Carbon fibre pendulum rod

[Bob StevensonParticipant @bobstevenson13909]

Brilliant thread that has demonstrated the strength of this site……..

 

Some points that might help in no particular order….

 

1) Invar has to be 'aged' to get the full CTE…it's not straightforward.

 

2)western red cedar has a reputed zero CTE and is not hydroscopic

 

3)bamboo has zero CTE but IS very hydroscopic….can be sealed with strong shellac etc

 

4) the best way of interpreting pendulum 'questions' is by evaluating their 'Q'…which is a measured calculation of their efficiency

Edited By Bob Stevenson on 10/08/2020 16:52:59

[Michael GilliganParticipant @michaelgilligan61133]

Recommended Reading

Matthys : Accurate Clock Pendulums

**LINK**

https://books.google.co.uk/books/about/Accurate_Clock_Pendulums.html?id=cJZBbsuCZRQC&printsec=frontcover&source=kp_read_button&redir_esc=y#v=onepage&q&f=false

MichaelG.

[Russell EberhardtParticipant @russelleberhardt48058]

The supplier of the pultruded CF tube I used for my regulator quoted -0.1 to -0.3 ppm/C. I used a threaded bottom fitting for the rod for use with a regulator nut at the bottom of the bob. I calculated that making the transition from CF to steel 30mm below the centre of the bob would compensate for 0.2 ppm/C for the rod with a small positive contribution from the suspension.

Seems to work fine.

Russell

[Former MemberParticipant @formermember32069]

[This posting has been removed]

[brickyParticipant @bricky]

My first clock was a regulater which as I had no tools was built with a home made turns and hand cut teeth.The pendulem was a length of straight grained deal with a heavy lead bob.When the clock was running I rated the pendulem with lead fishing weights and coins.The best I achieved was 2 seconds a day for a fortnight.I know that the post is about carbon fibre about which I know nothing,but the killer of all manual timepieces is friction ,reduce that and the difference between Invar and carbon fibre will hardly matter.

Frank

[John HaineParticipant @johnhaine32865]

Posted by Bob Stevenson on 10/08/2020 16:51:22:

………………………………………..

4) the best way of interpreting pendulum 'questions' is by evaluating their 'Q'…which is a measured calculation of their efficiency

Edited By Bob Stevenson on 10/08/2020 16:52:59

Hmm, well, up to a point! High (or low) Q doesn't help with CTE for example. High Q pendulums are more sensitive to horizontal motions of the whole clock, but less sensitive to internally generated noise (e.g. impulse variations). Very accurate clocks – Shortt and Fedchenko for example – have their master pendulums in vacuum for high Q (~100,000) and isolation from barometric variations; but Burgess' Clock B works in air, and has a Q of about 5,000 or less, but is probably the most accurate purely mechanical clock ever. On the whole I think high Q is to be preferred but it depends…and the Q topic is immensely divisive amongst horologists.

[Russell EberhardtParticipant @russelleberhardt48058]

I did measure the Q factor of my pendulum with all fittings. I can't find my notes just now but I think it came out at about 6000. Easy enough – set it going and count the number of cycles it takes to decay to 37% of it's starting amplitude. Multiply that number by 2 x Pi to get the Q. The most important factor is the air resistance so best to use lead or even uranium for the bob and an aerodynamic shape. I stuck with steel not having uranium to hand!

Russell

[Alan CrawleyParticipant @alancrawley27839]

I’m beginning to realise my IQ is too low for this conversation!!

Let’s try to simplify my original question- do I need any compensation for the Carbon Fibre rod?

My main interest is going into my little workshop and making nice bits. To be honest to me this engineer- designed regulator is running with amazing accuracy with common BMS rod.

This is not to say I am not grateful for the interest shown.

[Martin KyteParticipant @martinkyte99762]

For your purposes, and most others probably not.

regards Martin

[Alan CrawleyParticipant @alancrawley27839]

Edit:- comma after "to be honest"

[SillyOldDufferModerator @sillyoldduffer]

Posted by Alan Crawley on 11/08/2020 19:35:38:

…

Let’s try to simplify my original question- do I need any compensation for the Carbon Fibre rod?

My main interest is going into my little workshop and making nice bits. To be honest to me this engineer- designed regulator is running with amazing accuracy with common BMS rod.

…

Accuracy is the really big challenge in clockmaking. Every right to be pleased with a clock that runs and even more delighted if looks well-made. Unfortunately visitors will only be really impressed if it keeps time correctly. Experts know what to look for and a mechanical clock accurate to within a minute per year will enthuse anyone who knows about clocks.

By horological standards your clock can't be running with amazing accuracy because it's rate must vary with season, weather and the central heating due to the pendulum. Easy to do better than BMS by substituting a material with a low temperature coefficient of expansion: wood, Invar or Carbon Fibre. And maybe even better by also fitting a compensator. Probably not worth compensating a CF pendulum because other mechanical  shortcomings are likely to swamp temperature errors.

They way to find out is to monitor the clock's timekeeping over a long period. Does the rate of gain or loss vary with temperature? Worth fixing if the link is obvious.

Dave

 

 

Edited By SillyOldDuffer on 11/08/2020 21:20:27

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Alan Crawley on 11/08/2020 19:35:38:

[…]

Let’s try to simplify my original question- do I need any compensation for the Carbon Fibre rod?

.

That would depend upon what performance you are seeking

Seconds per day … probably not

Tens of seconds per year … almost certainly

MichaelG.

[Russell EberhardtParticipant @russelleberhardt48058]

Posted by Alan Crawley on 11/08/2020 19:35:38:

I’m beginning to realise my IQ is too low for this conversation!!

Surely not. Sorry if I have assumed too much.

To simplify things; Q is just a measure of how much energy is lost on each cycle of the pendulum and how much energy must be given back to it on each tick of the clock to maintain it. The higher the Q the smaller the energy is needed to keep it going and the less disturbance that it causes to the regularity of the clock.

Hope that helps.

Russell

[Alan CrawleyParticipant @alancrawley27839]

I fitted the carbon fibre rod as advised and all is well!

I recently used the term "amazingly accurate" when running on a length of 6mm EN3B. This was meant to be a relative term as it is amazing to me, my first ever clock in my ninth decade. It cost well over £300 for materials, and is not as accurate as a £4 clock from Aldi!

Another question, I learned that the term "Regulator" was used as it was the accurate clock in the clockmaker/repairers workshop to adjust other projects. What did they set the Regulator to? Or did it set standard time for the area it was in? I recall reading that railways had problems where different companies ran to different time. Much like today, it seems!

Many thanks for the help and guidance.

[Martin KyteParticipant @martinkyte99762]

Posted by Alan Crawley on 25/08/2020 14:09:49:

Another question, I learned that the term "Regulator" was used as it was the accurate clock in the clockmaker/repairers workshop to adjust other projects. What did they set the Regulator to? Or did it set standard time for the area it was in? I recall reading that railways had problems where different companies ran to different time. Much like today, it seems!

The regulator was important not so much for aboslute time but for 'rate'. If you have a clock that runs at a consistant rate say gains two seconds per day you have a time standard available to 'regulate'other time pieces particularly is you are attempting to 'prove' their temperature compensation. Absolute time really can only be determined by astronomical observation. In essence it is defined as such.

Regarding the railways, the need to produce timetables created the need to establish a counry wide 'universal' time rather than having each parish set its local time by sundial or local noon. That way you didn't have Bristol 20 mins different from London.

I am told that French railways as late as 1970 relied on large sundials to set the station clocks which could be achieved to within a few seconds. Many early longcase clocks were also delivered with a sundial for setting (how else would you do it) and some even showed sundial time on a separate dial to make for easier setting.

regards Martin

[SillyOldDufferModerator @sillyoldduffer]

Posted by Alan Crawley on 25/08/2020 14:09:49:

I fitted the carbon fibre rod as advised and all is well!

…

Another question, I learned that the term "Regulator" was used as it was the accurate clock in the clockmaker/repairers workshop to adjust other projects. What did they set the Regulator to? Or did it set standard time for the area it was in? I recall reading that railways had problems where different companies ran to different time. Much like today, it seems!

Many thanks for the help and guidance.

Congratulations, always good to hear of successes.

Standard time was originally local, ie Noon. So regulator clocks would be checked regularly against a superior sun-dial. Problem with this is Noon depends on longitude, making Bristol time about 10 minutes later than London. No good for marine navigation or railways. A 10 minute navigational error is out by roughly 100 miles, and timetabling trains was a nightmare!

Next step was to measure time with a transit observatory rather than a sundial. A telescope aligned rigidly with true North-South, ie on a Meridian, was used to detect exactly when stars pass directly overhead. The movement of the heavens is extremely accurate and knowing the observatory's longitude meant its time relative to Greenwich was also precisely known. Astronomical time was used to set and constantly measure the rate of a good mechanical clock in the observatory. If observations were missed due to clouds, the mechanical clock could still be read accurately by correcting it's rate error. The observatory would also set and measure the rate of other clocks, some of which would travel as necessary to set local regulators, often owned by clockmakers.

When telegraphs became available, Greenwich used the system to broadcast time signals to all major towns, making it easier to keep local clocks in step. Then by radio and now by satellite.

Not all Regulators were regulators; the term seems to have come to mean any accurate clock in that style.

Fascinating subject.

Dave

[Tim StevensParticipant @timstevens64731]

two further points:

The time set by a sundial will include errors due to the wobble of the earth's axis as it rotates – about 15 minutes (from memory) either way so it would need to be a clever dial or a set of look-up tables to get a universal noon (rather than a seasonal one). I think the term for the wobble is precession – if not I am sure to be bombarded with different corrections.

in the 1850-1860 period, as railways spread around the UK, lots of towns were persuaded to erect clock towers, so that locals could set their home clocks and so, catch the trains on time. In mid Wales, places like Rhyader, Dolgelly, Knighton, Kington etc still have these clocks. But the railways are not all there now – shame really.

Regards, Tim

[Martin KyteParticipant @martinkyte99762]

Hi Tim

time set by a sundial will include errors due to the wobble of the earth's axis

Depends on what you consider to be the more fundemental time reference. I would argue that time defined by the position of a location on the earth in it's orbit around the sun is what we define by 'the time'. Unfortunately the orbit is not perfectly circular and the axis has a tilt as Tim has pointed out. This makes it awkward to use as a universal reference for multiple locations. We have got used to Greenwich Mean Time as a time reference but local time in terms of where the sun is at your location on the earth is arguably the more 'fundemental' reference. If one wishes to convert the 'real time' to the artificlal time of GMT one needs to know how far out GMT is from real time on any particular day. So I would say the Sundial is correct and the variation in the timings of the orbit need to be used to calculate how far out GMT is.

If you are navigating around the globe by time you need the local noon and the time of the local noon at Greenwich in order to calculate your longitude. If you use GMT and local noon you are going to be in fix.

It's very true that we have come to regard clock time as the norm which is understandable, but just because we are used to it doesn't neccessarily mean that it's more fundemental than the actual position of the heavens.

Just making a curious point really.

regards Martin

 

Edited By Martin Kyte on 25/08/2020 15:58:27

[Tim StevensParticipant @timstevens64731]

Which you regards as 'fundamental' time depends on whether you regard time as passing steadily, or speeding up and slowing down in time with the earth wobbles. I suggest that steady time is the same everywhere in the universe, and so more 'fundamental'

Tim.

[John HaineParticipant @johnhaine32865]

Hmmm…

"Problem with this is Noon depends on longitude, making Bristol time about 10 minutes later than London. No good for marine navigation or railways. A 10 minute navigational error is out by roughly 100 miles, and timetabling trains was a nightmare!" – well actually there used to be a clock on Bristol Post Office near Temple Meads with two minute hands showing local time and London time, for the benefit of new arrivals from London. As long as you know about the shift you can allow for it in the timetables. And of course the longitude "error" is the essence of using a clock for navigation, long before railways.

Distributed telegraph time was introduced by the Post Office, partly so they could time-stamp telegrams synchronously everywhere, and also later to make sure exchange clocks were accurate for billing. Anyone who could pay could have a 10.00 GMT time pulse down a wire – for example the Robertson Clock at Bristol University used the pulse to phase lock its pendulum by adjusting its rate according to the measured phase difference. Telegraph time wasn't phased out until the 50s I believe.

The seasonal variation of solar noon compared to mean time is the "equation of time" and can be corrected by an "analemmatic sundial" or sets of tables. Some clocks incorporate a cam to show the error to make it easier to set them by the sun. There are dials where the shadow of the gnomon falls on an analemmatic curve and you can read off the right time if you know the date – there's one on the terrace outside the cafe at the National Maritime Museum.

Time is absolutely not the same everywhere in the universe – relativity shows that it varies with gravity, and every object in motion has its own proper time. This has to be allowed for in GPS to get accurate fixes and time distribution. For a very entertaining demonstration see this.

How did we get here from carbon fibre?

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Alan Crawley on 25/08/2020 14:09:49:

[…]

Another question, I learned that the term "Regulator" was used as it was the accurate clock in the clockmaker/repairers workshop to adjust other projects. What did they set the Regulator to?

.

Regulators [proper ones] were typically set to Sidereal time

I will post a very early description of how it was done, when I can find it in my ‘archive’

MichaelG.

.

Edit: As promised … See p91 of this transcription:

https://quod.lib.umich.edu/e/eebo/A60473.0001.001?rgn=main;view=fulltext

 

Edited By Michael Gilligan on 25/08/2020 18:27:25

[SillyOldDufferModerator @sillyoldduffer]

Posted by John Haine on 25/08/2020 17:23:04:

Hmmm…

…

As long as you know about the shift you can allow for it in the timetables. …

Ah, but how many timetables are needed? Passengers departing Paddington need one in London time while anyone meeting the train at Temple Meads either needs another timetable in Bristol time or does maths on the London one. Same problem in reverse, does the Temple Meads timetable show departures in Bristol time, or in London time? Most mid-Victorians weren't well educated and would have found the maths difficult. Even today many struggle to understand timetables.

Worse for anyone carrying on to Fishguard which is nearly 5 degrees West of Greenwich. Much easier to put everyone in the British Isles into the same time zone, and ignore noon being slightly off clock time.

I think pairs of timetables would be needed for all journeys with a significant east-west component, like London to Birmingham or Liverpool to York. Think of the work and the mistakes!

Dave

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Tim Stevens on 25/08/2020 16:18:08:

Which you regards as 'fundamental' time depends on whether you regard time as passing steadily, or speeding up and slowing down in time with the earth wobbles. I suggest that steady time is the same everywhere in the universe, and so more 'fundamental'

Tim.

.

But of course … ‘They’ abandoned all concepts relating to astrophysical behaviours, and Time [or more specifically ‘the Second’] is defined by the apparently simple process of counting.

MichaelG.

.

Ref.

The official definition of the second was first given by the BIPM at the 13th General Conference on Weights and Measures in 1967 as: "The second is the duration of 9192631770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the caesium 133 atom." At its 1997 meeting the BIPM added to the previous definition the following specification: "This definition refers to a caesium atom at rest at a temperature of 0 K."^[3]

Edited By Michael Gilligan on 25/08/2020 18:37:21

[Neil WyattModerator @neilwyatt]

I have finally found out how a Caesium clock works (hint: you won't find out from Wikipedia).

Neil

[Martin KyteParticipant @martinkyte99762]

What we mean by time is ultimately an instant of a particular configuration of physical objects and the passage of time can only be perceived by something physically changing configuration wether the oscillation of electron energy states in an atomic clock, the orbit of the planet or the oscillations of a pendulum or qurtz crystal. I quite agree that the access to a 'timebase' of constant period is essential but it can only be artificial. OK so noon exhibits a periodic variation when compared to a constant timebase but noon is noon and he use of the word error for a natural phenomenon seems a little out of place. I feel there is a philosophical difference between the absolute instant of time which is a configuration and a constant period of time which at best a constuction.

OK I am being a little pedantic here and maybe a little tongue in cheek too but when you really think about there is real or 'natural time' being the configuration of the universe and man made time contrived to generate a constant timebase in order to to measure the first.

As John Haine has already mentioned there is no universal clock as Newton envisaged where any event anywhere in space could be said to take place at a particular time rather with Einsteins spacetime no two spacially distant observers will agree on the timing of any event.

As I say it's a bit philosophical but interesting to think about all the same.

best regards Martin

[Author]

Posts