Fusee Clock keeping poor time

[4ster]

I have a late 19th century fusee mantle clock that keeps poor time. If fully wound-up it will lose around 5 minutes over the first couple of days before slowly regaining this time after around day four and then ending up about 5 minutes fast by the end of the week. For the size of the plates (5.5" x 4") and compared to other fusee's I have, the spring appears particularly strong and results in a substantial tick from the escapement. I suspect that the spring gauge is too thick and that I will need to fit a lighter spring. Does anyone recognize these symptoms and is there any other explanation for this type of activity in a fusee movement?

The pendulum is homemade and weighs 250 grams (perhaps a little heavy), could the weight have anything to do with this. The fusee appears typical and has a standard sized barrel about 55mm diameter & 42mm deep. I have not removed the spring from the barrel and don't know its gauge, but should this be in the region of 0.4mm?

[clockboy]

I presume you have tried adjusting the pendulum (ie the longer the slower it will run). The inconsistency could be down to a number of factors such as worn bushings or worn pivots. 

[oldhippy]

Sounds to me the clock needs a good clean and as clockboy has said worn holes and pivots could also be worn. The reason the fusee was invented was to counter act for the un-even thickness of the mainsprings, back in the old days it was not possible to manufacture springs of equal thickness from end to end. The fusee counteracts against that problem. The spring might also be at fault as well in your clock. It could be tired or a lack of oil or sticky with grease.  The best thing is it needs to be looked at before any more wear causes more damage. 

[watchweasol]

Hi One can only retierate the advice given a good service - clean and oil check bushes etc and then check again.

[rodabod]

How large is the pendulum swing? It could be partly related to excessive amplitude (circular error and lots of recoil).

[oldhippy]

2 hours ago, rodabod said:

How large is the pendulum swing? It could be partly related to excessive amplitude (circular error and lots of recoil).

If the movement is worn such as bad holes and poor pivots, that will cause power to be lost, that will effect the swing of the pendulum, also worn escape teeth and pallets. The recoil will be very poor due to power lose it will not increase.

[rodabod]

I queried the amplitude as the OP suggested that the mainspring seemed like it may be on the strong side. 
 

Worth mentioning also to the OP that a “substantial tick”  an sometimes be a sign of excessive drop as the escape wheel accelerates when it gets given the chance. 

[4ster]

Wow, my first time on the forum, thanks for the immediate responses and all the advice.

I picked this movement up on ebay a couple of years ago to fit in the first clock case I built some 45 years ago and to replace the old Perivale movement I had found at that time. It was short of gut, suspension and pendulum but seemed in remarkably good condition. There is virtually no sign of wear on the pinions and the bushes look original and nicely round. I was delighted, and once I new the pendulum length would fit in the case, the movement had a strip-down and a session in the ultrasonic bath, a polish of the pivots and holes and a little oil. Then it was in the case and on the bench for setting up. It was at this stage I realized that time keeping was inconsistent; running slow when fully wound, then faster as the spring wound down. At the extremes this is several minutes a day.

At the time the tension in the spring gave me concern. The torque required to turn the seemed excessive such that if it had a fusee chain I would be reluctant to wind it. Felling that this may be the issue I tried varying the spring pre-tension between nothing an around 3 full turns, but the time keeping seemed much the same. I have also substituted the pendulum with one having a bob of 145g  again with no improvement. At this time the next project came along and its sat on the self in the workshop ever since awaiting some inspiration. 

I have just taken a look at the swing which is 23mm at 135mm from the pivot point, by my reckoning this is near enough 5 degrees either side of center.

I'm just  at a bit of a loss as I have never encountered anything like this with a fusee. As you say, the whole reason for it was to overcome variation in torque as the spring wound down. Think it may be time to take it apart again and take a closer look. I suppose I was hoping that someone may confirm that excessive spring tension can totally mess-up the equalizing action of the fusee cone!   

[4ster]

Sorry. I got the swing of the pendulum wrong. Its swinging through 25mm at 210mm from the pivot; about 3.5 degrees either side of center

[4ster]

Just remembered what a work of art it was to get the movement in the case. Anyway it's out now.

Rodabob, Oldhippy

Whats your thoughts on the escapement drop?

[oldhippy]

From what I can see the escape wheel teeth look dry, I can't see the pallet faces, so I can't see if they are worn. The movement sounds about right and in beat, it does look a little shallow. Time keeping on a fusee  should be spot on, they are extremely good time keepers, no more then a minute a week. Escape wheel teeth look fine no sign of bad wear, and if the pallet faces and holes and pivots are good. It will be all down to the main spring. The tension should be so, that you cannot move the line on the barrel when the movement has run down, if so this will cause the line to be out of line when it comes to winding the clock up, it will not wind easy onto the fusee. I would change that line to the nylon type, they do not mark the barrel or unravel. The type you have can unravel and cause damage. Pendulums for fusee movements included.

https://www.cousinsuk.com/product/pendulums-english-dial-fusee

[wls1971]

Do you have a picture of the back plate of the clock one or two things make me think this is a English Williamson movement, I had one recently  that had a number of problems, the thing I noticed about that clock was the movement was of compact design with a strong spring and substantial chain, large pendulum with short suspension, beat correction on the crutch and a tick that could wake the dead once cased up.

[rodabod]

12 hours ago, 4ster said:

Sorry. I got the swing of the pendulum wrong. Its swinging through 25mm at 210mm from the pivot; about 3.5 degrees either side of center

Was that on a full wind?

 

[nickelsilver]

How much pretension is there on the barrel?

[4ster]

Currently about one & a quarter turns, but have tried between zero and three turns without improvement to time keeping.

[4ster]

Pendulum swing was measured with just two winds onto the fusee cone. I'll try measuring it at the start, mid point and fully wound.

[oldhippy]

5 minutes ago, 4ster said:

Pendulum swing was measured with just two winds onto the fusee cone. I'll try measuring it at the start, mid point and fully wound.

It shouldn't make any difference to the swing of the pendulum if you only have little power or if its on a full wind. 

[4ster]

Hi WLS 1971

Back plate of movement.

I never could find what is was or where made, interested to know

[rodabod]

2 hours ago, oldhippy said:

It shouldn't make any difference to the swing of the pendulum if you only have little power or if its on a full wind. 

In an ideal world, yes, the swing would be consistent. But this clock has largely varying rate according to state of wind, hence why I was asking questions regarding the swing of the pendulum.

[wls1971]

The clock movement is of H.Williamsons production, very nice movements the rear blued screws recessed into brass collets, its probably from the Salisbury factory so would date to 1903 to 1909 sometimes marked with Salisbury within a Key trademark, the factory burnt down in 1909 and production was moved to Coventry and these are stamped the same but with Coventry within a key the Coventry movements do not have the blued screws recessed into the collets.

I'm pretty sure yours would have originally had a chain and not a wire as the cone grove appears to be flat bottomed, Williamson did make wire fusee's but these where sold under the "Empire" brand of clocks.

This is a photo from the ebay listing of the clock I bought:

The pendulum on mine is the original, Williamson worked hard to standardise parts in their clock production and had just tooled up and spent a lot of money to upgrade the Salisbury factory, only for it all to be destroyed by fire, they where also under insured and could not afford to rebuild so moved production to their Coventry factory.

The problems with the clock above where one bent tooth and a chipped tooth on the escape wheel, the anchor had very slight wear and needed to be polished to a absolute mirror finish, the spring is on the strong side and the chain is very robust on these clocks, I put half a turn of pre tension on the clock enough to make the chain taut, it runs like a steam train and is very loud compared to any other clock I have, it runs slow to about 1 minute a week so they are very capable time keepers.

[rodabod]

2 hours ago, 4ster said:

Pendulum swing was measured with just two winds onto the fusee cone. I'll try measuring it at the start, mid point and fully wound.

I think this would be a good idea as it's easy enough to do. From your video, the drop did not look excessive. And there was not significant recoil. I'd be interested to see if the amplitude of the pendulum increases, or if there is noticeably more recoil when fully wound.

Even with a fusee in place to counteract the properties of a mainspring, it is only designed to counteract the properties of a given mainspring. At either end of the curve (fully wound or fully unwound) the curve becomes more exponential, so this would be very much exacerbated if an incorrect strength was used.

[oldhippy]

It is the mainspring that is causing the problem. You have to measure the recoil if it doesn't have a seconds hand, just by watching the recoil on a short video is not enough. 

[4ster]

wls1971,  thanks so much for the information and history on the movement, really fascinating. You are right the movement should have a chain, I just didn't want to go to the expense at the time.

rodabod & oldhippy,

More on the pendulum swing. I measured the swing at three stages of winding: 1, 7 & 14 turns on to the fusee cone.

Unwound 1 turn, swing = 36mm

Half wound 7 turns, swing = 32mm

full wound 14 turns, swing = 30mm

Sure looks like the cone does not match and equalize the properties of the spring

 

[4ster]

I have just taken a took at the escapement now the movement is near fully wound. As you can see from the video clip the is now very little recoil and less energy in the movement. Think I need to let it run down and take a look at just what type of spring is inside.

Pete

[4ster]

Sorry, video below