Temperature compensating a pendulum clock?

[jeff_wisnia]

We have a very nice pendulum wall clock in our kitchen which has the
type of movement where a battery powered electric motor winds the spring
of a traditional type pendulum movement.

The clock keeps excellent time in the summer, but during the cold months
of our New England weather, when we let the temperature in the part of
the haouse where the clock is located drop into the 50s during the
night, the clock gains about 3 minutes a day.

I presume this is due to the pendulum shaft shortening with temperature
and I can sort of compensate for it by adjusting the bob downwards a
bit, but it's still not accurate enough to suit my taste.

I was wondering if I could fasten one end of a strip of bimetal
scavengened from an old thermostat onto the back of the bob in such a
way that the free end moved downward as the temperature dropped and with
some trial and error tuning end up with a "temperature compensated
pendulum."

Comments appreciated, I'd really like to keep the existing movement in
that clock rather than jump to a quartz movement.

Thanks guys,

Jeff

--
Jeffry Wisnia
(W1BSV + Brass Rat '57 EE)
The speed of light is 1.8*10e12 furlongs per fortnight.

[d530]

"jeff_wisnia" <jwisniaDu...@conversent.net> wrote in message news:jenfgv$btj$1...@dont-email.me...
>

Hello Jeff, nice to hear from you again.

>
> I was wondering if I could fasten one end of a strip of bimetal
> scavengened from an old thermostat onto the back of the bob in such a
> way that the free end moved downward as the temperature dropped and with
> some trial and error tuning end up with a "temperature compensated
> pendulum."
>

If I understand your idea correctly fastening a bimetal to the bob won't influence the lenght of pendulum thus won't have any effect on it's period of swing.

> Comments appreciated, I'd really like to keep the existing movement in
> that clock rather than jump to a quartz movement.
>

I'm convinced you won't replace pendulum movement with quartz one.
Better place radio controlled clock in the kitchen as time measuring device and let pendulum clock work bearing in mind it's imperfect behaviour as temperature changes (just correct readings every day or so).


What others think ?

best, Wojtek

[jeff_wisnia]

d530 wrote:
> "jeff_wisnia" <jwisniaDu...@conversent.net> wrote in message news:jenfgv$btj$1...@dont-email.me...
>
>
> Hello Jeff, nice to hear from you again.
>
>
>>I was wondering if I could fasten one end of a strip of bimetal
>>scavengened from an old thermostat onto the back of the bob in such a
>>way that the free end moved downward as the temperature dropped and with
>>some trial and error tuning end up with a "temperature compensated
>>pendulum."
>>
>
>
>
> If I understand your idea correctly fastening a bimetal to the bob won't influence the lenght of pendulum thus won't have any effect on it's period of swing.
>

> What I had in mind was a horizontal strip of bimetal hidden behind the bob
with one end fastened to the back of the bob with something like a

>
>>Comments appreciated, I'd really like to keep the existing movement in
>>that clock rather than jump to a quartz movement.
>>
>
>
> I'm convinced you won't replace pendulum movement with quartz one.
> Better place radio controlled clock in the kitchen as time measuring device and let pendulum clock work bearing in mind it's imperfect behaviour as temperature changes (just correct readings every day or so).
>
>
> What others think ?
>
> best, Wojtek

To d530:

What I had in mind was a horizontal strip of bimetal hidden behind the
bob with one end fastened to the back of the bob with something like a
litle L shaped strip of metal. The bimetal would be oriented so that its
free end moved down as the temperature dropped. With a small weight
(amount to be determined) added to the free end of the bimetal strip.

So, when the temperature dropped the center of mass of the bob/bimetal
gadget moved lower. Fiddling with the amount of weight on the end of the
bimetal strip, or its position along the strip should get me the right
amount pf compensation.

Tell me why that couldn't offset the speeding up caused by the pendulum
shaft shrinking in length as it got colder.

It's the same principal as was used in those pendulums with mercury
filled columns isn't it? It moves the center of mass of the bob lower as
the temperature drops.

[Erik]

In article <jenfgv$btj$1...@dont-email.me>,

I know little in general about clocks... but remember hearing of better
clocks being compensated.

Check this, in particular the section on thermal compensation about half
way down;

<http://en.wikipedia.org/wiki/Pendulum_clock>

Erik

[gpsman]

On Jan 12, 3:23 pm, jeff_wisnia <jwisniaDumpThisP...@conversent.net>
wrote:

> I can sort of compensate for it by adjusting the bob downwards a
> bit, but it's still not accurate enough to suit my taste.

I can barely tell time, but I'm willing to suggest you change your
taste or clock.
-----

- gpsman

[dAz]

On 13/01/12 07:23, jeff_wisnia wrote:
>
> We have a very nice pendulum wall clock in our kitchen which has the
> type of movement where a battery powered electric motor winds the spring
> of a traditional type pendulum movement.
>
> The clock keeps excellent time in the summer, but during the cold months
> of our New England weather, when we let the temperature in the part of
> the haouse where the clock is located drop into the 50s during the
> night, the clock gains about 3 minutes a day.
>
> I presume this is due to the pendulum shaft shortening with temperature
> and I can sort of compensate for it by adjusting the bob downwards a
> bit, but it's still not accurate enough to suit my taste.

what is the pendulum rod made of?

the other thing to consider is the suspension spring in cold temps will
stiffen the spring and make the rate faster,

>
> I was wondering if I could fasten one end of a strip of bimetal
> scavengened from an old thermostat onto the back of the bob in such a
> way that the free end moved downward as the temperature dropped and with
> some trial and error tuning end up with a "temperature compensated
> pendulum."

hmm I think you find will that won't make much of difference, better to
improve the pendulum, if for example it's a brass rod brass/lead bob,
then simply fitting a steel rod will help or a timber one, if you could
get your hands on it invar would make a better rod as it won't vary much
in length.

[d530]

"jeff_wisnia" <jwisniaDu...@conversent.net> wrote in message news:jenqds$l8c$1...@dont-email.me...

> What I had in mind was a horizontal strip of bimetal hidden behind the
> bob with one end fastened to the back of the bob with something like a
> litle L shaped strip of metal. The bimetal would be oriented so that its
> free end moved down as the temperature dropped. With a small weight
> (amount to be determined) added to the free end of the bimetal strip.
>
> So, when the temperature dropped the center of mass of the bob/bimetal
> gadget moved lower. Fiddling with the amount of weight on the end of the
> bimetal strip, or its position along the strip should get me the right
> amount pf compensation.
>
> Tell me why that couldn't offset the speeding up caused by the pendulum
> shaft shrinking in length as it got colder.
>

Jeff, now when you fasten additional mass to free end of bimetal there is a chance of success.

Before you didn't mention about additional mass, only the bimetal.
As you imagine, the bimetal itself won't influence period of pendulum because it's mass is negligible compared to the mass of bob.

Additional weight fastened to the free end of bimetal might work.

May I suggest to experiment without bimetal first ?
Just fasten choosen weight to the pendulum with modeline or so and check how it influences period of pendulum.
Next alter the point of fastening and again check if it works.

Let us know about results.

best, Wojtek

[jeff_wisnia]

If I decide to take on the challenge (Just for Ss & Gs) I'll keep my eye
open for a glass rod of the right length and shape. I know the thermal
coefficient of glass is quite low. I should be able to replace most of
the length of the pendulum rod with a glass rod with epoxied on metal
ends, maybe then sprayed with bronze paint so it doesn't look so odd.

It's been a while since I've mucked around much with clocks, the most
interesting one I did about 45 years ago had a "verge" escapement. It's
still hanging on a wall in our home but don't even think about asking me
how good a timekeeper it is. <G>

http://home.comcast.net/~jwisnia18/jeff/clock.html

Jeff

--
Jeffry Wisnia
(W1BSV + Brass Rat '57 EE)
The speed of light is 1.8*10e12 furlongs per fortnight.
>
>>