English Longcase rocking automation

[digitalpan]

Hi everyone

Grandparent duties over Christmas and new year have prevented any progress on a William Harris of Chippenham long case clock until this week. It's in for a complete strip and clean, with new bushes where needed plus attention to the date mechanism as reported in a previous post. But it's the rocking figure in the arch cutout that is making me ponder.

Here's the dial, the figure rocks in the cutout in front of a nicely painted starry sky.



I had not realised until I dismantled it how "top-heavy" the rocking figure makes the pallets and crutch.
The rocking figure of Old Father Time is connected directly to the pallet arbor by a bent rod:



The brass connection piece is a tight push fit on the pallet arbor so it can be adjusted for position.
Here's a view from above of the pallet arbor in the frames:



The weight of the figure and rod is far greater than the pallet arbor and crutch so it makes the whole thing "top-heavy". My immediate feeling is that this weight will cause all sorts of problems with setting the clock in beat and with the subsequent running. Am I right or just being silly? Is this setup correct or has it been modified? I can't see any signs of missing levers or attachment points. I've seen similar automation on other clocks but have not been able to examine the operating mechanism, nor can I find any photos or descriptions in any of my reference books.

Does anyone have any views on this, or can you point me to any reference?

Thanks.

 

[shutterbug]

It looks to me that the moving figure is original to the clock. The weight of the pendulum will counter any top heavy issues, but you might have to adjust timing with the figure as much as the crutch when the time comes.

 

[Ralph]

Other than essentially putting it in beat, by adjusting it so it's mass travels equally either side of center, there is not much you can do. Your pendulum will swamp the small amount of mass of the figure, vs it's mass.

It might require forming the rod, the figure is attached to, if the collet isn't a slip fit.

More than likely, the clock will run regardless of it's setting.

Ralph

 

[Willie X]

Nothing to do, it will work nicely as long as something isn't draging on the dial. I've never seen the "father time" motif. Usually it's a rocking ship.
Willie X

 

[shimmystep]

Nothing to do, it will work nicely as long as something isn't draging on the dial. I've never seen the "father time" motif. Usually it's a rocking ship.
Willie X

Agree with willie X.

 

[bangster]

You have, in effect, a "compound pendulum", with weight both below and above the pivot point. The upper weight affects the period of the assembly (makes it run a tiny bit slower than if it weren't there), but that was taken into account in the length of the lower pendulum.

It has no effect on beat. Set that as you normally would.

 

[shutterbug]

I'm not sure if weight added above the crutch would have the same effect as a compound pendulum. I'll have to ponder that one

 

[bangster]

See All About Pendulums

 

[BigAl]

Bangster, Thank you for the link to your pendulum information sheet and for taking the time to write it in the first place. I had not seen it before. I have bookmarked the page; it is now my priority clock reading. Previously I had looked on the net as well as other posts in this MB but found them far to technical for my simple, and old, brain. You get as many stars as I can give for this one.


Cheers


Alan

 

[shutterbug]

See All About Pendulums

Good information. But it doesn't answer my quandary about weight added to the top of a crutch

 

[bangster]

It's weight added above the suspension point. Doesn't matter where it's attached to the pendulum assembly. The crutch is part of the pendulum assembly. Technically, the mass of the crutch has to get into the equation. "It's always something...." --Gilda Radner

 

[harold bain]

Good information. But it doesn't answer my quandary about weight added to the top of a crutch

Lets just say it won't change the center of balance for the pendulum, so shouldn't have an effect on timekeeping.

 

[bangster]

It's weight added above the suspension point. Doesn't matter where it's attached to the pendulum assembly. The crutch is part of the pendulum assembly. Technically, the mass of the crutch has to get into the equation. "It's always something...." --Gilda Radner

I yam wrong (show embarrassed red face). The crutch AIN'T part of the pendulum assembly; it's part of the escapement. And I have no idea what adding mass above the anchor arbor will do to anything. The EW will have a bit of an easier time impulsing it, but somebody else will have to say what that means in terms of time keeping.

Yoda

 

[shutterbug]

Lets just say it won't change the center of balance for the pendulum, so shouldn't have an effect on timekeeping.

That's what I think too

 

[digitalpan]

Apologies for not replying earlier, due to a bit of a domestic crisis, and thanks for the interesting discussions especially to bangster for the reference to his piece on pendulums that I had not seen before.

I think I'm leaning toward the view that theoretically the rocking figure will turn the system into a compound pendulum, with weight above and below the pivot point but that in practice the weight of the rocking figure is so much lower than the pendulum that it shouldn't make any difference. To get a better idea I weighed the parts: pendulum bob 812g (28ozs), pendulum rod 106g (4ozs), rocking figure and support rod 18g (1/2oz), pallets and crutch 18g (1/2oz). So I guess I'm worrying about nothing!

What I will do, as suggested, is try to make sure the figure is centred over the pallets. I think I can do this by getting the clock running properly (level and in beat) without the figure, then to mark the position of the crutch with the pendulum hanging straight down. Then add the rocking figure, turn the movement upside down (no weights of course) and adjust the position of the figure until the crutch is in the marked position.

Then again I might just pop it on in roughly the correct place and see how it goes!

 

[Randy Beckett]

If you suspect, or experience, a problem, I suppose you could add a small lead weight around the crutch shaft, using it to balance the anchor assembly. The only influence on the pendulum then would be the impulse.

 

[bangster]

Not a compound pendulum. I wuz wrong (see post above). The figure isn't part of the pendulum; it's part of the escapement.

 

[harold bain]

If anything it will add some impulse to the pendulum in a positive way.

 

[tom427cid]

Interesting,I have set up a number of "rocking ship" movements and have experimented(nothing major) a bit.
I have found that the running characteristics seem to be improved if the addition-ship or in this case "father time"- is as close as possible to a straight line with the pendulum.
While this may not be by definition a compound pendulum the additional weight does have an effect. IMO the effect seems to be somewhat diminished amplitude of the pendulum-or put somewhat differently-it diminishes a bit of overswing.
Hope this might help-these are just observations. There might be a scientific explanation that someone else can explain.
tom

 

[shutterbug]

I think balance is the key. If it's working exactly in line with the anchor it should be an asset.

 

[Tinker Dwight]

It is, still, a compound pendulum but the typical pendulum, for this clock, will have
a bob heavy enough that it will dominate both the inertial mass and the center
of balance.
Removing the top piece will have the effect of speeding up the clock, requiring the
lowering of the bob, to compensate.
As was mentioned, beat setting will be best if it is aligned vertical with the
middle position of the anchors drops. This will help keep the gap at the
crutch from having to support too much off centered weight.
As I said, unlike many compound pendulums that use a light bob weight,
it should adjust normally in that the bob being lowered will cause the clock
to run slower.
Being compound, the weight of the bob has an effect on the rate, as well
as its position, does. A heavier bob will make is some slower.
Tinker Dwight

 

[harold bain]

Perhaps our poster might try it without the animation, then with it to see if there is a noticeable difference in timekeeping.

 

[shutterbug]

I have a hard time with that, Tinker. The crutch is not the pendulum and doesn't really have much effect on it. The opposite is more in line with what I observe.

 

[bangster]

It is, still, a compound pendulum but the typical pendulum, for this clock, will have
a bob heavy enough that it will dominate both the inertial mass and the center
of balance.
Removing the top piece will have the effect of speeding up the clock, requiring the
lowering of the bob, to compensate.
As was mentioned, beat setting will be best if it is aligned vertical with the
middle position of the anchors drops. This will help keep the gap at the
crutch from having to support too much off centered weight.
As I said, unlike many compound pendulums that use a light bob weight,
it should adjust normally in that the bob being lowered will cause the clock
to run slower.
Being compound, the weight of the bob has an effect on the rate, as well
as its position, does. A heavier bob will make is some slower.
Tinker Dwight

Tink, I don't understand how it can be compound pendulum, unless we count the anchor, anchor arbor, and crutch as part of the pendulum assembly. Are those also factors in determining the behavior of the pendulum?

 

[Tinker Dwight]

I have a hard time with that, Tinker. The crutch is not the pendulum and doesn't really have much effect on it. The opposite is more in line with what I observe.

Setting the beat will balance the anchor relative to the crutch and pendulum.
It will then be set correctly to the pendulum.
Since the beat setting usually adjust the crutch on the arbor, it makes sense
to set it relative to the anchor.
If the anchor slips on the arbor then it would make sense to align it to the pendulum
but still one wants it balanced over the anchor when done.
Tinker Dwight

 

[Tinker Dwight]

Tink, I don't understand how it can be compound pendulum, unless we count the anchor, anchor arbor, and crutch as part of the pendulum assembly. Are those also factors in determining the behavior of the pendulum?

To the dynamics of the pendulum, any swinging mass is part of the pendulum system and
effects the operation. Other than a possible small gap between the crutch and the
pendulum, the effects of its mass makes little difference as to where it is mounted on the pendulum
or on the other end of the anchors arbor.
It will still effects the center of gravity of the pendulum system, the same as a classical compound pendulums.
It will still effects the moment of inertial of the pendulum system, the same as a classical compound pendulums.
I think that qualifies it as a compound pendulum. It is not a traditional one but still acts exactly like one.
Yes, all the parts have some effect. The crutch is usually light and the anchor is real
close to the arbor. The effects are small but yes they are all part of the pendulum system
and, if doing an exact calculation of the system, need to be included.
The fact that this piece has some distance from the arbor and significant weight
makes it a significant part of the system.
Tinker Dwight

 

[Dave Diel]

To the dynamics of the pendulum, any swinging mass is part of the pendulum system and
effects the operation. Other than a possible small gap between the crutch and the
pendulum, the effects of its mass makes little difference as to where it is mounted on the pendulum
or on the other end of the anchors arbor.
It will still effects the center of gravity of the pendulum system, the same as a classical compound pendulums.
It will still effects the moment of inertial of the pendulum system, the same as a classical compound pendulums.
I think that qualifies it as a compound pendulum. It is not a traditional one but still acts exactly like one.
Yes, all the parts have some effect. The crutch is usually light and the anchor is real
close to the arbor. The effects are small but yes they are all part of the pendulum system
and, if doing an exact calculation of the system, need to be included.
The fact that this piece has some distance from the arbor and significant weight
makes it a significant part of the system.
Tinker Dwight

The interesting thing about Tinker's post's is they don't make sense the first time you read them. Then, you re-read them and they kind of make sense. By the third time it's beginning to soak in. The fourth time you realize how smart he really is.

 

[bangster]

Not bring really smart, I need some more explainin' before I get it. I regard Tink as our resident expert on pendulums, and I'll never say he's wrong about anything in that domain. I just say, I don't get it. I prefer to understand something before I accept it.

The crutch is a foreign entity with respect to the pendulum proper. The pendulum bangs into it periodically as it swings. Anything the pendulum bangs into (or rubs on) during its swing is going to affect its operation. But that doesn't make it part of the pendulum, does it? It's an external force working on the pendulum.

I think there's some fudging on the notion of "the pendulum system." Tink defines it as the pendulum, the crutch, the anchor, and the arbor. And then everything he says is clear enough.

But Tink's "pendulum system" isn't a pendulum. Its a composite of a pendulum and several other parts. Those other parts are going to affect the behavior of the pendulum, and the running of the clock, just like any external force will. But they don't alter the center of oscillation/center of gravity of the pendulum. Rather, they nullify it to a certain extent.

In a classical compound pendulum (mass above as well as below the suspension point) the center of gravity (and center of oscillation) is determined by the mass of the three elements and the distance of the weights from the suspension point. You "affect the center of gravity" by moving mass with respect to the suspension point. The change is internal to the pendulum. Nothing outside is working on it except gravity

Tink says substantially the same thing about his complex "pendulum system". And no doubt he's right. But being analogous to a compound pendulum isn't the same as being a compound pendulum. The extra parts of his "pendulum system" affect the center of gravity of the system; but not the center of gravity of the pendulum. They don't somehow turn it in to a compound pendulum

Is what I think.

 

[Ralph]

Bangster,

I tink you better tink again.

Ralph

 

[bangster]

Bangster,

I tink you better tink again.

Ralph

Thanks, but I'm waiting for Tink himself to straighten me out. He's the guy that get's it; I'm the guy that wants to get it.

 

[shutterbug]

So far, I'm with Bangster on this. A compound pendulum is a pendulum with forces added to the upper part to compensate for inadequacies in the lower part. Usually they are designed to slow down a pendulum that is in a case too short to allow adjustment in the normal way. In the case in question, the crutch is affected by a weight added above it. In no way does it affect the pendulum ... just the crutch. It might have a VERY MINIMAL affect on the operation of the pendulum, but certainly not enough to cause any issues. Look at all of the cuckoo clocks that run a saw from the crutch. They are about the weakest clock mechanism you could encounter, and the pendulums are among the poorest out there for construction ... and yet they saw away, hour by hour with negligible effect on the escapement. There's something here that is just not adding up. Like Bangster, I respect Tinkers knowledge a lot. But this one has me scratching my head.

 

[Ralph]

Look at it as a pendulum system. Everything connected to the pendulum is part of it, even if loosely linked, as the crutch.

So the rocker appendage, the arbor, the anchor, the crutch, and the pendulum with it's suspension, make up the pendulum system.

Ralph

 

[bangster]

Look at it as a pendulum system. Everything connected to the pendulum is part of it, even if loosely linked, as the crutch.

So the rocker appendage, the arbor, the anchor, the crutch, and the pendulum with it's suspension, make up the pendulum system.

Ralph

That's what I find suspicious. Everything connected to the pendulum is part of a system that includes the pendulum; it's not part of the pendulum.

Is what I think.

We may be arguing just about words. The timing of a clock isn't regulated solely by its pendulum; it's regulated by the entire escapement, which Tinker and Ralph like to call "the pendulum system". But the pendulum, whether simple or compound, is only an element in that system. The rest of the elements are not parts of the pendulum, and no matter what other machinery surrounds the pendulum, the pendulum will have its own physics, which it may not follow because the other stuff interferes with it.

Is still what I think.

 

[Dave Diel]

I find limit analysis to be useful in cases like this where the effect of a component on a system is not obvious. What would happen if the mass attached to the anchor arbor was increased to 1/2 the mass of the pendulum bob and placed 1 meter above the anchor?

 

[bangster]

I don't think the EW would be able to drive it.

 

[bangster]

Hey Dwight, where are you hiding? Come help me out here.

 

[Ralph]

I don't think the EW would be able to drive it.

Why?

If it remains a pendulum under the influence of gravity, as long as the power to the escapement is enough to replenish losses due to friction , it should continue working.

The period will change as the physical characteristics dictate. The compound effect would be more apparent.

IMHO, Ralph

 

[bangster]

The mechanism we're talking about is this: the anchor arbor, with its pallets, a crutch hanging down a few inches from one end, and sticking straight up from the crutch, a meter-long rod with a substantial weight on its upper end.

Ignore the pendulum of the clock. It has no role here. The mechanism I just described is itself a crude compound pendulum: a pivoted unit with weight both above and below the pivot (and the pallets off to one side, which I'm going to ignore).

This mechanism is way overbalanced on its upper half. On the metronome principle, IF it could wag by itself, it would wag very slowly. But I daresay it could not wag by itself. Push the meter-long upper weight to one side, and the mass of the crutch ain't gonna be sufficient to reverse its swing. The only thing that could make it swing back is an impulse from the EW. And at the end of that swing, it won't return by itself. It'll need another push from the EW. It's not self-sustaining, even for a wag or two.

The EW is at an enormous leverage disadvantage when it comes to driving this lopsided mechanism. And with a normal power source from a mainspring, I think it would lose the battle. The EW tooth would arrive at the pallet, go "UNNGH", and stop. It couldn't get that meter-long extension to wag the other way.

Is what I think.

Tink, where the heck are you?

 

[Dave Diel]

The EW is at an enormous leverage disadvantage when it comes to driving this lopsided mechanism. And with a normal power source from a mainspring, I think it would lose the battle. The EW tooth would arrive at the pallet, go "UNNGH", and stop. It couldn't get that meter-long extension to wag the other way.

I agree that it will not work without the pendulum, but with the pendulum the escape will work normally.

If the clock is designed so that the point of inflection of the suspension spring is right at the center of the anchor arbor, you have two separate pendulums swinging on a common axis that happen to be connected by the pin on the crutch that goes through the pendulum leader. They will behave as a single complex pendulum.

 

[bangster]

I think you've got me convinced. This setup doesn't change the center of gravity of the clock pendulum, as I thought had been suggested. But that's no longer relevant, if we think of this whole contraption --clock pendulum, crutch, upper extension-- as a single unit, with a bit of wiggle in the middle where crutch meets pendulum.

That whole unit will constitute a compound pendulum, and the relevant factor will be the center of gravity of the whole unit, not that of the clock pendulum. Maybe that's what Tink was saying all along, and I didn't get it. So now I Get It. I see, even without the compounding factor of the upper extension, why the mass of the crutch enters into the equation.

Thanks, dad.
(And thanks, Tink)

 

[shutterbug]

I suppose you COULD say you have two pendulum contraptions working together, making it in effect a compound pendulum .... however I think you'd be wrong. The scenario of an ultra long appendage above the crutch is just an unbalanced crutch that has no use in a clock. In order for it to function at all, you'd have to have a very heavy pendulum to overcome the added resistance, and a heavier weight or stronger spring to run the pendulum. The pendulum would not likely have to be longer, just bigger. The whole setup would be counter productive. The motion in question has the same off balanced anchor that is overcome by a pendulum big enough to compensate for it. It is light enough though, to make that compensation doable.

 

[bangster]

No, I think it would work. The upper bob is specified to have 1/2 the mass of the pendulum bob. Suppose the clock pendulum is 18" long, half that of the upper extension. Without getting into equations, I imagine the two would pretty much balance out.

As I said before, the overbalanced crutch by itself would never wag; the EW impulse couldn't overcome the leverage of the upper extension. But when we add in the clock pendulum, the situation changes. If you push the upper extension to one side, it won't be just the EW trying to reverse its swing; the crutch would push the clock pendulum to the opposite side, and gravity will be helping as well. With that help, the swing would reverse and the entire contraption would wag. True, you'd need enough mass and length in the clock pendulum to provide the additional gravitational push. But given that, it would run.

Is what I think.

 

[shutterbug]

OK, banster. Lets say it runs. The bigger question is: will it run substantially slower. My guess is that it's not really affecting the anchor like a compound pendulum would, even though it's putting a huge drag on the escapement. We once had a discussion on crutch length. The consensus was that it makes no difference at all to the escapement.

 

[harold bain]

OK, banster. Lets say it runs. The bigger question is: will it run substantially slower. My guess is that it's not really affecting the anchor like a compound pendulum would, even though it's putting a huge drag on the escapement. We once had a discussion on crutch length. The consensus was that it makes no difference at all to the escapement.

My experience on crutch length is the opposite, SB. Shorter crutch, faster clock. https://mb.nawcc.org/showthread.php?1056-Crutch-height-and-effect-on-regulation

But, this animation really has nothing to do with changing the crutch length. If anything it should increase the pendulum swing.

 

[shutterbug]

Thanks Harold. I remember that now, and still wonder about your findings. I do agree with your idea that the pendulum swing would increase with the scenario postulated. However, I wonder how significant that increase would be on the time keeping.

 

[bangster]

My experience on crutch length is the opposite, SB. Shorter crutch, faster clock. https://mb.nawcc.org/showthread.php?1056-Crutch-height-and-effect-on-regulation

But, this animation really has nothing to do with changing the crutch length. If anything it should increase the pendulum swing.

For any clock, when we get the mass of the crutch into the equation, the "pendulum" is now the clock pendulum + the crutch. So it's as if the clock pendulum had been top-loaded by the amount of the crutch. The shorter the crutch, the higher the center of gravity of the "pendulum" and the faster the clock.

Is how I see it.

Back to the example with the upper extension. The pendulum assembly = clock pendulum + crutch + upper extension. The pendulum assembly is functioning as a compound pendulum, suspended at the arbor. Like any compound pendulum, its effect on the rate of the clock will depend on the effective length of the upper extension. Longer upper extension = slower clock.

Is how I see it.

So in the OP's clock, the little figurine will make the clock run slightly slower than it would have run without the figurine. Easily dealt with by the rating nut.

Is how I see it.

Tinker, WHERE ARE YOU? dammit!

 

[Tinker Dwight]

Hi
I'm back.
The weight above the anchor is moving with the pendulum.
It is part of the system. If you don't like calling it a compound pendulum
you have to understand it has the same effect on the pendulum speed.
Two things happen when there is a mass added above the main pendulum.
One is that the increase in mass slows it because of increase inertia.
The place where gravity pulls moves up the pendulum ( center of gravity )
as well. This means gravity has less leverage. Both changes are the same
in that they both slow the pendulum.
More mass less leverage. Slow/Slow. Can't get around that.
Weight added below the fulcrum can be more complicated to analyze.
If the center of mass moves down or the moment of inertia is
reduced, it runs faster. Remember the mass added at the center was
making it run faster. The center of mass was moving down the rod faster
than the increased inertia was slowing it.
There is an effect that is the increased friction of the anchors arbor. I'm sure
this was compensated for by the design of the clock.
When there is more mass added above the fulcrum of the pendulum as in this
case, it has, in all cases, been called a compound pendulum.
The hole system is a mass that is just the sum of the distances squared and the mass,
while the center of gravity has been moved up. The only distinction is that the crutch
may have some play so that for a tiny amount of the swing, the two pieces, are separate.
For most of the swing, it is like they are part of the same rod.
Tinker Dwight

 

[Dave Diel]

I stumbled around on the internet and found the equations for a "generalized" pendulum and attempted some calcs.

For a massless 1 meter long pendulum rod with a 1 kg bob, the period is 2.0 seconds, which is the time for the pendulum to swing from left to right and back to left.

If you add a 1/2 kg bob 1 meter above the pivot point, the period becomes 3.47 seconds.

 

[Tinker Dwight]

Yes Bang
You've got it.
Still you are thinking there is some magic center
of push thing.
I keep trying to explain, there are to components.
Center of Mass
and
Moment of Inertia
The lower the center of mass, the faster ( it means gravity has more leverage )
The farther away the masses are from the fulcrum the more moment of inertia, making it slower.
( as it is the square of the distance, it doesn't care where in a circular distance from
the fulcrum it is, just how far away it is ).
We have added moment of inertia ( you know, more mass harder to push or pull ).
We've moved the center of mass up so gravity has less leverage.
Only center of mass cares about where around the fulcrum it is.
Moment of inertia only cares about how far away it is from the
fulcrum.
Both are slowing it down.

 

[digitalpan]

Hi everyone, OP here.

What a fascinating discussion! I can now report that the clock does indeed run slower with the figure attached.

I set it up without the figure and measured it with the only tool at my disposal, the ClockMaster app for iPhone. (Aside: please don't rush to tell me how poor/inaccurate/misleading the app is. I use it to create a data file of the tick-tock timings which I download to a PC for analysis. That way I can examine the raw data for spurious values and remove them. I will try to post a critique of the app very soon).
Then I added the figure, balanced it as best as I could, and measured again. The difference was approximately 0.06secs per revolution of the escape wheel (0.06 secs/minute), which is nearly a minute and a half a day - more than I expected, but easy to rectify. I still need to do some more work on the numbers to see whether the figure is influencing the smoothness of motion of the escape wheel - it already shows at least one tooth is slightly bent and/or clipped. But that's the sort of thing I ought to cover separately.