C7 Temperature stability

[Paul Galpin]

I have a Gents C7, which has been running nicely for a few years now.
Accuracy is controlled by a Satellite time device, which usually controls it to less than a second, compared to Satellite Time Signal. This loads a ball-bearing off and on to a carrier about halfway up the pendulum shaft.
However, every Spring and Autumn it runs out of range, and then I have to adjust the knob at the bottom of the pendulum, on which the bobweight sits.
It is obviously a compensated pendulum, by the design, but it is overcompensated, i.e. it runs slower in winter than in summer.
Only two things appear to me to be relevant;
(a) the suspension spring was broken, and I replaced it with an 0.015 feeler gauge.
From what I read, no-one seems to think this important.
(b) This clock worked fine with its first owners (SABC), but it lived underground in a more-or-less constant temperature environment. I don't have that option, and the temperature in my lounge ranges from 5C to 30C ( 28.3C as I type!)
Should a professional clock be able to handle this range?
Has anyone else had experience of wide temperature variations?

Paul Galpin

 

[John UK]

The Gents pendulum rod should be Invar (unless it is a VERY early clock pre about 1910 which used a wooden rod). Invar should in itself give a good degree of temperature stability, but 25 degrees C is a wide range. I don't believe there is any 'compensation' as such used, but I may be wrong on that.

One suggestion (since you are using a see saw type corrector) is to increase the degree of correction it can apply (i.e. a larger weight added/removed).

I run a Synchronome (Invar rod, lead bob on mine) and it does need a small change in weight summer to winter, which I do by (manually) adding 1 gram weights and 6BA washers to the weight tray - but my temperature variation is smaller, (typically less than 15 degrees C).
A Silent Electric (brass bob and presumed Invar rod) varies rather more than the Synchronome.

All of these clocks were designed 100+ years ago - when they were considered very accurate and it would have been difficult to find anything more accurate against which to check them pre radio 'pips' and speaking clock, but in 100 years, expectations of accuracy (quartz, radio control, GPS etc.) have grown.

 

[Paul Galpin]

The Gents pendulum rod should be Invar (unless it is a VERY early clock pre about 1910 which used a wooden rod). Invar should in itself give a good degree of temperature stability, but 25 degrees C is a wide range. I don't believe there is any 'compensation' as such used, but I may be wrong on that.

One suggestion (since you are using a see saw type corrector) is to increase the degree of correction it can apply (i.e. a larger weight added/removed).

I run a Synchronome (Invar rod, lead bob on mine) and it does need a small change in weight summer to winter, which I do by (manually) adding 1 gram weights and 6BA washers to the weight tray - but my temperature variation is smaller, (typically less than 15 degrees C).
A Silent Electric (brass bob and presumed Invar rod) varies rather more than the Synchronome.

All of these clocks were designed 100+ years ago - when they were considered very accurate and it would have been difficult to find anything more accurate against which to check them pre radio 'pips' and speaking clock, but in 100 years, expectations of accuracy (quartz, radio control, GPS etc.) have grown.

Thanks for the info!

In fact, my C7 has an SABC-designed corrector on it. This was triggered by an "Observatory Pulse" contact closure every hour. Using a polarity changing relay and a Carpenter bi-stable relay, it adds on or takes off a ball bearing. This involves very small gaps between the carrier and the support bracket. I have the largest size ball on it that does not cause interference.

The basic temperature compensation is performed by the bobweight being supported at the bottom of the pendulum shaft. As the shaft grows with temperature, the bobweight grows up the shaft. I did the maths, and the temperature coefficients of the Invar shaft and the cast-iron bobweight (assumption?) cancel out exactly!
But it doesn't work out like that in real life!
Perhaps it's not cast iron.

Paul G

 

[John UK]

Gent's own "see saw" mechanism can be seen on Clockdoc.org under Gents/Synchronisers/see saw

 

[colintd]

I have a Gents C7, which has been running nicely for a few years now.
Accuracy is controlled by a Satellite time device, which usually controls it to less than a second, compared to Satellite Time Signal. This loads a ball-bearing off and on to a carrier about halfway up the pendulum shaft.
However, every Spring and Autumn it runs out of range, and then I have to adjust the knob at the bottom of the pendulum, on which the bobweight sits.
It is obviously a compensated pendulum, by the design, but it is overcompensated, i.e. it runs slower in winter than in summer.
Only two things appear to me to be relevant;
(a) the suspension spring was broken, and I replaced it with an 0.015 feeler gauge.
From what I read, no-one seems to think this important.
(b) This clock worked fine with its first owners (SABC), but it lived underground in a more-or-less constant temperature environment. I don't have that option, and the temperature in my lounge ranges from 5C to 30C ( 28.3C as I type!)
Should a professional clock be able to handle this range?
Has anyone else had experience of wide temperature variations?

Paul Galpin

I've got a Gents C7 (converted to opto-magnetic drive) with exactly the same kind of negative rate vs temperature (slower cold). Mine has a change in going of around ~ -150ms/day/C. Like you I've done the calculation on the assumption on an Invar rod, and steel/iron bob, and whilst I believe it is over-compensates, I can't generate anything like the observed degree of change.

I'm currently experimenting with an alternative bob, spring and upper chops to see if this makes any difference, and will report back.

The pendulum from my Synchronome in the same case (with adaptation blocks for the mount) remains almost perfectly temperature compensated. As expected it is shorter and fatter than the Gents bob.

 

[colintd]

I've got a Gents C7 (converted to opto-magnetic drive) with exactly the same kind of negative rate vs temperature (slower cold). Mine has a change in going of around ~ -150ms/day/C. Like you I've done the calculation on the assumption on an Invar rod, and steel/iron bob, and whilst I believe it is over-compensates, I can't generate anything like the observed degree of change.

I'm currently experimenting with an alternative bob, spring and upper chops to see if this makes any difference, and will report back.

The pendulum from my Synchronome in the same case (with adaptation blocks for the mount) remains almost perfectly temperature compensated. As expected it is shorter and fatter than the Gents bob.

Fyi the density of the bob on my clock is 7.99g/cm3, which combined with the machining marks, suggests mild steel

 

[colintd]

Looking back through my notes, with the original pulsynetic mechanism, a large part of the going change appeared to be down to a decrease in amplitude with increasing temperatures. Looking at the pulse timings, this mainly seemed to be down to change of impulse instant and length, perhaps down to expansion of the brass count arm?

With my replacement opto-magnetic drive, my issues may instead be down to the top of the bob being too close to the drive coils. The squatter synchronome bob definitely has more clearance.

 

[Paul Galpin]

Members may have seen this discussion on another group.
One thing I got from that was that the bob was originally cast iron, and that was what I based my calculations. However, at some stage, production changed to machined steel.
If, and I stress the "if", the dimensions were unchanged, the different density would bring about the over-compensation problem that I have.
"Have" or perhaps "had". From last winter to present, I have only been adjusting the pendulum in small increments, just enough to get it back into the working range. Now we are heading into Winter again, and maybe I won't have to re-tweak.

I feel that I must repeat that this criticism is not basically of Gents design. It is due to my desire to have the clock in my study, and the large temperature variations therein.

 

[colintd]

Members may have seen this discussion on another group.
One thing I got from that was that the bob was originally cast iron, and that was what I based my calculations. However, at some stage, production changed to machined steel.
If, and I stress the "if", the dimensions were unchanged, the different density would bring about the over-compensation problem that I have.
"Have" or perhaps "had". From last winter to present, I have only been adjusting the pendulum in small increments, just enough to get it back into the working range. Now we are heading into Winter again, and maybe I won't have to re-tweak.

I feel that I must repeat that this criticism is not basically of Gents design. It is due to my desire to have the clock in my study, and the large temperature variations therein.

Fully appreciate your perspective, and I wasn't trying to be overly negative about the Gents design. It love them, and have several in my collection (some working, other in parts.)

Having said that, I've got both Gents and Synchronome systems, and after exploring them for a number of years, to my mind there are definitely quite a few elements of the Gents design which seems more driven by a need to avoid the core Synchronome patents than precision time keeping.

Not using a pallet and gathering arm directly attached to the rod, but instead using the indirect system with extra pivots, gaps, etc being a good example...

 

[bruce linde]

why not try a different suspension spring and see what happens?

the original feeler gauge may not offer the same flex characteristics as suspension material offered specifically for clocks.

(sounds like you guys are all measuring/monitoring far more than i do, but my less rigorous/subjective opinion is that my C7 needs minor pendulum tweaks with more extreme temperature/season changes... but... i only run it from 9am to 9pm every day as it's in a bedroom on a wi-fi programmed outlet. an alarm goes off at 859 and i walk in and start it swinging in time with a welch spring and company regulator. the pendulums swing in perfect harmony all day long... until the seasons change and said minor tweak is required.)

since suspension springs are relatively easy to replace, it would be interesting to compare results of feeler gauge vs. something from (for example) timesavers.

 

[colintd]

why not try a different suspension spring and see what happens?

the original feeler gauge may not offer the same flex characteristics as suspension material offered specifically for clocks.

(sounds like you guys are all measuring/monitoring far more than i do, but my less rigorous/subjective opinion is that my C7 needs minor pendulum tweaks with more extreme temperature/season changes... but... i only run it from 9am to 9pm every day as it's in a bedroom on a wi-fi programmed outlet. an alarm goes off at 859 and i walk in and start it swinging in time with a welch spring and company regulator. the pendulums swing in perfect harmony all day long... until the seasons change and said minor tweak is required.)

since suspension springs are relatively easy to replace, it would be interesting to compare results of feeler gauge vs. something from (for example) timesavers.

I have some 0.15mm spring steel stock on my desk for exactly that purpose! I'll let you know how it goes.

 

[colintd]

Suspension changed, and I'll let you know when I have some results.

In the meantime, I've seen suggestions that the Gents pendulum rod might be "Super Invar" / "Sinevar" not standard Invar. From memory that has a much lowers coefficient of expansion around room temperature. Does anyone know if this is the case?