Adjusting a pendulum clock timekeeping
Thread Starter
Join Date: Sep 2004
Location: Berkshire, UK
Posts: 768
Adjusting a pendulum clock timekeeping
I recently brought an eight day, hours and halfs striking pendulum clock back into use. I had bought it from a junk sale 20 years ago for a surprisingly small sum because there was no key or pendulum. I stripped it down, cleaned, reassembled and lubricated it and then went in search of a key and temporary pendulum. A key was the easy bit. The pendulum is a two part affair with the upper part of the rod from the suspension down through the crutch ending at that point with a hook, the lower part comprising the remainder of the rod and the weight having been unhooked for transporting and subsequently lost. I found a quick substitute in the form of one of those supermarket trolley tokens with its short chain and quick release catch which "looked" about the "right length" for the job. I slipped it onto the hook and set it swinging and the clock ran for some 5 days gaining about 6 minutes an hour. One the 5th day we went away for a week, the clock had stopped when we got back.
I do understand the logic behind pendulum weight and length and timing. As the bob on the end of my temporary pendulum is not big in relation to the mass of the rod I decided to add a second trolley token onto the catch which has increased the effective length of the temporary pendulum. After winding and restarting the clock it then ran for 10 days during which time it gained 3 minutes an hour over the first 5 days and then lost about 1 minute an hour until I put it right when the clocks went back. The timings are approximate, it doesn't suddenly switch from gaining 3 mins to gaining 1 min but this is the best I can see as we are not always here.
A further observation. The upper half of the pendulum rod is a thin length of brass strip with a hook cut into the lower end, it is suspended perpendicular to the clock plates, ie the hook faces the rear of the clock. The lower half should have a slot which is a snug fit over the upper rod and a pin through it would engage with the hook to give a joint which has little play in the direction of swing. My temporary lower pendulum rod has a short length of chain with a ring which engages the hook and the coupling is flexible in the direction of swing, When the pendulum takes impulse from the recoil escapement it is likely that the joint in the pendulum rod pivots a bit one way and when the pendulum drives the recoil it pivots the other way. At other times it will be straight ish due to the weight and free swinging action.
Questions.
If I make (I will eventually!) a new lower pendulum section and bob which has a snug, non flexing, joint at the hook, will the non linear time keeping be reduced?
Is there a way to improve the non linear timekeeping by adjusting something else? I don't know where to start with the pallets and escape teeth, they don't appear to have suffered noticeable wear. The clock is somewhat older than me but I don't know how long it was in service before I bought it.
An internet search has pulled up much information about adjusting the pendulum to be in beat but little else.
Rans.......
I do understand the logic behind pendulum weight and length and timing. As the bob on the end of my temporary pendulum is not big in relation to the mass of the rod I decided to add a second trolley token onto the catch which has increased the effective length of the temporary pendulum. After winding and restarting the clock it then ran for 10 days during which time it gained 3 minutes an hour over the first 5 days and then lost about 1 minute an hour until I put it right when the clocks went back. The timings are approximate, it doesn't suddenly switch from gaining 3 mins to gaining 1 min but this is the best I can see as we are not always here.
A further observation. The upper half of the pendulum rod is a thin length of brass strip with a hook cut into the lower end, it is suspended perpendicular to the clock plates, ie the hook faces the rear of the clock. The lower half should have a slot which is a snug fit over the upper rod and a pin through it would engage with the hook to give a joint which has little play in the direction of swing. My temporary lower pendulum rod has a short length of chain with a ring which engages the hook and the coupling is flexible in the direction of swing, When the pendulum takes impulse from the recoil escapement it is likely that the joint in the pendulum rod pivots a bit one way and when the pendulum drives the recoil it pivots the other way. At other times it will be straight ish due to the weight and free swinging action.
Questions.
If I make (I will eventually!) a new lower pendulum section and bob which has a snug, non flexing, joint at the hook, will the non linear time keeping be reduced?
Is there a way to improve the non linear timekeeping by adjusting something else? I don't know where to start with the pallets and escape teeth, they don't appear to have suffered noticeable wear. The clock is somewhat older than me but I don't know how long it was in service before I bought it.
An internet search has pulled up much information about adjusting the pendulum to be in beat but little else.
Rans.......
Join Date: Nov 1999
Location: UK
Posts: 1,985
I think some clocks have a device to counter pendulum length changes with changing room temperature?
An acquaintance of mine who worked on church clocks had an electronic device you clip onto the pendulum to measure its timing. The acquaintance has since passed away, and I don't know what the device is called.
I suspect you will have to research clockmakers forums to find your answer.
.
An acquaintance of mine who worked on church clocks had an electronic device you clip onto the pendulum to measure its timing. The acquaintance has since passed away, and I don't know what the device is called.
I suspect you will have to research clockmakers forums to find your answer.
.
Last edited by Uplinker; 15th Nov 2021 at 08:52. Reason: Typos and nonsense.
Thread Starter
Join Date: Sep 2004
Location: Berkshire, UK
Posts: 768
In an ideal or theoretical pendulum ALL of the weight is concentrated in the bob. Under these conditions the actual weight makes no difference to the time period. In the real world the pendulum is a combination of the weight of the bob and the weight of the rod so its effective length is the centre of gravity of the assembly measured down from the suspension. If a small weight is added to the bob, as in the penny added to the Big Ben clock, it increases the effective length by lowering the centre of gravity of the assembly and slows the rate a tad. This is why what I did makes my clock run slower.
Rans6................
Rans6................
Join Date: Feb 2000
Location: UK
Posts: 1,536
Can I suggest the following link as an excellent place to join and pose your query. https://www.model-engineer.co.uk/
many members have built and restored clocks and other time-pieces and are quite generous with advice.
many members have built and restored clocks and other time-pieces and are quite generous with advice.
Paid...Persona Grata
Join Date: Aug 2004
Location: Between BHX and EMA
Age: 76
Posts: 238
I think some clocks have a device to counter pendulum length changes with changing room temperature?
Steel is better than brass for a pendulum as it has a lower co-efficient of expansion.
Join Date: Oct 2004
Location: California
Posts: 356
I use an iOS app called Clockmaster for setting my pendulum length. I've got two clocks next to one another. One runs at +/-0 seconds per month (approximately). The other at -15 seconds/week. I can't get it closer because I need to turn the clock around to set the pendulum, and it changes speed slightly when I turn it back.

Join Date: Apr 1998
Location: Mesopotamos
Posts: 1
Great observation there SRFred, thanks. I guess if the pendulum clock was not firmly planted onto the ground then a similar variance of time could result.
Paid...Persona Grata
Join Date: Aug 2004
Location: Between BHX and EMA
Age: 76
Posts: 238
Have a family one that is over a hundred years old and keeps fair time but one thing that was pointed out was it has to be level. A slight amount off level and the time varies.
It had worked fine for a hundred years or so on a stone floor, but the carpet had enough give in it to allow this to happen. The cure was a screw through the back of the case into the wall.
Join Date: Apr 2008
Location: Out in the sticks in DE56
Age: 84
Posts: 559
when the weight dropped to the same level as the pendulum, it would experience a sympathetic oscillation, to the extent that it made the clock sway in time with it, reducing the pendulums effective swing to the point that it wasn't swinging far enough to trip the escapement.
That's fascinating. How long did it take for your friend to puzzle it out? Bet it was a Eureka! moment!
Join Date: Oct 2007
Location: Here
Posts: 865
Join Date: Aug 2010
Location: Malaysia
Posts: 143
A bit of drift... the book Longitude by Dava Sobel covers John Harrison's quest for perfect timekeeping. Amazing that his clocks only lost a few second on a voyage to the Caribbean in the 18th Century. Recommended if you are into horology and/or navigation.
Anilv
Anilv