Fusee mainspring tension

[SeanFD]

Hi All,

First post here - first clock too - so please bear with my almost complete lack on knowledge! I only wish I'd found the forum earlier.

So I'm working on a fusee wall clck with no identifying markings except for some numbers engraved here and there.

The problem was it was running erratically; would run down and then refuse to go again. After a variable time during which I would try to get it started from time to time, it would run for a while (hours) then stop again. Then, after maybe a few more attempts at starting, it would go for several days until run down again. Time keeping was good when it did decide to do so.

I dismantled it, cleaned everything and replaced mainspring and the steel wire with synthetic gut.

The advice I found online regarding setting the tension on the mainspring varied and was, in any case, none to clear. Finally, the advice I chose to follow was a combination of these two:

Quote

 

I start with the fusee chain wound entirely around the fusee, except for a portion sufficient to reach the barrel. At this point, the spring is unwound. Then I give the mainspring barrel winding arbor a couple turns, and let the clock run until it stops, run down. Then give the winding arbor another turn, and set the clock running again. I do this until it runs 8 or 9 days. THEN I wind it fully at the fusee, and check the stop works when it is wound.

After the clock is all together,lubed etc. I advance the spring barrel one notch at a time,and at the point when the movement has a good beat,pend swing,etc I stop,Then wind the clock to full wind,and let it run. If the time keeping is consistant,in other words doesn't loose toward the end of the run time then I feel pretty good about the setting. Also of note this generally works out to not less than one turn and not more than a turn and a half to one and three quarters.

 

I advanced the barrel click 3-4 clicks at a time, expecting the escape wheel to start moving. However after a number of 3-4s – foolishly I wasn't counting – I thought I had gone well past the 1 ½ -2 ½ turns I had seen mentioned as the range required. I then started doing it one click at a time, for maybe another ½ rotation, but still no response from the escape wheel.

Finally I chickened out and decided just to try it out. It's now ticking away contentedly. 

I've attached teh hands to see if it's anywhere near keeping time, but what should I do once it has run down?

Thanks for any thoughts.

[nevenbekriev]

See, there ia another, more correct way to do this job. I will describe it and will use word 'chain', as in the most cases it really is one.

1. The movement is assembled without the chain and the lever (or the verge, if it is a verge escapement), the train is oiled and the wheels turn free.

2. Attach the chain to the barrel and rotate the barrel to the direction to wind entire cain on the barrel. Then attach the other end of the chain to the fusee and force the barrel to strain the chain and to bring the train to motion untill no more motion is possible, and the chain is pulled to the end, at this point it will be perpendicular to the fusee wheel.

3. Holding the barrel as to strain the chain, put the ratchet wheel, wind the spring a little (about 1/2 turn) and engage the click.

This is the procedure to put the chain in place.

Next is the first winding and the check and adjustment of the torque on the fusee.

Block the train and start to wind the spring (by turning the fusee arbour) Observe how the spring goes in the channel of fusee and if needed, help it to go in the channel correcty, do not let it go out of the channel. This is only when first winding, next times the chain should go where needed by itself. Observe if the end of winding is limited by the limiting device correctly. Then release the train and let the spring unwind by train turning. The torque on the fusee wheel must be equal all the time (this is the aim of the fusee), this can be evaluated by the 'speed' of the train spinning. If near the end of reserve speed reduces, then the initial tension of the spring must be increased ( by the ratchet wheel position) and wice versa. As You have replaced the spring, have in mind that the fusee has been calculated and made to equalize the original spring torque, so it is not obligatory to be able to fully equalize the torque of the new spring. Also, if increasing the initial tension to much, it is possible to reach a momment when winding the spring to the end wil get limited by the spring end, but not by the limiting device, which is not acceptable.

[oldhippy]

nevenbekriev, has described it completely correct. How did you put the new spring in the barrel I hope with a clock mainspring winder.   

[SeanFD]

Thanks for the replies.

It is an anchor escapement, not verge, if that makes a difference.

Yes I used a main spring winder.

The mechanism has been running now since Sunday, nicely in beat.

The fusee is just short on half empty.

The clock is losing time (about 50 mins since Sunday) - easy to see even though there is no face on as yet. 

Thanks also for the advice, but ....

 I have to attach the chain (actually a cord) before assembling the mechanism; once assembled I cannot access the attachment point for either barrel or fusee. I could reach this point by waiting for the mechanism to run down so the fusee is "naked" if that would work.

During step 2, is the spring completely unwound in the barrel, no tension at all? At the end of this step, the fusee is still empty and directly connected to the barrel?

Where is the best place to block the train? Using a wooden thingy, I imagine?

"Observe how the spring goes in the channel of fusee" Should this read chain?

 

"Then release the train and let the spring unwind by train turning." With or without the pendulum attached?

 

"Also, if increasing the initial tension to much, it is possible to reach a momment when winding the spring to the end wil get limited by the spring end, but not by the limiting device, which is not acceptable."  Sorry, I don't quite understand this.

 

So based on the above, I think I should  – this is a question 

1. Remove pendulum and let the spring unwind.

2. I think I have put too much tension the spring, so at this point I should release the click and let down the spring.

3. Then restart the procedure from Step 3 above.

Quote

The clock is losing time (about 50 mins since Sunday) - easy to see even though there is no face on as yet. 

I post this separately as it is a slightly different issue.

I cannot find anything that explains the relative functions of teh spring and pendulum in regulating the time keeping.

I know the length of the pendulum is critical, but if the spring in this case is over-tensioned that effect will that have? I would imagine it would make the clock run fast, but it can't as the whole mechanism is governed by the periodicity of the pendulum.

How does it work then on a normal - non-fusee - clock. Does it just but more stress on the mechanism and result in greater wear on the teeth etc?

An (simple) explanation, or a reference, would be much appreciated.

[oldhippy]

When the fusse cone is empty of the chain or line, there should be a tension in the spring in the barrel, the chain or line should be tight on the barrel you shouldn't be able to move it from side to side. The tension is caused by putting a wind or two on the large ratchet wheel on the barrel arbor. The ratchet is normally held with a pin, if you have put a  straight steel pin in you need to bend it into an S shape so it cannot slip out this is a safety tip. 

If the suspension spring is too strong the clock will gain the opposite if the spring is to weak  the pendulum must also be of the correct weight and length. The more modern clocks have springs that are made of the same thickness and strength from end to end so time keeping should be good again providing correct suspension spring and pendulum are fitted.    

[nevenbekriev]

10 hours ago, SeanFD said:

 I have to attach the chain (actually a cord) before assembling the mechanism; once assembled I cannot access the attachment point for either barrel or fusee. I could reach this point by waiting for the mechanism to run down so the fusee is "naked" if that would work.

OK, in You case You will assemble the movement with the line attached to the barrel and fusee and will  wind the line  entirely on the barrel after that. It is important when replacing the line to know the correct lenght of the line. The rule is that when line wound on the fudsee entirely, about 1/4 turn still to remain on the barrel.

10 hours ago, SeanFD said:

During step 2, is the spring completely unwound in the barrel, no tension at all? At the end of this step, the fusee is still empty and directly connected to the barrel?

Yes, no tension. The ratchet is not placed yet, the spring in the barrel is complitelly unwound. At the end of the step the fusee is empty and the line is attached to it and strained perpendiculary to it, not on tangent.

10 hours ago, SeanFD said:

Where is the best place to block the train? Using a wooden thingy, I imagine?

The escape wheel. Use whatever that will not damage anything.

10 hours ago, SeanFD said:

"Observe how the spring goes in the channel of fusee" Should this read chain?

Yes, this is typing mistake, sorry.

10 hours ago, SeanFD said:

"Then release the train and let the spring unwind by train turning." With or without the pendulum attached?

The word 'pendulum' here is not correct, You should use 'anchor' and yes, it is removed from the movement at this point, according my instruction 1 (Ihave used the word 'lever' there):

On 2/20/2024 at 9:01 AM, nevenbekriev said:

1. The movement is assembled without the chain and the lever (or the verge, if it is a verge escapement), the train is oiled and the wheels turn free.

 

10 hours ago, SeanFD said:

"Also, if increasing the initial tension to much, it is possible to reach a momment when winding the spring to the end wil get limited by the spring end, but not by the limiting device, which is not acceptable."  Sorry, I don't quite understand this.

Ok, let sey the barrel arbour makes 6 full turns when winding the spring in the barrel from unwound to fully wound state of the spring. Let sey that the barrel makes 4.5 full turns when winding the clock from fully unwound (the line is on the barrel) to fully wound (the line is on the fusee) state. This means that You should not wind the spring in the barlel when adjusting the initial tension to more than 1.5 turns of the arbour of the barrel.  If You make 2 full turns, then when winding the clock, the barrel will be able to turn to only 4 turns, so the limiting device will not limit the rotation of the fusee arbour and You will be able to force the line much more than when the limiting device acts, also the power reserve will be less. Hope this is clear now...

10 hours ago, SeanFD said:

So based on the above, I think I should  – this is a question 

1. Remove pendulum and let the spring unwind.

2. I think I have put too much tension the spring, so at this point I should release the click and let down the spring.

3. Then restart the procedure from Step 3 above.

Yes, but again - not the pendulum, but the anchour.

 

About the loosing time...

In this kind of escapement, the frequency pritty much depends on the torque. The higher torque - the highrer frequency. This is because this escapement forces the pendulum to oscillate significantly faster than it's own resonant fequency. The frequency depends on the 'depth' of the escapement too - the deeper escapement is, the lower the frequency, and the higher amplitude of pendulum oscillations. The heavier the pendulum is, the harder to the movement (the escapement torque) to force the pendulum to oscillate faster than it's own frequency. And in the end, the own pendulum frequency depends only on the pendulum lenght and a little on the suspension spring 'strenght' This are all the relations between all the factors.

You can try to use the old spring (if it is not broken) and see if the clock will work faster with it. Yes, the old springs of fusee clocks sometimes give bigger torque than modern ones, no matter if they seem to be 'set'

You can shorten the pendulum to achieve correct frequency. If the torque is reduced, but enough for the movement to work reliably,  then reduced torque will only lead to lessen the wear. The torque in fusee movements is more or less constant all the time, this is the function of the fusee.

You can use the Clock Tuner  free app for android to adjust faster the clock rate. You will need to know the BPH of the movement, so count the teeth of the wheels and pinions and calculate the BPH

Edited February 23 by nevenbekriev