Wall Clock chiming problem

[Steve72]

Hello Everyone, first post about a first clock repair so please bare with me if I use the wrong terminology. I picked up a pendulum wall clock at an auction recently, basically house clearance type stuff so not much money, thinking if it works I will put it on the wall and if not it will be a fun project. It was listed as late 19th century, so I have no real idea of its real age or maker. I'm in the UK so 99% chance it is UK made. It was filthy and covered in oil so did not run. I stripped it down cleaned and reassembled. I have had it running continuously for over a week now and it keeps pretty good time.

My problem is with the chiming. It will every now and again miss a chime so that when it does chime, it strikes the wrong hour of half hour. For example it misses half past six and then strikes half six at seven o'clock, then seven o'clock at half past seven and so on.

It does not skip at any particular time. Sometimes it will skip once or twice a day and sometimes it will run for a couple of days without skipping at all. When I have been near it when it has missed a chime it only takes the slightest touch to one of the wheels arrowed (warning wheels?) in the picture below or even a little tap to the clock itself and it will turn again. I can see no obstruction and no rough or broken teeth. The uppermost wheel in the drivetrain that meshes with fly and the fly itself have a little wear at the pivot points. Is this just a case of a bit of wear causing a random jam?

Thanks for any help,
Steve

P.S. I have taken the hammer off in the pictures for a clearer view.

 

[SuffolkM]

Hi Steve

Compliments on the good photos, and nice clean clock!

How much rotation do you get on the warning wheel (should be about half a turn) before the strikes? If this wheel's position gives it only a small amount of warning rotation, or none, then the inertia of the fly can be too much for the strike train to start. You could be right on that limit. The warning wheel is the most likely thing as it is generally in a random setting when a clock is reassembled, and only by luck it might be in the right one! If it's not that then I'd also check for bent pivots toward the top of the strike train, eg. accidentally bent closing the plates. Even quite small deflections on the last few wheels and fly pivots can add enough friction.

Michael

 

[Willie X]

Suff's advice is sound.

The strike train could be stalling. It does this when it's not producing enough power or has to much load. If the fly is quick to start up the power is probably OK but this could be changing over time causing a no start now and then.

After checking the warn, as Suff mentioned. First thing I would do is to check the hammer lift. It should be about 1/2". If more, this can be reduced by raising the hammer stop, or by adjusting the hammer tail, according to how your clock is made. Often the stop has been beat on for so long the hammer lift has become two or three times what is is supposed to be.

If all this checks out, worn bushings, bent teeth, bent arbors, bent pivots, etc. would be some of the likely suspects.

Just watching the clock run over time, prodding the suspected areas wilh a bamboo skewer, will often (eventually) reveal yer problem.

Willie X

 

[roughbarked]

It is a nice clock too. Try not to bend any pivots.

 

[JimmyOz]

Looking at the photos, it looks to me that the count lever is not all the way in the cut out for the count wheel, then looking at the front photo the lever is sitting up even with the spring on it. I would re-bend the spring in the front photo so it is putting pressure on the lever and not on the arbour.

 

[Steve72]

Thanks for the advice folks. I will have a look at the points mentioned and see what happens.

Steve

 

[shutterbug]

To me, it looks like the pinion driving the count wheel is barely engaging, and might be skipping teeth.

 

[Steve72]

SuffolkM, If I understand you correctly the top warning wheel rotates just over a half turn before it engages with the back of the V shaped lever controlled by the minute hand on the front of the clock. This happens every time the clock gets ready to strike without a problem. When the lever drops to allow the clock to strike, this is when it will jam, with the wheels not moving and staying still.
Everything is clear of the pins on the warning wheels and there is nothing to see that will stop them from turning but something obviously is. I hope the drawing helps you to see the relationship between the wheels better than I can explain it.

WillieX, Sometimes the train starts quickly sometimes it is slower and speeds up. All the pivots are as far as I can see true. This makes me think something is binding but I can not see what. I have not cleaned the spring in the barrel. I did look at it and decided I would wreck it if I tried taking it out, but it does not seem weak to me. The hammer lifts about half an inch.

JimmyOz, the count lever does sit fully in the count wheel cut outs normally. I think I took that picture when it was ready to strike.

Shutterbug, all the wheels mesh properly. The picture might be misleading, sorry.

If it is needing to be re-bushed I can live without a chime as I do not think the value of the clock is worth having it done.

Thanks again for the advice,
Steve.

 

[SuffolkM]

Hi Steve,

You must be due a certificate for all these lovely diagrams!

Regarding the spring, what sort of condition was it in when you had a look? If it looked like it was full of black goop, or worse green scale, that is probably a bad sign. On the other hand, if it looked basically clean and not caked in dirt that would bind the layers together, and it seems to take your turns on the winding arbor without jolts etc. then it is likely ok, producing enough power at least for the moment.

This clock does not look like it needs bushing to me (although you can only really tell looking from the inside of the plates outwards). But bent arbors or pivots can still be very elusive. If you are willing to disassemble, I'd suggest a progressive build of the strike train to eliminate this. I do this when a clock is being mysterious, because it discloses all the flaws quite quickly. The aim is to get individual wheels turning fast and quietly. They won't do that convincingly if there's a bent part, a deeply scored pivot, a wonky bushing repair, etc. You may hear noises and the wheel will certainly come to a halt. Caught red handed.

In this case we could say we're just looking at the strike train and so you can leave everything from the going train aside. Skip the mainspring, too (you drive with a finger in this test). Start by putting each wheel into the plates in the sequence of the train, starting with the second wheel after the mainspring. Spin the wheel between the plates, looking for it to freewheel for perhaps 10s regardless of how you orientate the clock (longer is easily possible). If it seems to stop rather abruptly or makes rattling noises, then that wheel, either its pivots or the pivot holes or maybe a bent arbor, is a problem. If it seems totally fine, you can remove that wheel and then go to the next one (or you can test two at a time, which is good for listening to the pinions). By the time you get to the fly, you have tested everything. This might sound like a lot of work, but you can add a few parts between the plates at a time, and it's considerably quicker than closing the plates on a full build. Even putting the whole strike train in, and driving it round with a finger, will give you confidence that it all turns without a lot of force when it's healthy. As I mentioned, this sort of approach leaves nothing untested and can help you bust a mystery wheel sticking like a powerful search light.

Someone else might chip in with a more specific suggestion for you, but it's an option I'd be mulling over!

Michael

 

[bangster]

Here's what I think.
The count wheel has no accommodation for a half-hour strike, so it's a passing strike and we can ignore that.

When it misses a regular hour strike, it's because the count wheel didn't complete its cycle: the count lever is resting between notches. So something has interrupted the strike train in mid-cycle. The simplest way to halt a strike train is for the stop pin to encounter the stop lever…when it's not supposed to. That's where I'd look: at the stop wheel and stop lever. It might just need something as simple as resetting the stop wheel a tooth or two. But at any rate, that's where I'd start looking.

 

[bangster]

Well?

 

[Steve72]

I took the clock apart again and have tested all the wheels as described. All run freely for some time apart from the fly which turns easily but does not spin for anywhere near the length of time as the other wheels. I don't know if it should or not but I can see no damage to it. Having said that I have loosened the fly spring a touch to hopefully allow it st start a little easier.

The stop is clear of the stop lever all the times I have been able to look at the clock when it has jammed.

I have set it going again and for now it is behaving, although a little quickly for the first couple of chimes as I have loosened the fly. It has only been running for a day so will see what happens over the rest of the week.

Thanks again.

 

[shutterbug]

I think you may have gone the wrong way on the fly. It should spin as long as the other wheels on the strike side. If it doesn't, it usually means the arbor is spinning but the fly part is not ... so it needs to be tighter, not looser.

 

[Steve72]

A little update. The clock ran perfectly for about four days then started to miss the occasional strike again. I have re-tightened the fly again to see what happens but it has made no difference, in fact the first two or three strikes run so quickly that it is hard to hear them individually. After a day or so it has slowed down a bit now, maybe this is just because I have wound the spring up fully. Is it possible that the fly arbour is too smooth with wear?

 

[Willie X]

Sounds like the fly is still to loose. It's not that obvious but a loose fly might be your only problem. The meter of the striking should never vary. Well, maybe a little slow after a week's run but even then the meter should be exactly the same for every count. Willie X

 

[SuffolkM]

Please can you tell us about the condition of the springs, Steve? I am going to assume you didn't service them (possibly they remained in the barrels uninspected?). Although your mainsprings do produce more torque when they are fully wound, so it's not unexpected that with a loose fly the strikes set of faster after a full wind, it is just possible the condition of the middle portion of your springs is set. I wouldn't necessarily dive in here and change the springs, but as I was commenting in an earlier post, if they are bound with old gunge (you might hear and feel strange slips when you are winding) then this could rob the strike train of power.

Tip: to remove the barrel caps, tap with a nylon or leather hammer on the back of the winding arbor (this is better than prying off the lid and much quicker). You can twist the winding arbor anticlockwise to get it out of the centre of the spring, and then send some photos so we can take a good close look.

Safety note: don't try and get the springs out of the barrels if you haven't got a mainspring winding tool! (Edit: I am in the UK and can help you with this if it turns out to be your problem).

As you've worked through rather a lot of details and nothing has really stuck out yet, the spring could be it...

Michael

 

[Steve72]

OK, I'll take it apart again and take some more pictures when I get the chance.


Thanks.

 

[shutterbug]

The fly has to be removed from the arbor in order to tighten it. It won't work otherwise.

 

[Mike Phelan]

Weak or sticky mainsprings is it's OK for a short while after winding?

 

[Willie X]

Mainsprings probably do need to be serviced but I can't see where this is a power related issue. Willie X

 

[Steve72]

I have taken a couple of pics of the mainsprings. Striking spring on the left and time on the right. I know the teeth look like they are chipped and worn at the edges in the pictures but they are not.


I did open the springs when I was cleaning the clock and wiped some surface grease/old oil from the top of the coil and around the barrel but did not want to go any further than that. I left what old lubricant that could be in the spring where it was.

It did work for 4 days recently from a full wind before being problematic again but the problem seems to be completely at random.

I managed to break the fly spring in half when I was straightening it to tighten the fly again. I was worried that it might fatigue easily and it did. I have temporarily replaced it with a thin piece of brass sheet and it seems to be holding for now. The clock is striking at a steady rate now and I will see how long that it runs for this time.

If the problem is the fly, I can understand the wheels not turning when it is too tight, but why would they not turn if it is too loose?

Thanks again for your time.

 

[Mike Phelan]

First spring looks a bit "tired" and probably needs taking out, cleaning and oiling with thick oil before putting back in barrel.

Second spring is too long and cannot wind sufficiently.

 

[SuffolkM]

The number of spring laps is around the same in both (around 16 to 17) despite how it looks. Having said that, it does seem like too many turns - maybe, quite a bit too many based on the springs I've handled recently. Checking a few photos where I could see springs I've replaced recently, I count 8 or 10 turns, but not more. Willie will have a better instinct for this and confirm I'm sure.

Bad luck on that fly spring, but good to hear you were able to solve it as well.

Michael

 

[Mike Phelan]

The number of spring laps is around the same in both (around 16 to 17) despite how it looks. Having said that, it does seem like too many turns - maybe, quite a bit too many based on the springs I've handled recently. Checking a few photos where I could see springs I've replaced recently, I count 8 or 10 turns, but not more. Willie will have a better instinct for this and confirm I'm sure.

Bad luck on that fly spring, but good to hear you were able to solve it as well.

Michael

In which case, that spring is probably thicker. What do you mean regarding "number of laps"? The number of turns to wind it is usually about five - if the spring is either too long or too short it will be less.
Looking at the picture the right-hand spring is definitely wrong. When a spring is new, in the barrel it should occupy about a third of the distance between the barrel and its arbor.

 

[Willie X]

Left spring to narrow, right spring to long. So, I guess the Evil Spring Fairy has already visited your clock!
Look up "Willie's turns of power". This can give you a good idea of what's going on there with those springs.
Even though your springs are screwed up, I don't think it would affect your original strike problem, at least not until after several days of run time. Now that narrow one might not be strong enough to give you reliable start ups, especially if it's narrow and thin too. Willie X

 

[shutterbug]

Here's a handy calculator to help you determine what you need. I'd try something around .0165" for the thickness.

 

[SuffolkM]

Steve, you can get replacement springs rather easily from HS Walsh or Cousins in the UK. If you are not equipped to change them, drop me a note so I can save you the horror and peril of trying without the correct tools.

 

[Mike Phelan]

Ditto here - we're 2 miles from M62 J26

 

[Steve72]

Willie, I could not find anything under "Willie's turns of power" when I did a search, is it on this forum or did you mean google.

Do you think that changing the springs would make a difference to how the clock is running? I understand changing them to the correct size is something worth doing but would it help the problem? Out of curiosity, should both springs be the same depth as the barrels?

I have had a look at the mainspring length and thickness calculator and I don't know if my clock would be a 4 or 5 wheel layout as it is set out differently then the diagram beside the calculator. Here is a picture of the layout, before cleaning and minus one wheel from the striking side. Can I use the same calculator for the striking spring?


Thanks.

 

[JimmyOz]

It may be my eyes, however the fly looks odd where the arbour/pivot is, it looks fat, maybe someone has made a bad job of re-pivoting?

 

[Steve72]

I think it is just that the end is out of focus. Here is another pic of it.

 

[JimmyOz]

You are right, that looks okay.

 

[shutterbug]

For the most part, the springs should be about 1mm lower than the cap. In other words, almost as wide as the barrel is deep. The thickness of the spring and the width have a lot to do with the amount of power imparted. The length is calculated to allow the spring to wind and unwind properly within the confines of the space allotted by the barrel. The thickness I mentioned before is a "typical" number that might work. But that number can be tweaked a bit, and will affect the length that the calculator suggests. Thicker is stronger, and some clocks use different thicknesses for the time and strike trains. In your case, I suspect they will be the same. You could look up the thickness that Hermle uses for their two train movements, and match that if you want. They should be pretty close to what you need.
Your post #21 shows one spring about the right width, and the other considerably short. That short one was likely a replacement with something the guy had on hand. You could probably match what is in the other barrel and be safe.

 

[Steve72]

I have been trying to work out the length and thickness of springs to replace the ones I have and was wondering if what I am doing is along the right lines. So here are some more details.

The barrel is 34.5mm wide and 20.5mm deep (from the bottom of the cap seat).
The arbor is 9mm wide.
The mainspring is .45mm thick and 18mm high.
The striking spring is .35mm thick and 14mm high.

If I enter the numbers for the mainspring I have into the spring calculator on this site it suggests a length of 967.74mm/38.1" but the only springs I can see on the Cousins site that are near that length are only .25mm thick or are listed for a 400 day movement (which strangely enough I was wondering what that type of clock was called only the other day).

Am I making a mistake somewhere or am I being too hopefull that there will be a modern spring which will slip perfectly into an old irregular sized barrel?

Thanks.

 

[Willie X]

The thicknesses you have there are all wrong too. The strike is usually thicker than the time. Bug's estimate is good.
I would probably go with something near .017' for the strike and something .0155" or .016" for the time. Willie X

 

[SuffolkM]

Hi Steve,

My little spreadsheet that accounts for arbor sizes etc. suggests that the length of your spring needs to be around 1250mm. That's got big error bars either way and is just approximate, so you could say it's anything from 1100mm to 1400mm quite reasonably.

I'd aim for a 19mm high spring in both barrels. The extra 5mm of height that entails in your strike mainspring is a very considerable extra bit of power compared to the 14mm spring. Also to add, if you use springs with much less height than the barrel they can take on strange contorted shapes inside the extra space. It's unlikely the maker would have made the barrel unnecessarily large given the waste of materials that would entail, either. Have a look at where your winding arbor catch is. If it's in the centre of the barrel when the arbor is seated in its normal location, then that suggests the spring centre is half-way as well (useful supporting evidence).

I would try part 194336A (19 x 0.43 x 1143 x 36mm). That's about 0.017" thickness if you're reaching for a calculator to check Willie's suggestions, too! Don't worry too much about the 36mm seeming to be larger than the barrel diameter here as they do allow a bit of extra. Alternatively, part 193535 (19 x 0.35 x 1400 x 35mm). You'll find it winds a bit more easily than the fatter spring, but all the power is still there as it's longer too.

It's tempting to try and supercharge your clock with the biggest springs you can crowbar in, but it's not the right answer. There is no need to over-power your clock as it just increases wear, and although you might like to investigate different thicknesses, I suspect you will have no problem with 0.35mm on both trains here. If you want, you could try 194336A on just the strike train perhaps.

Michael

 

[Steve72]

Thanks for the replies again folks it is extremely helpful.