Fluttering

[Nathan S]

Three 400 day clocks II've had running fairly accurately for a few months have all seemed to be fluttering suddenly. They all have new Horolovar temperature compensated suspensions. It has turned warmer here and wonder if that is the cause.
Any ideas?

 

[Wayne A]

Flutter can be intermittent enough to go unnoticed as flutter, just looked fast. Never associated flutter with small room temperature changes. Given enough power most 400 day clocks will flutter, so warmer room perhaps more power being developed. Another thing is did the fork to pin gap change with temperature, more gap promotes fluttering. Don't think the suspension spring would be a cause.

Wayne

 

[Schatznut]

The whole premise of the Horolovar spring is its stability over temperature (low coefficient of thermal expansion, or CTE), so that likely is not the issue. Whatever the cause, the solution is to move the forks ever so slightly upwards towards the top blocks. Do not assume that because you've installed new Horolovar suspensions that the fork-to-top block distances are optimized for your particular clocks.

For some fascinating background, look up Invar on Wikipedia.

 

[demoman3955]

The whole premise of the Horolovar spring is its stability over temperature (low coefficient of thermal expansion, or CTE), so that likely is not the issue. Whatever the cause, the solution is to move the forks ever so slightly upwards towards the top blocks. Do not assume that because you've installed new Horolovar suspensions that the fork-to-top block distances are optimized for your particular clocks.

For some fascinating background, look up Invar on Wikipedia.

good reading and info on invar. now im wondering what if anything is used on race engines, because expansion of some parts can kill a motor when you have clearances in the hundreds of thousands, like the thickness of some suspension springs

 

[Nathan S]

The whole premise of the Horolovar spring is its stability over temperature (low coefficient of thermal expansion, or CTE), so that likely is not the issue. Whatever the cause, the solution is to move the forks ever so slightly upwards towards the top blocks. Do not assume that because you've installed new Horolovar suspensions that the fork-to-top block distances are optimized for your particular clocks.

For some fascinating background, look up Invar on Wikipedia.

Regulator clocks I have made have Invar specified as the pendulum rod, but I learned that carbon fibre is even less prone to thermal change. One clock that I put a carbon fibre rod on is in a terrible place above a radiator and keeps excellent time. I have been told that carbon fibre actually contracts microscopically as it warms up. And it is much cheaper!
Bet this has opened up a can of worms!
As far as engines, my experience back last century on jet fighter engines was that parts were toleranced to be at the optimum size when the unit was at optimum operating temperature.

 

[Schatznut]

As far as engines, my experience back last century on jet fighter engines was that parts were toleranced to be at the optimum size when the unit was at optimum operating temperature.

Yeah, it's disconcerting listening to a turbofan engine windmilling on the ramp - sounds like a coffee can full of nuts and bolts slowly being turned over. Even though that's normal, it's a sound I never got used to.

 

[Nathan S]

Yeah, it's disconcerting listening to a turbofan engine windmilling on the ramp - sounds like a coffee can full of nuts and bolts slowly being turned over. Even though that's normal, it's a sound I never got used to.

Is flutter only ever caused by the fork position?

 

[KurtinSA]

Location of the eccentric and pallets can also result in fluttering. Fluttering is the result of failure to lock the escape wheel during rotation.

Kurt

 

[Berry Greene]

I just gotta say that it has been my experience that these clocks are always on the edge of flutter. Its a most misleading phenomenon as the clock can be rated to go slowly and use the random flutter to make up the lost time. It is therefore not truly isochronous. We can ask if that matters and were it to be a small and constant inconsistency but nevertheless averaging out to what is required, perhaps it would not matter. However, flutter is NOT consistent. While I don't have a total answer I can agree that raising the fork by tiny amounts does appear to reduce it. However, I have noticed that the smaller clocks suffer the worst and that it is precipitated by CHANGES of temperature ahead of all else. Maybe downward fluctuations that you get overnight in particular.
So we have two or maybe three things: The scale, weight, and temperature change. Now take ye a deep breath and step with me - not behind or ahead. I need your help here as much as you might need mine. As the fork turns under the energy of the rotation it moves the anchor & pallet(s) via the pin until it "escapes" its detent. The slope of the tooth then gives it a tiny lift and an impulse is delivered to the wire in a twisting motion that will keep it in torsion. If the rate of lift is too great or too fast it will snap back as the pallet drops of the tooth. Here is where we have the beginnings of flutter because the anchor may "snap back " precipitating a bouncing action and further escapes. A lot now depends on the weight of the fork, the mass of the pallets and anchor. Heavier parts will need more energy for a given flutter speed. An energy that might not be available.
To see how this works just lightly assist the minute hand and as the pendulum cluster unfurls the torsion spring it will flutter as it escapes. You can certainly reduce the incidence of flutter by fitting a simple wind brake to the arbour of the anchor. Read elsewhere here my other posts and see the photo. You can't be bothered, so I will re-post the photo IF I can find it....! {See below). A small clear plastic vane fixed with Rodico to the ancho arbour. It changes the resonance.
I have had quite startling results from this simple reversible modification applying it to several clocks in my collection. However, it isn't suited to every clock and frankly I don't know why!
Another allied issue is the drive. A mainspring output is not level / constant. You would do well, despite its "anniversary" name, to wind the clock more frequently, Say once a month and keep the mainspring centred in its winding up. A certain number of pawl clicks each month.
One more thing. The Atmos allows for fine adjustment by adjusting the actual length of the suspension spring. However most torsions don't provide a fine adjustment. This ball clusters are frequently far too loose and difficult to finely adjust so here are two more tips. Stiffen it up by applying a small band of rubber (elastic) to hold the balls in better tension. Now adjust the clock until it is running slightly fast. Bring down the speed by adding small weights. I use tiny balls of Blue-Tac. That's right I do that! I fancy adding little dishes for lead shot BUT I do not have the time or inclination. Distribute the the balls in equal opposite pairs at least and maybe to all four balls in the cluster. You can get good accuracy by this method. However, any flutter will ruin whatever you do. Be rid of it first. Raise the fork and fit a vane!
Now tell me, have I helped or hindered you? I have broad shoulders - I won't be offended... much!

Rgds, Berry G

 

[Berry Greene]

Reading through my post of yesterday I find myself wanting to expand or edit it slightly. Narrowing down the cause of flutter is one thing. A cure is something different. I think we are talking here about the power in the system matching to the natural rotation and furling / unfurling of the suspension spring. You would think it was automatic. That larger impulses would throw a wider arc of the pendulum - and to an extent this is so. However, the speed and extent of furling/unfurling changes as you approach the top block - where it is zero. When the height of the fork is very nearly matching the speed of the lift, a maximum use is made of the impulse and there will be no "twang" to precipitate flutter. This of course depends not only the amount of power from the mainspring, but that it is consistent and smooth. Therefore clean polished pivots and wheel teeth are a pre-requisite.
Some "experts" decline any use of oil at all. You can now see why although a tiny drop on the ES teeth will ensure they are polished. The teeth of the EW must be smooth and consistent, the pallets clean, smooth and precise in their landing. A study of the action will reveal how critical the speed of rotation of the EW, when it is set free, is to matching the rocking of the pallets. This is function of power reaching the wheel. Yet we know that changes of temperature could and would alter the power delivery from the mainspring, through the train, its teeth, pivots, and frictional losses. Now do we see all too clearly the shortcomings of this form of escapement.
Taking the ATMOS as our guide, there are quite a few differences that add up in its favour. The disc pendulum, the precise solid weight adjustment, and the precise spring length adjustment. The greatest single difference is the power source from an atmospheric operated diaphragm that keeps the power so consistent. Only now do we need to consider the actual suspension wire and its temperature stability. In my view it is way down the list of important features. I have one or two 400 day items that are not fitted with a Horolovar suspension that keep time just as well as those that are. Marketing might include a stretching of the truth eh? BerryG

 

[Nathan S]

Reading through my post of yesterday I find myself wanting to expand or edit it slightly. Narrowing down the cause of flutter is one thing. A cure is something different. I think we are talking here about the power in the system matching to the natural rotation and furling / unfurling of the suspension spring. You would think it was automatic. That larger impulses would throw a wider arc of the pendulum - and to an extent this is so. However, the speed and extent of furling/unfurling changes as you approach the top block - where it is zero. When the height of the fork is very nearly matching the speed of the lift, a maximum use is made of the impulse and there will be no "twang" to precipitate flutter. This of course depends not only the amount of power from the mainspring, but that it is consistent and smooth. Therefore clean polished pivots and wheel teeth are a pre-requisite.
Some "experts" decline any use of oil at all. You can now see why although a tiny drop on the ES teeth will ensure they are polished. The teeth of the EW must be smooth and consistent, the pallets clean, smooth and precise in their landing. A study of the action will reveal how critical the speed of rotation of the EW, when it is set free, is to matching the rocking of the pallets. This is function of power reaching the wheel. Yet we know that changes of temperature could and would alter the power delivery from the mainspring, through the train, its teeth, pivots, and frictional losses. Now do we see all too clearly the shortcomings of this form of escapement.
Taking the ATMOS as our guide, there are quite a few differences that add up in its favour. The disc pendulum, the precise solid weight adjustment, and the precise spring length adjustment. The greatest single difference is the power source from an atmospheric operated diaphragm that keeps the power so consistent. Only now do we need to consider the actual suspension wire and its temperature stability. In my view it is way down the list of important features. I have one or two 400 day items that are not fitted with a Horolovar suspension that keep time just as well as those that are. Marketing might include a stretching of the truth eh? BerryG

This is a clear and comprehensive explanation of something close to Black Magic! There is much well thought-out theory in there to be considered.
I don't know why, but it made me remember a domestic plumbing problem of fifty years ago. We were suffering from the dreadful phenomenon called "water hammer" and an old plumber of many years experience fixed it.
I thought he was joking! He tied a piece of ordinary string around the feed pipe and it ceased for ever.Harmonic vibrations spring to mind.
You do add fuel to my original musings about temperature change as three clocks that had run for months all started fluttering at around the same time.
Without doing anything they are all ok again.(one I raised the fork and it stopped, so put it down again) I have been using a very simple usb microphone to hopefully hear a flutter and not have to watch nothing happen!
Another thing that I think about is that the position of the fork also affects the length of spring able to flex.
I once was happy making all the parts and building 30-day regulators and the like!

 

[Schatznut]

Barry, lots to think about in your posts. I've worked on many 400-day and 1000-day clocks, and I've often wondered about the finish on the escape wheels. Every one of them I've seen has been stamped, and not one of them shows evidence of any secondary process to smooth them. I've examined a lot of them under a microscope, and they all have the characteristic surfaces left by the stamping dies. It would appear that the manufacturers didn't see the need to smooth the ramp faces. If there was any benefit to doing so, it apparently was not worth the cost. I agree that just slightly touching the escapement teeth with very light oil provides a tangible improvement in operation.

There is an uneasy balance (pun intended) in the location of the fork - too high and the portion of the suspension wire between the fork and the top block becomes so stiff that the clock will not run, and too low, flutter may occur. It seems to me that the resonance is more likely to be in the long portion of the spring, and upsetting the resonance point should be possible by changing that length, perhaps by simply repositioning the bottom block, or by making up a new suspension with a slightly different lower length. The period is much more strongly affected by the cross-sectional area of the spring than its length, so this would not have a significant effect on regulation.

Nathan, re the water hammer, whenever we would turn off the water in our shower, that whole end of the house would vibrate with a strong resonance for up to 30 seconds. This one was so sensitive that it didn't matter whether we closed the valve quickly or slowly, and I searched in vain for it for a very long time. One day, as this racket was going on, the toilet tank cover started rattling, and I removed it to find a resonance pattern in the water in the tank, centering on the fill valve assembly. I touched the top of the valve, and the racket immediately stopped. I replaced the valve with one of a different style (and obviously a different fundamental frequency) and peace has reigned supreme ever since.

 

[Berry Greene]

This is a clear and comprehensive explanation of something close to Black Magic! There is much well thought-out theory in there to be considered.
I don't know why, but it made me remember a domestic plumbing problem of fifty years ago. We were suffering from the dreadful phenomenon called "water hammer" and an old plumber of many years experience fixed it.
I thought he was joking! He tied a piece of ordinary string around the feed pipe and it ceased for ever.Harmonic vibrations spring to mind.
You do add fuel to my original musings about temperature change as three clocks that had run for months all started fluttering at around the same time.
Without doing anything they are all ok again.(one I raised the fork and it stopped, so put it down again) I have been using a very simple usb microphone to hopefully hear a flutter and not have to watch nothing happen!
Another thing that I think about is that the position of the fork also affects the length of spring able to flex.
I once was happy making all the parts and building 30-day regulators and the like!

Ha Ha! Oh you gotta laugh. A bit of string...... Thanks for that response Nathan S.
It turns out that the length of the suspension spring has a quite tiny effect on the rate. Its thickness works contra to common sense, and flutter is like the work of a leprechaun. Thirty day regulators do at least keep in regular time. These 400 day torsions have relied on slowing everything down and being very economic with the power - to such an extent that individual tiny pulses can be very varied and there's a lot less of then to produce a reliable average. That is so BEFORE flutter steps into the reckoning! For each one of several 400 day designs in my keeping, a different amount of rotation is required. From 180 to maybe 400 degrees. I can see it diminish slightly over a couple of months as the mainspring output changes. Wind it 4 to 8 weeks worth of clicks and the rotation goes back up!
I have tried all sorts of ideas in order to get around the flutter. I imagine a "once only once" device that will not allow fast repetition. Yet do I have no idea how it would work...! A very light toggle (snap) action on the verge arbour by virtue of a cats hair is as near as I have got to. It would create hysteresis. You see I know the words but I cannot convert them into a mechanical solution.
I've even thought of an electronic latch that looks at the position of the pendulum cluster before unlocking the verge to a single escape. Oh man who would buy into that? The beauty and the simplicity of my air vane solution deserves to work better than it does. It does have a minor impact on the hysteresis but its main contribution is air resistance. I have used a piece of clear plastic so as to be less visible. Approx 20 x 12 x 0.1mm held horizontally to the shaft with a small piece of Rodico has showed me best success. Just so long as the action doesn't stall you will be amazed at how much slower the time passes! Flutter can have a lot to answer for. Best of, - Berry G.
P.S. BTW - you DO have to oil the mainspring! Maybe not the pivots.

Barry, lots to think about in your posts. I've worked on many 400-day and 1000-day clocks, and I've often wondered about the finish on the escape wheels. Every one of them I've seen has been stamped, and not one of them shows evidence of any secondary process to smooth them. I've examined a lot of them under a microscope, and they all have the characteristic surfaces left by the stamping dies. It would appear that the manufacturers didn't see the need to smooth the ramp faces. If there was any benefit to doing so, it apparently was not worth the cost. I agree that just slightly touching the escapement teeth with very light oil provides a tangible improvement in operation.

There is an uneasy balance (pun intended) in the location of the fork - too high and the portion of the suspension wire between the fork and the top block becomes so stiff that the clock will not run, and too low, flutter may occur. It seems to me that the resonance is more likely to be in the long portion of the spring, and upsetting the resonance point should be possible by changing that length, perhaps by simply repositioning the bottom block, or by making up a new suspension with a slightly different lower length. The period is much more strongly affected by the cross-sectional area of the spring than its length, so this would not have a significant effect on regulation.

Nathan, re the water hammer, whenever we would turn off the water in our shower, that whole end of the house would vibrate with a strong resonance for up to 30 seconds. This one was so sensitive that it didn't matter whether we closed the valve quickly or slowly, and I searched in vain for it for a very long time. One day, as this racket was going on, the toilet tank cover started rattling, and I removed it to find a resonance pattern in the water in the tank, centering on the fill valve assembly. I touched the top of the valve, and the racket immediately stopped. I replaced the valve with one of a different style (and obviously a different fundamental frequency) and peace has reigned supreme ever since.

Thanks for the response Schatznut. On the strength of what you say I think I might turn away from 400 day madness to plumbing...... ! It's a good job we can laugh isn't it!

 

[Nathan S]

Barry, lots to think about in your posts. I've worked on many 400-day and 1000-day clocks, and I've often wondered about the finish on the escape wheels. Every one of them I've seen has been stamped, and not one of them shows evidence of any secondary process to smooth them. I've examined a lot of them under a microscope, and they all have the characteristic surfaces left by the stamping dies. It would appear that the manufacturers didn't see the need to smooth the ramp faces. If there was any benefit to doing so, it apparently was not worth the cost. I agree that just slightly touching the escapement teeth with very light oil provides a tangible improvement in operation.

There is an uneasy balance (pun intended) in the location of the fork - too high and the portion of the suspension wire between the fork and the top block becomes so stiff that the clock will not run, and too low, flutter may occur. It seems to me that the resonance is more likely to be in the long portion of the spring, and upsetting the resonance point should be possible by changing that length, perhaps by simply repositioning the bottom block, or by making up a new suspension with a slightly different lower length. The period is much more strongly affected by the cross-sectional area of the spring than its length, so this would not have a significant effect on regulation.

Nathan, re the water hammer, whenever we would turn off the water in our shower, that whole end of the house would vibrate with a strong resonance for up to 30 seconds. This one was so sensitive that it didn't matter whether we closed the valve quickly or slowly, and I searched in vain for it for a very long time. One day, as this racket was going on, the toilet tank cover started rattling, and I removed it to find a resonance pattern in the water in the tank, centering on the fill valve assembly. I touched the top of the valve, and the racket immediately stopped. I replaced the valve with one of a different style (and obviously a different fundamental frequency) and peace has reigned supreme ever since.

Isn't it wonderful, the tangents our brains take of on!

 

[Schatznut]

Ha Ha! Oh you gotta laugh. A bit of string...... Thanks for that response Nathan S.
It turns out that the length of the suspension spring has a quite tiny effect on the rate. Its thickness works contra to common sense, and flutter is like the work of a leprechaun. Thirty day regulators do at least keep in regular time. These 400 day torsions have relied on slowing everything down and being very economic with the power - to such an extent that individual tiny pulses can be very varied and there's a lot less of then to produce a reliable average. That is so BEFORE flutter steps into the reckoning! For each one of several 400 day designs in my keeping, a different amount of rotation is required. From 180 to maybe 400 degrees. I can see it diminish slightly over a couple of months as the mainspring output changes. Wind it 4 to 8 weeks worth of clicks and the rotation goes back up!
I have tried all sorts of ideas in order to get around the flutter. I imagine a "once only once" device that will not allow fast repetition. Yet do I have no idea how it would work...! A very light toggle (snap) action on the verge arbour by virtue of a cats hair is as near as I have got to. It would create hysteresis. You see I know the words but I cannot convert them into a mechanical solution.
I've even thought of an electronic latch that looks at the position of the pendulum cluster before unlocking the verge to a single escape. Oh man who would buy into that? The beauty and the simplicity of my air vane solution deserves to work better than it does. It does have a minor impact on the hysteresis but its main contribution is air resistance. I have used a piece of clear plastic so as to be less visible. Approx 20 x 12 x 0.1mm held horizontally to the shaft with a small piece of Rodico has showed me best success. Just so long as the action doesn't stall you will be amazed at how much slower the time passes! Flutter can have a lot to answer for. Best of, - Berry G.
P.S. BTW - you DO have to oil the mainspring! Maybe not the pivots.

Thanks for the response Schatznut. On the strength of what you say I think I might turn away from 400 day madness to plumbing...... ! It's a good job we can laugh isn't it!

Yep, the rules of plumbing are much simpler - hot on the left; cold on the right; water flows downhill and payday is on Friday!

 

[Wayne A]

Besides raising the fork to reduce flutter I've had success in eliminating it by increasing the locks. Deeper locks moves the drops further into the rotation where there would be a little more torque on the suspension spring which helps to prevent flutter. This really helps on flutter prone overpowered clocks. Think all 400 day clocks will flutter given enough power, so the thing is to tune that resonance to be higher than the available power, so no flutter.

Wayne

 

[KurtinSA]

This really helps on flutter prone overpowered clocks.

I think that is a contradiction!! Overpowered 400-day clocks!!

Kurt

 

[Schatznut]

I think that is a contradiction!! Overpowered 400-day clocks!!

Kurt

Roger that...

 

[Wayne A]

I think that is a contradiction!! Overpowered 400-day clocks!!

Kurt

Lol, have a few that meets that description!

Wayne