Pendulum Length

It is possible the carpet is permitting a little bit of sway to he top of the clock whereas the bare floor isn't. What happens if the carpet goes back under the clock?

It's hard to tell how the mechanism works from the photos, but it looks like there is a pivot point on the underside of the brass piece that holds the pendulum. It looks like the brass piece with the 3 rounded grooves pivots on that spindle. It may be that by moving the pendulum to the outermost (deeper) groove there is more friction between the brass support and the spindle on which it rotates. More friction would cause the pendulum to swing at a slower speed, thus making the clock run slower - in this case, at normal speed.

When you placed the pendulum on the innermost groove there was less friction between the brass piece and the spindle, so it ran faster.

The brass piece with the 3 rounded grooves is fixed. The visible pivot supports a shaft, on one end of which is the anchor escapement. On its other end the L-shaped lever (crutch) is fixed, taking up the movement of the pendulum and transmitting it to the escapement.

Twenty minutes fast in 24 hours means the period is about .986 the correct time. Pendulum period is proportional to square root of length, so if the problem were due to pendulum length, the length would be .9724 the correct length, or about 1/4 inch short for every 10 inches of pendulum length. You are right, the notch difference is far too small.

The only other thing I can guess is that there is a little flex in the pendulum support bearing and moving the weight of the pendulum closer to the bearing is affecting the escapement, allowing it to occasionally "skip a notch".

If it's a grid iron pendulum move the outer support rods to the middle position.

My WAG is the outer rods are over compensating for temperature.

The gridiron pendulum had been invented when this clock was made, but this is a standard (I will not say simple) pendulum for a long-case clock. It is suspended from a fixed point, with a short length of spring steel providing the flexibility to swing, with a period of 2 seconds.

It looks like from the picture that the inner most notch that the pendulum hangs from would closely align the pendulum to the L-shaped mechanism possibly reducing resistance of the mechanical movement....My guess as is others the carpet had some effect, possibly some slight wobble or angle....

Could you repeat the experiment? Move the pendulum back to the outside notch and see if the clock again runs fast. More than once have I been puzzled by something that made no sense only to finally retrace my steps and find that something else had happened.

PWSlack had a good idea. If you have a piece of carpet, it might be interesting to see the effect of the clock sitting on carpet. It's my guess that it would tend to run faster on carpet because the pendulum would swing about a point below the pivot if the rest of the clock could sway in the opposite direction of the pendulum.

If it doesn't make sense, do some more experiments. Sooner or later it's got to reveal the secret.

I have moved the pendulum back to the first position and will report on the timekeeping. Regrettably I shall leave the carpet experiment alone. Owners of old clocks have enough hassle setting them up in a stable position in the first place.

Methinks the 3 pivot locations provide 3 options for coarse height adjustment and the finer adjustment is the vernier screw at the bob. Similar to this Whitehurst one below which only has 2 locations available.

Is it likely that in the old location before the clock was moved that the pendulum support being used is not the same one used after relocation ?

I have tried the first experiment, moving the pendulum back to the notch nearest the clock face (left in my original photo). At the same time I had to bend the L-shape so that the brass block did not bind in the fork. The clock ran faster, though not as fast as originally.

Try doing this. Hang a string from the the brass block with the 3 notches. Then tie something not too heavy to the other end of the string. See how plumb it is.

Gravity varies with underground rock formations, substrate densities, and water deposits. Also elevation is a factor. You may be in a "twilight zone" for gravitation.

Altitude changes the air pressure and density so the clock could run faster.

That triple notched suspension point is definitely unique. Do you have the name of the movement maker? I suspect that if you put a straight edge across the top and measured the depth of the notch they would follow some fixed ratio like 1:2:4. The intent is not to alter the distance from the suspension point to the actuating lever (the crutch) but to alter the overall length of pendulum. Even though the difference is minimal it should be enough to affect the period. It could also be there to adjust for large seasonal changes in the ambient temperature since they will affect the period as the average air density changes, and all the owner had to do was shift its position without retiming the bob with the adjusting screw.

What does the pendulum weight (the bob) look like and is it the original? How many days will it run between windings?

The maker's name is Whitehurst of Derby (UK), a very well respected family of clockmakers. John Whitehurst senior lived 1713-1786 and had Benjamin Franklin, Erasmus Darwin and Josiah Wedgwood as friends. I suspect, however, that this clock was made by son or grandson. It runs for 8 days between windings. The case is relatively crude by comparison with the quality of the movement.

Attached is a picture of the pendulum bob (at present back to front).

Are you available to broker my next car purchase? I like your attention to detail!

Thanks Wayne. I am intrigued by watch and clock movements. If you ever get the opportunity to study movements with complications (repeaters, day/date, moonphase, chronographs, etc) you'll get an appreciation for how the watchmakers kept the regular watch operating accurately, while at the same time the complication drew power from the mainspring. Some had a separate mainspring (I have a 1950's Cricket Alarm watch that has two mainsprings) which the complication ran off of, but most didn't.

Just a point of clarification: had you definitely been using the notch nearest the clock face in the old house?

That I cannot say. The point remains, however, that a minimal change in the height of the suspension point results in a substantial change in timekeeping. I am grateful to Autobroker #10 for his comment.

you could try adding to, or removing a bit of the weight of the pendulum!

I believe that the pendular principle is that the position of the centre of gravity is important, but not the actual weight. In any case I'm not going to interfere with a working part of an antique. I'm coming round to the idea at which Autobroker and others have hinted, that the spring steel portion of the pendulum may be as important as the pendulum length

I am pretty sure the answer lies in location of application of driving power to the pendulum.

Left to it's own devices, the period of the pendulum swing depends solely on gravity and the length of the pendulum. However, as soon as you apply force to drive the pendulum and extract force to pull the mechanism free of the escapement wheel, you start affecting the force balance, velocities and timing of the pendulum.

I see that there is a brass block of sorts that engages the fork that both pushes and pulls on the pendulum. The point at which the fork presses on the block and the force applied has an affect on the pendulum period and shortening or lengthening the distance between the driving force and the top of the pendulum has a significant effect on the period.

Astronomical clocks of the period commonly used a different escapement that would push the pendulum for 1 swing and would then disengage for the next several swings so that the pendulum would swing with a more consistent and "natural" average period. During the same period, a lot of effort also went into making the driving force consistent, such as verge-fusee type watches where the spring drove a case with a snailshell sort of track on top with a drive cable or chain running on the track. This compensated for the change in spring force between fully wound and almost unwound.

In short, the notches on the pendulum support change the point of application of lateral forces on the pendulum. The change in moment arms affects the period of the pendulum when under power.

BTW. I have had situations where clocks moved to a hard floor from carpet will interact more than they used to. The swing of the pendulum wants to make the clock sway which puts a stress on the floor. The floor structure can transmit the movement to the base of a nearby clock, causing it to sway minutely and affect the pendulum swing. You can get some weird things like one clock loses swing amplitude while the clock next to it develops overswing and starts banging the case with the pendulum. In that case, both clocks were on a floor where the joists ran perpendicular to the wall and the floor flexed like a standing sine wave. You could stop and restart one clock out of sync, but eventually they would sync in and something strange would happen. The fix was to move one of the clocks a little bit and add a bunch of boxes of shotgun ammunition into the bottom of the clock case.

shortening or lengthening the distance between the driving force and the top of the pendulum has a significant effect on the period. I think this is the conclusion we are coming to.

During the same period, a lot of effort also went into making the driving force consistent, such as verge-fusee type watches where the spring drove a case with a snailshell sort of track on top with a drive cable or chain running on the track. This compensated for the change in spring force between fully wound and almost unwound. That is true, and I have a couple of spring-driven clocks with fusees. However, this one is weight-driven, making the driving force constant.

As a digression, one of my other clocks has an Ansonia (New York) movement. The driving and striking trains have been combined into a single driving train and fitted into a ship's clock casing. The oddity for me is that both the movements have to be wound anticlockwise, a feature I have seen from no other clock manufacturer.

Thank you all for your contributions.

Pocket watches with fusee chains wind backward - counterclockwise. The fusee chain is like a tiny bicycle chain that winds (loads the spring) and unwinds to power the movement. Someone was smart enough to create a cone shaped piece that the chain wound on. As the spring unwound (and lost power) the chain would move down the cone and compensate for the difference in spring power.

When you have a chance, do a Google search on Abraham Brequet. He was a watchmaker in the late 1700 to early 1800's. He did some amazing things with pocket watches. A mechanical engineering genius.

The central puzzle is "why does changing the notch change the period of the pendulum". The difference of 20 minutes per 24 hours is about 1.4 percent. Two things come to mind, both related to friction.

As has been mentioned before (Usbport #2), the pendulum weight is further from the visible bearing increasing the force on that bearing (and the other bearing not visible) and could increase the friction, lengthening the period.

The second factor is that hanging on a different notch, the pendulum swings in a slightly different path. If it is swinging closer to back wall of the clock for example, perhaps the increased wind resistance could slow it down.

GA to you! I was trying to explain it with my description, but yours clears things up better.

If someone still doesn't see how this works, consider the following:

Let's make up an extreme case where the first notch is in front of the clock face (not possible, but we're hypothesizing), so the pendulum will then be at a 45 degree angle to the clock face. Let's call gravity a force in the Y axis and the escapement imparts a force in the X axis. The pendulum swings in the XY plane (left and right) and not in the Z axis (front to back). The moment arm of the pendulum in the XY plane will be root 2/2, because of the 45 degree angle. In the XY plane, the pendulum is only 0.707 as long, thus you've effectively shortened the pendulum which reduces the arc, which shortens the period and makes the clock run fast.

Aha, I think we have a misunderstanding here. The slot in the crutch (the L-shaped lever), is sufficiently long to allow the pendulum to hang vertically, whichever suspension notch is chosen.

I think you got it. Thanks, that makes sense. Lots of times there's no substitute for practical experience! GA from me.

This probably has nothing to do with the problem to hand.

But it is interesting.

32 Metronome Synchronization - YouTube

Perfect example of the two clock interaction.