG

400 day Kundo damping vane 21Sep2021 (3)_resized.jpg 400 day Kundo damping vane 21Sep2021 (1)_resized.jpg
 

Wayne A

NAWCC Member
Sep 24, 2019
834
133
43
Country
Region
All but one of my 400 day clocks are miniatures and midgets. Some designs are more prone to flutter, light suspension springs and over powered mainsprings need extra attention. I've been able to stop flutter with just the basic adjustments on all of them. The usual things, raising the fork, eliminating slop, adjust pallets or a combination of all.

Wayne
 

Schatznut

NAWCC Member
Sep 26, 2020
604
242
43
SoCal
Country
Region
All but one of my 400 day clocks are miniatures and midgets. Some designs are more prone to flutter, light suspension springs and over powered mainsprings need extra attention. I've been able to stop flutter with just the basic adjustments on all of them. The usual things, raising the fork, eliminating slop, adjust pallets or a combination of all.

Wayne
I generally have had the same experience as Wayne. The only time I've encountered flutter is when I've moved the fork down the suspension spring too far. I've never had a problem with too much power from the mainspring - quite the opposite. I regularly work on Schatz 1000-day clocks and they're basically a miniature movement with a huge honking mainspring. They've got fixed pallets so any adjustment must be made by the positioning of the saddle, which carries the rear pivot for the anchor. You'd think they'd be prone to flutter when fully wound, but I've never had one exhibit that trait.
 

Wayne A

NAWCC Member
Sep 24, 2019
834
133
43
Country
Region
I generally have had the same experience as Wayne. The only time I've encountered flutter is when I've moved the fork down the suspension spring too far. I've never had a problem with too much power from the mainspring - quite the opposite. I regularly work on Schatz 1000-day clocks and they're basically a miniature movement with a huge honking mainspring. They've got fixed pallets so any adjustment must be made by the positioning of the saddle, which carries the rear pivot for the anchor. You'd think they'd be prone to flutter when fully wound, but I've never had one exhibit that trait.
Schatz 1000 days have a long spring and fairly strong for sure as evident by the low rotation of the mainspring per year, 3.9rpy. Have three of these running and only one had flutter, was able to make that stop with the usual adjustments. Have Edgar Henn's and Koma's with heavy springs. My Koma 1392 only rotates the mainspring 3.34rpy with a heavy pendulum, takes some torque to do that. Now a Kundo 1379 mainspring makes 5.28rpy with light mainspring. Wish I'd started logging the gearing of my clocks sooner, oh well I'll get to them eventually.

Wayne
 

clksmyhobby

Registered User
Jan 29, 2011
111
6
18
Central Florida
Country
I forgot about this post until I saw the replies this evening, and I thank each for the follow-up. Barry's comments about resonances gave me a perspective on flutter I had never considered.

After another round of tear-down and inspect, I finally noticed that the top block was kind of sloppy in the bracket. I carefully closed the gap a bit and the fluttering was gone and has not come back.

I recently did an overhaul on a Schatz 53 and actually got the suspension bracket in a sweet spot on the first go, a novelty for me. Love those little clocks but have had my frustrations with them too.

Best regards,
Mike
 

Schatznut

NAWCC Member
Sep 26, 2020
604
242
43
SoCal
Country
Region
I forgot about this post until I saw the replies this evening, and I thank each for the follow-up. Barry's comments about resonances gave me a perspective on flutter I had never considered.

After another round of tear-down and inspect, I finally noticed that the top block was kind of sloppy in the bracket. I carefully closed the gap a bit and the fluttering was gone and has not come back.

I recently did an overhaul on a Schatz 53 and actually got the suspension bracket in a sweet spot on the first go, a novelty for me. Love those little clocks but have had my frustrations with them too.

Best regards,
Mike
Thanks for sharing this information. It feeds perfectly into Berry Greene's theory of secondary resonance.
 

Berry Greene

Registered User
Oct 2, 2017
473
31
28
Chichester
Country
Region
Sometimes I have been able to eradicate flutter by adjustment only to see it return after a few weeks of running. Even with monthly mid-winding it appears. Yes the smaller ones are worse. The K&S mini - all awful! The Haller & Schatz 53 (approx18cm) are taller but much better. I do not own a very tall version such as the Gustav-Becker, but I believe they can be much better BUT are still disappointing timekeepers.
BTW - I meant to include a word about setting the beat. Get a stethoscope from ebay/Amazon very cheap and listen to the escape while watching the over-swing. You want at least 1/2" of o/swing at each end of the rotation after the escape clicks. This is the power reserve within the pendulum. You can hear this through the base without removing the glass cover.
 

Schatznut

NAWCC Member
Sep 26, 2020
604
242
43
SoCal
Country
Region
Sometimes I have been able to eradicate flutter by adjustment only to see it return after a few weeks of running. Even with monthly mid-winding it appears. Yes the smaller ones are worse. The K&S mini - all awful! The Haller & Schatz 53 (approx18cm) are taller but much better. I do not own a very tall version such as the Gustav-Becker, but I believe they can be much better BUT are still disappointing timekeepers.
BTW - I meant to include a word about setting the beat. Get a stethoscope from ebay/Amazon very cheap and listen to the escape while watching the over-swing. You want at least 1/2" of o/swing at each end of the rotation after the escape clicks. This is the power reserve within the pendulum. You can hear this through the base without removing the glass cover.
I'm curious about your problems with the Kern minis. I've got a couple of them and they gave me no grief whatsoever when I was setting them up after overhaul. They're the later versions with the additional width between the plates; maybe that's why. Agreed about the stethoscope - that and the little beat gauge printed in the Horolovar guide are indispensable when setting beat and fork height.
 

Berry Greene

Registered User
Oct 2, 2017
473
31
28
Chichester
Country
Region
Thanks for the response Schatznut. it is much appreciated.
I was helped here on this Forum some 4 years ago when I was tackling my first Torsion clock. I had much to learn and owe a great deal such people as John Hubby, Bangster, Victor Miranda, Kurt and several others towards whom I feel a depth of warm gratitude. Along with other types of clocks and watches, my collecting hobby has proceeded with unexpected success and satisfaction that I cannot claim in full for the 400 day types. I must have around 10 here of my own and have "serviced" some more for friends. Yet in truth they all fall short of expectations. Of course a great deal is expected. A year or three of accurate running is, in my experience, out of reach. Now you might, along with me, question my methods and be absolutely correct in that I continually fall down the same hole. However, every one of these clocks seems set to disappoint me. I have accepted the variation of a spring drive power will produce changes of rate. However what I see is an inconsistency entirely in keeping with spasmodic fluttering. Indeed I set up an electronic beat counter, the results of which seemed to support flutter as the cause of a gain in indicated time against the numbers of beats described.
In the mean time I would read other reports on these clocks and it seems that they are well known in "the trade" for their poor relationship with real time accuracy. From time to time my mind returns to the subject wondering if there is somewhere to be found an elegant cure for flutter. Having allowed the dust to settle as it were, I have recently returned to the subject and experimented with the air damping vanes. I do seem to see an improvement. There is so far just one exception which is a K&S that stalls when a vane is fitted and does a lot better without a vane. I shall experiment with a reduced size vane. The gap between the plates of all 3 of my Kern & Sohne's is 20mm and the beat is 12bpm.
I have only recently read that even the Gustav Becker is also a disappointment. Indeed that ALL torsion models under perform.
I suppose my purpose in posting here again is to see if there is some shortfall in my procedure. It would help me to know the exact extent of your pendulum rotation. The exact position of the fork from the top suspension points and the angle, if any, of that fork. The exact spec of your chosen suspension spring and any other salient details that might help me to get better results. Especially that of curtailing the propensity to flutter. Believe me it is not from lack of trying that I approach this.
Sincerely, BerryG.
 

Schatznut

NAWCC Member
Sep 26, 2020
604
242
43
SoCal
Country
Region
Thanks for the response Schatznut. it is much appreciated.
I was helped here on this Forum some 4 years ago when I was tackling my first Torsion clock. I had much to learn and owe a great deal such people as John Hubby, Bangster, Victor Miranda, Kurt and several others towards whom I feel a depth of warm gratitude. Along with other types of clocks and watches, my collecting hobby has proceeded with unexpected success and satisfaction that I cannot claim in full for the 400 day types. I must have around 10 here of my own and have "serviced" some more for friends. Yet in truth they all fall short of expectations. Of course a great deal is expected. A year or three of accurate running is, in my experience, out of reach. Now you might, along with me, question my methods and be absolutely correct in that I continually fall down the same hole. However, every one of these clocks seems set to disappoint me. I have accepted the variation of a spring drive power will produce changes of rate. However what I see is an inconsistency entirely in keeping with spasmodic fluttering. Indeed I set up an electronic beat counter, the results of which seemed to support flutter as the cause of a gain in indicated time against the numbers of beats described.
In the mean time I would read other reports on these clocks and it seems that they are well known in "the trade" for their poor relationship with real time accuracy. From time to time my mind returns to the subject wondering if there is somewhere to be found an elegant cure for flutter. Having allowed the dust to settle as it were, I have recently returned to the subject and experimented with the air damping vanes. I do seem to see an improvement. There is so far just one exception which is a K&S that stalls when a vane is fitted and does a lot better without a vane. I shall experiment with a reduced size vane. The gap between the plates of all 3 of my Kern & Sohne's is 20mm and the beat is 12bpm.
I have only recently read that even the Gustav Becker is also a disappointment. Indeed that ALL torsion models under perform.
I suppose my purpose in posting here again is to see if there is some shortfall in my procedure. It would help me to know the exact extent of your pendulum rotation. The exact position of the fork from the top suspension points and the angle, if any, of that fork. The exact spec of your chosen suspension spring and any other salient details that might help me to get better results. Especially that of curtailing the propensity to flutter. Believe me it is not from lack of trying that I approach this.
Sincerely, BerryG.
It is important that expectations are set realistically. These are simple movements that depend on everything being tuned for minimum friction to meet the objective of running for 400 days or more. Considering that even young ones are 50 or more years old and they originally were lubricated with oils that oxidize and dry out over time, it is evident that removing any trace of accumulated dust, dirt and dried oil is the most important step to restoring these clocks to maximum efficiency. Fortunately modern synthetic oils are readily available and inexpensive. Relative to petroleum-based oils, they neither gum up nor dry out, so as long as the clocks are kept under their domes one may expect them to run as well as the lubricants and the technique of the clock repairer allow them. They are never going to be superb timekeepers unless one has infinite patience and a deft touch, and we must content ourselves with that.

I have perhaps 40 torsion pendulum clocks that I've overhauled. Some run longer than others. The service formula is the same - get everything absolutely spotlessly clean, pay particular attention to the mainspring: after thoroughly and meticulously cleaning it, lubricate it with synthetic oil with friction modifiers in it so that the coils don't stick or bind against each other. Polish pivots and peg the pivot holes so they are clean as viewed under a microscope. Install the wheels one at a time and check for any interference, drag, friction and binding, correcting as necessary. Do not make any adjustments to the pallets unless you know exactly what you are doing and why. The only time to do so is if there is evidence someone has been in there and messed things up (a frequent ailment) or there is wear in the escape wheel and anchor pivots (an infrequent ailment). Oil sparingly. Install a new torsion spring using the Horolovar Guide as a reference. Set the beat. Patiently get it into regulation.

In his book, Joseph Rabushka devoted an entire chapter to the forks of these clocks. It is recommended reading.

If accurate timekeeping is of paramount importance, go to the local dollar store, plunk down a buck for the cheapest quartz clock you can buy, and be happy with the result.
 
  • Like
Reactions: KurtinSA

Berry Greene

Registered User
Oct 2, 2017
473
31
28
Chichester
Country
Region
Thanks for that thorough response.
I'm on top of everything you mentioned except the mainspring. I haven't got a spring winder so I can only wash it out and re-lube it. I have some thin synthetic oil that I use. I wonder if that's the common link to my failure? I do the cleaning and test for free running too.
No I'm not expecting super quartz type accuracy. A slow loss or gain - even both - would be acceptable. What I can't tolerate are these quite huge skips in time - always gains. I've tried to catch it the act. Only a couple of times in my presence. It can be promoted by putting more power in the train as the pendulum passes through its mid-point. I have an idea that its probably temperature related as many times its overnight after the day temperature drops. I don't have air-con here but I have tried various places for the clocks and can say that a firm vibration and draught free shelf is a distinct advantage. However it still doesn't eradicate the flutter gains.
While musing over this I dream of some kind of interlock that rules out any extra escapes until it is unlocked again. That would have to be a function of the pendulum position and it can hardly be loaded in any way. Hence non contact electronics .... and the end of reality and any chance of an elegant solution!!
I'll go back to the mainspring operation in my cases and see if I can find a way to make a better job of it. It could be sticking - of course it could - and temperature changes might well promote it too.
Thanks for your thoughts. Much appreciated.
Best regards, BerryG
 

Schatznut

NAWCC Member
Sep 26, 2020
604
242
43
SoCal
Country
Region
Poor mainspring maintenance will cause the clock to stall, not run fast. The symptoms you describe are indicative of flutter, as you've surmised. Rabushka suggests raising the fork on the suspension spring just to the point where the clock no longer flutters. If the clock then stalls, something is robbing power from the movement or the mainspring is not providing sufficient power.
 

Berry Greene

Registered User
Oct 2, 2017
473
31
28
Chichester
Country
Region
That very succinctly describes my experience Schatznut. Thanks. At first I struggled so much that I would have, and did for a while, settle for the clocks just maintaining their operation and their fascination. However, just prior to reading your posts I was motivated to try the flutter damper idea. Its early days yet but I can report a transformation amongst 5 of the worst performers. I have no doubt it is flutter. Your Posts revived something within me to try and chase this down even further. Thanks for your interest and your help. I must try to get a copy of the Rabushka book.
Rgds, BerryG
 

Ibehooved

NAWCC Member
Jun 9, 2021
95
17
8
71
Country
Region
It is important that expectations are set realistically. These are simple movements that depend on everything being tuned for minimum friction to meet the objective of running for 400 days or more. Considering that even young ones are 50 or more years old and they originally were lubricated with oils that oxidize and dry out over time, it is evident that removing any trace of accumulated dust, dirt and dried oil is the most important step to restoring these clocks to maximum efficiency. Fortunately modern synthetic oils are readily available and inexpensive. Relative to petroleum-based oils, they neither gum up nor dry out, so as long as the clocks are kept under their domes one may expect them to run as well as the lubricants and the technique of the clock repairer allow them. They are never going to be superb timekeepers unless one has infinite patience and a deft touch, and we must content ourselves with that.

I have perhaps 40 torsion pendulum clocks that I've overhauled. Some run longer than others. The service formula is the same - get everything absolutely spotlessly clean, pay particular attention to the mainspring: after thoroughly and meticulously cleaning it, lubricate it with synthetic oil with friction modifiers in it so that the coils don't stick or bind against each other. Polish pivots and peg the pivot holes so they are clean as viewed under a microscope. Install the wheels one at a time and check for any interference, drag, friction and binding, correcting as necessary. Do not make any adjustments to the pallets unless you know exactly what you are doing and why. The only time to do so is if there is evidence someone has been in there and messed things up (a frequent ailment) or there is wear in the escape wheel and anchor pivots (an infrequent ailment). Oil sparingly. Install a new torsion spring using the Horolovar Guide as a reference. Set the beat. Patiently get it into regulation.

In his book, Joseph Rabushka devoted an entire chapter to the forks of these clocks. It is recommended reading.

If accurate timekeeping is of paramount importance, go to the local dollar store, plunk down a buck for the cheapest quartz clock you can buy, and be happy with the result.
This is the best one page summary I have ever read about the care and feeding of these clocks. Now , off to a balky Schatz Standard
 

Forum statistics

Threads
169,753
Messages
1,481,666
Members
49,143
Latest member
Great Leslie
Encyclopedia Pages
1,060
Total wiki contributions
2,965
Last update
-